Sam's Laser FAQ, Copyright © 1994-2009, Samuel M. Goldwasser, All Rights Reserved.
I may be contacted via the
Sci.Electronics.Repair FAQ Email Links Page.

  • Back to Sam's Laser FAQ Table of Contents.

    HeNe Laser Testing, Adjustment, Repair

    Sub-Table of Contents

    Flavio's Comments on HeNe Tube Mirror Alignment

    (Portions from: Flavio Spedalieri (

    Below is a simple diagram that shows the end configuration of a typical internal mirror laser tube:

    	 \ __   __
    	--|  |_|  |-
              |  |_|  | |====> Laser Beam
    	--|__| |__|-
    	 / ^    ^
           /   |    |
               |    +--- Adjustable part of mirror mount
               +--- Fixed part of mirror mount

    The end of the mount is divided in two with a gap between the first and second sections. At the time of manufacture, HeNe laser tubes are aligned for optimum power output.

    On some HeNe tubes (as well as internal mirror argon ion laser tubes), this gap may me covered by a ring with three (3) adjustment 'grub' screws as shown below:

           \     ___
    	 \ _|   |_
    	--| |   | |-
              | |(X)| | |====> Laser Beam
    	--|_|   |_|-
    	 / ^|___|^   (X) denotes adjustment 'grub' screw (1 of 3 shown).
           /   |  ^  |
               |  |  +--- Adjustable part of mirror mount
               |  +--- Ring with three (3) grub screws
               +--- Fixed part of mirror mount

    Or see A HREF="3slcmg1.jpg">Three-Screw Locking Collars on Melles Griot HeNe Laser Tubes for photos.

    If the tube has the metal ring with the grub screws, some people have been tempted to re-adjust this - very BIG mistake.. and the reason is this: With the ring in place and the screws tight and sealed from the factory, the whole assembly is very solid. Now, if you try to re-adjust the grub screws, trying to extract more power, more than likely you will throw out the entire tube out of alignment. The screws are so tight, that very slight, and gentle and precise alignment is very difficult to achieve.

    For tubes that DO NOT have this assembly, once the mirrors are out of alignment, it is extremely difficult to re-align the tube. Been there, done that. :(

    Now, the reason that there are troubles with realigning a tube so it is stable are two-fold:

    1. Mirror mounts are very difficult to physically change position (very slight movements) and maintain that position. You are physically bending metal so it is easy to overshoot the desired position. In addition, the mount will not relax to its final position and stay there - there may be some drift or creep over time especially after multiple thermal cycles.

    2. Heat - even if the tube is only on for a few seconds, the very slight temperature differences can be enough to change the mirror alignment. At the factory, the alignment is optimized after a warmup of 30 minutes or more and minimum output power is usually specified after a similar period of time.
    In some older HeNe tubes, I have seen the affects of thermal expansion, the beam will drift in and out of alignment, and this start to occur only after about 5 to 10 sec after power-up.

    WARNING: All the adjustments that you do on the tube, unfortunately have to be done while the tube is powered up - so you have at least one end of the tube (usually the anode) floating at 2 kV or more once the tube is running (and even after power down due to tube and power supply capacitance). If during your adjustments, the tube decides to drop out, and re-start, you will have the 8 to 15 kV starting voltage - so please be very careful!

    As the tube is powered, try and push the mirror mount, and watch the beam, once you get a nice bright output, try and hold that position, and see if it will hold the output as you support that position - Note in which direction / movement you used to achieve this.

    If you have the ring/grub screw assembly, moving one of the screws will not necessarily adjust the mirror in the direction that you want, so you may have to use different adjustment/pressures on all three screws.

    (From: Sam.)

    If the alignment is nearly correct - gentle force or just touching the mirror mount results in full power - I would suggest as an alternative: Instead of actually attempting to bend the mount, add an external 3 screw adjuster to the problem mirror mount. This will operate within the elastic limit of the mounts so the risk of breaking them off from repeated unsuccessful attempts at bending them back and forth is eliminated. Let the tube warm up for at least 30 minutes, then gently adjust the screws to optimize power output.

    Rich's Procedure for External Mirror HeNe Laser Alignment

    As written, this would appear to be apply more to determining if a combination of HeNe tube and mirrors will lase. Modify as appropriate where you are doing this with an existing laser.

    (From: Richard Alexander (

    1. Use a good optic axis (very important). A good rail is worth the money.

    2. Use a second Hene laser. I pity those who lack this option.

    3. Use good adjustable mounts.

    4. Mount everything except the mirrors onto your optic axis.

    5. Once you have the beam of the 2nd HeNe shining down through the tube of your 1st HeNe (and out the other end), mount the far mirror on the optic axis.

    6. Adjust the mirror mount so that the beam of the 2nd HeNe reflects back through the tube of the 1st HeNe and strikes the 2nd HeNe next to the Output Coupler of the 2nd HeNe.

    7. Put the other mirror on the optic axis.

    8. If your tube is functional, you could apply power to it, and then fiddle with the last mirror until you get a beam.
    With practice, this method can be completed in less than an hour, though 4 or 5 hours is not unusual, either. If you have a 3rd HeNe, or better equipment than we had in tech school, you might get done much faster. Rich's Procedure for External Mirror HeNe Laser Alignment As written, this would appear to be apply more to determining if a combination of HeNe tube and mirrors will lase. Modify as appropriate where you are doing this with an existing laser.

    (From: Richard Alexander (

    1. Use a good optic axis (very important). A good rail is worth the money.

    2. Use a second Hene laser. I pity those who lack this option.

    3. Use good adjustable mounts.

    4. Mount everything except the mirrors onto your optic axis.

    5. Once you have the beam of the 2nd HeNe shining down through the tube of your 1st HeNe (and out the other end), mount the far mirror on the optic axis.

    6. Adjust the mirror mount so that the beam of the 2nd HeNe reflects back through the tube of the 1st HeNe and strikes the 2nd HeNe next to the Output Coupler of the 2nd HeNe.

    7. Put the other mirror on the optic axis.

    8. If your tube is functional, you could apply power to it, and then fiddle with the last mirror until you get a beam.
    With practice, this method can be completed in less than an hour, though 4 or 5 hours is not unusual either. If you have a 3rd HeNe, or better equipment than we had in tech school, you might get done much faster.

    Dave's General External Mirror Alignment Techniques

    (From: Dave (

    This is my new method of laser alignment. This works well for most narrow bore HeNe and ion lasers. As of today, it is the best yet. :-)

    Ever hold a HeNe or other laser tube in your hand and just hold it up to your eye and sight through the bore and look at something across the room and target it? Quite easy to repeat. I always wished I could shine a laser beam down the same tube with the same accuracy and speed. Especially when trying to align a laser!! I have aligned quite a few lasers over the years via this same tedious method and to be frank, I am sick of it. :-) In the beginning of the hobby I really enjoyed doing this and worked it down to a science but it is still a pain , all that laser light splashing all over the place , fiber optic effect of the light zig-zaging back and forth the bore etc. Well this method works for me and I'm sticking with it :-)

    You will need:

    Here is the procedure:

    1. Remove OC and place light bulb locked in the 3rd hand 8" to 14" in front of the laser where you removed the mirror.

    2. Look through the HR and down the bore, and line up the bulb filament so it is centered, this should be quick and easy. :-) In my case laser is on my right pointing left at the bulb on my left, lined up.

    3. Hold LED flashlight (diffuse source) through HR down the bore.

    4. Place glass slide (beamsplitter) between bulb and laser and orient at 45 degrees so you can sight the LED flashlight, again real quick and easy. :-) Lock in place with 3rd hand, or in this case with 4th hand. :-)

    5. While looking through the slide, twist HR mirror mount adjustments til you see the bulb filament through the slide, center it, done. Again quick and easy. :-)

    6. Remount OC, fire up laser and fish for the beam, your done.

    I like this method because all critical alignment is accurately sighted directly and quickly by eye. This satisfies my natural wanting to look directly down the bore and immediately align the mirror directly by eye. :-)

    I can't believe I haven't tried this before.

    I read the HeNe laser in SciAm and it is pretty much the same setup, but they do not mention to shroud the Brewsters which helps greatly to maximize the contrast. Shrouding the SP-907's Brewsters made it a snap. :-) Yup, tried it a couple of times on the SP-907 and it works awesome. I use 1 steering mirror with the 907 for the tube is too long and this way I sit at the HR-end and tweak while looking down to the mirror.

    Dave's Quickie External Mirror Alignment Technique for the SP-120

    The following works for the Spectra-Physics model 120 and other lasers with spherical OCs where the optics and machining are most excellent. Interchange OC and HR in the procedure below if your laser only has a spherical HR. I doubt it would work reliably depending only on close tolerances for a planar mirror.

    (From: Dave (

    As far as the terrible 3 point mirror mounts on the SP-120, I have developed a way to get the mirrors aligned without any cards or another laser. Just my two hands and a hex wrench. Within 5 minutes I get it every time. :-) I have also been applying this technique to the longer lasers with some good results.

    As you know, if one mirror is aligned correctly, the other is a cinch. I tighten down the OC and then back off each screw 3/4th of a turn. Then I loosen up the HR so it has a lot of play. I put my finger over the HR and wiggle in a repeating all over the place while hunting for a flash out of the OC. When I get a repeatable flash on the OC that's it, no problem to tweak in the HR. Works every time on the SP-120. :-)

    Dave's Preferred External Mirror Alignment Technique for Long Lasers

    For this procedure, both mirrors are left in place. The OC of the Laser Under Alignment (LUA) is facing a Reference Laser (Ref-L). Note that like many of the other techniques, this does require that the Ref-L's beam is sufficiently narrow and collimated that it can get through the LUA's bore with at most minimal wall contact. To what extent this is possible will depend both on the beam characteristics of the Ref-L and the curvature of the LUA's HR (which affects the size/divergence of the reflected beam).

    1. Set up the Ref-L on a scissor lift with XY adjustment and the aperture of Ref-L and LUA approximately 6" apart. For aligning a red HeNe, the Ref-L should be some other wavelength like green which will pass more easily through both of the LUA mirrors.

    2. On opposite end of LUA, a convex lens is held in a "3rd hand" projecting onto a white card or the wall to expand the emerging beam to say an inch or more so you can see what you are doing way over on the other end of the setup while getting the Ref-L perfectly aligned with the bore.

    3. Once the bore is perfectly aligned, insert a glass slide held in another "3rd hand" in between the lasers and reflect the light coming from the laser to be aligned up onto the ceiling or wall or screen.

    4. What you will see reflected if the LUA is totally out of alignment are two spots reflected back from the OC: The mirror surface (bright spot) and the AR-coated surface (dim spot).

    5. Pick an axis and rack this mirror (OC) a few turns to get its reflections out of the way.

    6. Take a walk to the other end of the laser and start twisting the HR mount till you get a 3rd spot projected on the ceiling (preferably over the center of the laser. :) Center this spot for a good clear round spot and if there is some scatter lopsided about this spot, it is OK (bore is slightly out). You want the spot to be round and clear.

    7. The best part: Superimpose the bright OC spot over the HR spot and suddenly there will be an eruption of laser light from the laser that was out of alignment. :-)

    This procedure is nice and easy to perform but even better, REAL EASY to SEE what's going on - no squinting down the bore to look for a light bulb you can't reach. :-)

    Cat's Eye Mirror for Hassle-Free Alignment

    Haven't you always dreamed of just dropping the mirrors into an external mirror laser cavity, power up, and have the laser operating at near optimal performance without ever touching an alignment screw? Well, there is a relatively simple optics configuration that has the potential to make this possible

    While many large bore lasers like the M60 Tank ruby rangefinder laser have used optical roof prisms or even corner cube reflectors for the HR, this isn't practical for narrow bore HeNe and ion lasers. Wny? Well, for one thing, no roof prism or corner cube has perfect edges so there will be some scatter in the region where they are - but for a laser with a 1 mm diameter beam, that's a relatively large percentage of the mode cross-section and effectively kills lasing.

    However, there is a combination of a curved mirror and convex lens called "cat's eye" due to its similarity to the arrangement in, well, a cat's eye. This behaves much like a corner cube reflector but without its problems (at least over a small angle). Rays entering the lens will be reflected directly back in the direction they came, at least close to the optic axis for small angles. The optimal arrangement has an AR coated convex lens placed at one focal length (f) from a mirror with an RoC of f, coated as an HR or OC for the laser wavelength. In principle, any narrow beam laser could benefit from this. The cat's eye reflector has been tested with HeNe lasers but would certainly work as well for other narrow bore lasers - which are those creating the most problems with mirror alignment. Of course, the manufacturing cost would be higher but how much is your time and sanity worth? :)

    Apparently, using such a setup allows the mirror assembly to be held by hand or with a pair of tweezers and get stable lasing. Now, I can do this with my 1-B HeNe laser tubes and a normal HR or OC, the cat's eye makes it even easier. ;-) Of perhaps more importance, since angular sensitivity of the mirror is greatly reduced, it would be possible to say goodbye to the annoying power drift due to alignment changing that often occurs as the laser warms up.

    This was presented in the paper: "Adjustment-free cat's eye cavity He-Ne laser and its outstanding stability", Zhiguang Xu, Shulian Zhang, Yan Li, and Wenhua Du, 2005 Optical Society of America, Optics Express, vol. 13, no. 14, pp. 5565-5573, 11-July-2009.

    Here is the abstract (spelling and grammer NOT corrected):

    "This paper introduces an innovative He-Ne laser which exhibits many advantages to current He-Ne lasers. With cat's eye reflector as the reflecting mirror, the new laser can solve the conventional problems of laser adjustment and power stability. Comprehensive experiments are carried out both in a half-external cavity and a full-external cavity He-Ne laser. Then the results from the cat's eye cavity, plane-concave mirror cavity and concave-concave mirror cavity are compared, which show that in halfe-xternal cavity laser, cat's eye cavity can improve the laser stability up to 10 times better than other cavities and lower the power drift significantly. And in the full-external case the improvements are much greater even up to 60 times and power drift is minished greatly too. The adjustment problem is also considered and solved. A stable and adjustment-free He-Ne laser is finally realized. The examination of a cat's eye reflector is described."

    Now, while it's quite likely that the benefits of the cat's eye configuration were recognized long ago, the added complexity and cost - and especially the losses through the AR coated lens - would likely have prevented it from even becoming widely used. Only with a very long HeNe laser like a Melles Griot 05-LHR-927 or Spectra-Physics 127 would those losses be relatively small compared to the gain. But even so, would still result in a significant reduction in power, not to mention the issues of making sure 2 additional optical surfaces are perfectly clean. One way around the loss problem might be to use an aspherical HR mirror reflecting off-axis to the spherical cavity mirror instead of a convex lens, but then the cost of manufacturing that very special mirror would probably be totally ridiculous.

    My question - which I'm not sure is answered in the paper - is: What happens if cat's eye reflectors are used at both ends of the laser? Is the thing then totally self-aligning? :)

    Mirror Alignment with just an Optical Power Meter

    Here's a way of aligning mirrors very quickly on small to medium length external mirror (one and two-Brewster) lasers. With care, it may work on large frame lasers as well. This approach takes advantage of the fact that the far mirror will be aligned when maximum bore light is returned to the front of the tube. Unlike the laser wavelength, most bore light passes through the mirror and there is ample power to monitor. It works great on one-Brewster HeNe lasers as well as the very difficult to align PMS LSTP-0010 and LSTP-1010 tunable HeNe lasers. I would recommend it only for lasers with screw adjustable mounts, not for HeNe laser tubes with three-screw locking collars or less.

    All that is needed is an optical power meter (laser, photographic, etc.) with enough sensitivity to respond well to the bore light. One with a "suppression range" feature is best but this is not essential. (The suppression range enables the constant light to be cancelled out so that the sensitivity to changes can be increased.)

    To align one mirror, place the sensor of the optical power meter at the other end of the laser, located to pick up the bore light. Set up the meter on a range that allows the maximum deflection of the meter while keeping it on scale, and/or set the suppression range to cancel out most of the constant bore light.

    Now all that's required is to twiddle (technical term!) the far mirror to maximize the power reading. With kinematic or gimbal mounts, this will actually be quite easy. The peak is broad so each axis will have an effect even if the other axis is way off. As the alignment approaches optimal, the reading will increase and with a bit of luck, will then spike as lasing occurs (assuming the other mirror was already aligned).

    For a laser with two adjustable mirrors, just repeat the procedure for the other mirror.

    It takes literally only a couple of minutes to do this for a PMS tunable laser (which uses a 1-B tube with permanently adjusted internal OC), which with its narrow bore is very difficult to align with any of the other techniques.

    Inconsistent Behavior of HeNe Laser Alignment

    With a healthy HeNe laser, the adjustment of mirror alignment will result in behavior that is smooth and repeatable. Except for the mode cycling variations in output power, the result will be fairly stable even from one power cycle to the next. Even the very rudimentary adjustments using the screws on Melles Griot locking collars perform generally in this way. Unless....

    You may come across a laser tube or head where nothing works as expected. After peaking power, the output may drop after a few minutes such that adjustment is again needed. After that, the same thing may happen again. And again. The output power may be extremely sensitive to mirror alignment even to the point were gently clamping the tube in a head cylinder using the nylon screw may cut output power in half or worse. Or, supporting the tube at various points will significantly affect it. And just the weight of a popsicle stick on one of the mirror mounts will change power significantly. What's going on?

    If the laser is from a surplus supplier or eBay, it's quite possible - actually quite likely - that it was a reject, possibly due to a bad design (yes laser designers make mistakes!) or improper manufacturing. In particular, if the mirror specifications were not correct and matched to the bore, the stability of the resonator with respect to the various modes could be so low that each one sees a significantly different gain. So, after optimizing one set of modes, as they drift with respect to the gain curve, there could be very significant power fluctuations. Guess where such tubes end up? :)

    Another possibility is contamination like a hair, fiber, or metal sliver, inside the tube. If it extends into the lasing mode volume (the intracavity beam), then peculiar behavior could result, and could change with time, orientation, vibration, etc.

    The effects of IR (3.39 um) mode competition can appear similar but are not likely to show up with most reasonably modern red (632.8 nm) HeNe lasers, though they can be significant for "other-color" tubes.

    I have several tubes that exhibit this sort of behavior. One is a Melles Griot 05-LHR-990, a 10 mW (rated) tube about 18 inches long. It was obtained in a batch of tubes I bought from one of the well known laser surplus outfits mainly to salvage mirrors. So this tube was even considered a reject by them! (It only cost me $2.) The output power is extremely sensitive to any pressure on the mirror mounts (even with the locking collars tight), pressure applied to the sides of the tube with the nylon screws in the head cylinder, and temperature gradients. By adjusting the locking collars, it's possible to achieve over 14 mW by careful tweaking. However, after a few minutes, the power declines to 12 or 13 mW and realignment of the mirrors at one or both ends is required to get back to the high level. The power is still well above the spec'd value but the variation is annoying. It's a nice tube otherwise. :)

    Another Melles Griot tube I have with a similar but more severe symptoms is also of similar size, though I don't know the exact model number. It's output can vary from 4 to 9 mW with almost no change in mirror alignment, and may switch to a multi-transverse mode (TEM01/10 or something stranger) at the lower powers. I rather suspect that there may be something somewhere in along the length of the bore or inside one of the mirror mounts though I can't find it. There is what looks like hair stuck inside one of the mirror mounts but it doesn't extend anywhere near the beam. But, perhaps, it's buddy is hanging out somewhere else.

    Cleaning and Alignment of the Spectra-Physics 907 HeNe Laser

    This is longest HeNe laser that had been available from Spectra-Physics (now Newport) and the model 127 which uses this tube was listed on their Web site and in the Edmunds catalog until recently. However, many of these are still available surplus either as the SP-127 or just the tube designated 907, 107, or the older 082. The design hasn't changed noticeably in 25 years.

    Unfortunately, one of the deficiencies of this laser is that the probability of it remaining aligned during shipping, even if packed with 10 inches of foam all around - is small. Usually, it's just a slight power loss but I've seen cases where there was no lasing at all and a full realignment was needed.

    Being external mirror lasers with Brewster windows, the optics can get dirty, especially if the rubber sealing boots at each end are cracked, as some tend to be, possibly from overzealous removal by a previous user.

    Sam's Procedure and Comments

    There are 4 sets of alignments in the SP-907 (listed in the order in which they should be dealt with for a total realingment):

    Recommended cleaning

    None. :) Actually, if the rubber boots are in good condition and sealing well and there is no reason to suspect that someone before you has messed with them, it's probably best to leave them in place and not to attempt to clean the Brewster windows or mirrors, at least not until you're determined to eak out the last few photons/second of performance. But here are some procedures that work:

    As noted, put the boots back on as soon as possible as a gradual decline in power is inevitable from from dust collecting on the B-windows and mirrors. For a laser like this, even slight contamination not obvious by eye can result in a power reduction of a few mW.

    Total alignment procedure

    Touch up the mirror alignment after the bore centering.

    Keep in mind that all of these adjustments interact to some extent. So, it's possible to be at a local maxima and lose sight of the global optimization. For example, changing the position of one of the bore centering adjustments may reduce the power initially, but adjusting the other one and realigning the mirrors may get it back and more. These are third order effects though, so doing the procedure above should get you most of the may to a happy laser. :)

    Michael's Procedure

    A complete sequence of steps for cleaning and realigning of the SP-907 from a non-lasing condition can be found at: Dragon's Eye Doings: Laser Alignment Heck. This includes descriptions and many photos. My mirror of the content of this may be found at Sam's Copy of Dragon's Eye Doings: Laser Alignment Heck. This isn't the only way to do it (see my summary in the previous section) but is known to work and is straightforward. Bore straightening isn't covered but can be dealt with as an add-on. :)

  • Back to HeNe Laser Testing, Adjustment, Repair Sub-Table of Contents.

    Collimation of HeNe Lasers

    Reasons for Poor Collimation

    The output of most HeNe lasers is a very well collimated beam - approaching the theoretical diffraction limited optimum possible for a given bore diameter. It isn't expected to look like that of a flashlight! However, a number of factors can affect this performance - and some are by design:

    Improving the Collimation of a HeNe Laser with a Beam Expander

    The following applies to any laser which outputs a substantially parallel beam but is written specifically for HeNe lasers. Collimation of laser diodes require a slightly different approach - see the section: Beam Characteristics of Laser Diodes.

    Although the divergence of a HeNe laser is already pretty good without any additional optics, the rather narrow beam as it exits from the tube does result in a typical divergence between 1 to 2.5 mR (half of total angle of beam). 1 mR is equivalent to an increase in beam diameter of 2 mm per meter.

    As noted in the section: HeNe Laser Tubes and Laser Heads, beam divergence is inversely proportional to the beam diameter. Thus, it can be reduced even further by passing the beam through beam expander consisting of a pair of positive lenses - one to focus the beam to a point and the second to collimate the resulting diverging beam. Though the beam will start out wider, it will diverge at a proportionally reduced rate.

    A small telescope can be used in reverse to implement a beam expander to collimate a laser beam and will be much easier to deal with than individual lenses. (This is how laser beams are bounced off the moon but the telescopes aren't so small.) Using a telescope is by far the easiest approach in terms of mounting - you only need to worry about position and alignment of two components - the laser tube and telescope. The ratio of original to expanded beam will be equal to the magnifying power of the telescope. Even a cheap 6X spotting scope will reduce divergence six-fold.

    If you want to use discrete optics:

    For optimal results, the ratio of collimating lens diameter to focal length (D2/F2) should greater than or equal the ratio of HeNe beam diameter to focusing lens focal length (D1/F1). This will ensure that all the light is captured by the collimating lens.

    The beam will be wider initially but will retain its diameter over much longer distances. For the example, above, the exit beam diameter will be about 10 mm resulting in nearly a 10 fold reduction in divergence.

    Adjust the lens spacing to obtain best collimation. A resulting divergence of less than 1 mm per 10 meters or more should be possible with decent quality lenses - not old Coke bottle bottoms or plastic eyeglasses that have been used for skate boards. :-)

    Note that some HeNe tubes have wide divergence by design using an external negative lens glued to the OC. For these, removing this lens with a suitable solvent may be all that is needed to produce the divergence you want. See the section: HeNe Laser Beam Characteristics.

  • Back to HeNe Laser Testing, Adjustment, Repair Sub-Table of Contents.

    Beam Polarization of HeNe Lasers

    Typical Polarization Characteristics and Problems

    Sealed HeNe tubes with internal mirrors which are linearly polarized usually incorporate an internal Brewster plate to suppress the unwanted polarization orientation - and these HeNe tubes are more expensive. Where mirrors are external, the Brewster windows on the plasma tube accomplish this function and the output of these is always linearly polarized.

    Common inexpensive internal mirror HeNe tubes produce a beam that is either randomly polarized or slowly changing in polarization (as the tube heats) - possibly with a combinations of polarization states present simultaneously. Placing a polarizing filter in the beam of one of these lasers results in a variation in brightness, usually taking place over a few seconds possibly with sudden shifts as various modes compete for attention inside the resonator. The presence of any of these characteristics makes such a laser unsuitable for many experiments and applications. These tubes are normally designated as 'random polarized' (with an 'R' somewhere in the model number) which translates as: "The manufacturer has no idea of what the polarization characteristics will be at any given time". :)

    If the polarization were truly random, meaning all polarization states are present simultaneously (or on a short enough time scale that it doesn't matter), a simple polarizing filter in the beam path will produce a linearly polarized beam at the expense of at least one half the output power (that which is blocked because its polarization orientation is wrong and because of losses in the filter). However, where the polarization orientation of the laser is slowly changing, this approach will result in unacceptable varying output intensity from the polarizing filter. Additional optics including polarizing beamsplitters, mirrors, and combiners can in principle, at least, produce a stable polarized beam but these are complex and expensive.

    Determining if a Laser Tube is Linearly Polarized

    There are several ways to determine if a HeNe (or other internal mirror) laser tube or head has random or linear polarization (all external mirror laser with Brewster angle windows are linearly polarized): Note: Random polarized lasers are generally not truly random but will tend to (longitudinal) mode cycle and some modes will tend to favor a particular polarization orientation. This will result in the polarization changing gradually or suddenly (especially as the tube is heating up) so at any given time, they may emit a polarized or partially polarized beam but its orientation will not be predictable. Thus, to confirm that your tube is polarized requires that the polarization remains constant - not just for an instant (30 seconds to 1 minute should be enough time to wait to know for sure). If the beam is reflected off of a non-metallic reflective surface (which acts somewhat as a polarizer), you may see a large variation in brightness due to the polarization changing especially if it is a low gain or short tube where fewer modes are active simultaneously.

    Even a polarized tube may show a small amount of variability of the low intensity beam passed by a polarizing filter or reflected from a Brewster angle plate - this is normal and one reason why the specifications only say 500:1 or 1000:1 and not infinity:1. The reason is that the tube's linear polarization results from the cavity gain being maximized by the internal Brewster plate at the polarization angle. However, gain function with respect to angle is not a singularity - there is still enough gain for a few degrees on either side to maintain oscillation. And, some samples are better than others. Also see the section: Typical Polarization Characteristics and Problems.

    Unrandomizing the Polarization of a Randomly Polarized HeNe Tube

    The best option where a polarized beam is required is to start with a HeNe laser that produces a polarized beam! However, for the experimenter, there is at least one alternative - magnets to the rescue!

    I have found that placing powerful magnets alongside a random polarized tube will result in a highly linearly polarized beam. While this may be common knowledge at the Afternoon Teas attended by laser physicists (assuming they drink tea), it certainly isn't something found in popular books on lasers.

    A type of magnet that works quite well has a strength of several thousand gauss. The ones I used came from the voice coil positioner of a moderate size hard disk drive. They are rare earth magnets with dimensions of about 1.25" x 2.5" x .375" with the broad faces being the N and S poles. The amount of polarization is most pronounced by placing one of the broad faces of the magnet against the tube near its mid-point. Some adjustment may be needed to optimize the effect. I do not know how much magnetic field strength is needed but even moving this magnet 1/4" away from the tube surface greatly reduced the ratio of light intensity in the two orthogonal polarization axes.

    CAUTION: These types of magnets are very powerful. In addition to erasing your credit cards and other magnetic media, they will tend to crush, smash, or shatter anything (including flesh or your HeNe tube) between them and/or between them and a ferrous metal. Some portions of a HeNe tube or laser head may contain parts made from iron or steel. These rare earth magnets also tend to be quite brittle. In addition, the violent uncontrolled movement may place you and a HV terminal in the same space at the same time as well! Take care.

    With the magnet's N or S pole placed on the side of the tube, the result was a vertically polarized beam. By rotating a polarizing filter in the beam path, beam intensity could be varied from nearly totally blocked to nearly totally transmitted and the polarization orientation followed the magnet as it was rotated around the tube.

    The control wasn't perfect - a small amount of light with a slowly varying polarization did sneak through. However, it was significantly less than 1 percent of total beam power for these particular tube and magnet combinations (I have tried this with 2 different tubes with similar results). The constant portion of the residual beam may have just been a result of the imperfect nature of the polarizing filter.

    By using two similar magnets - one on either side of the tube with N and S poles facing each other (mounted on an aluminum U-channel for support and so they would not crush the tube), the variation in residual beam intensity was virtually eliminated. I do not know if this effect was due to the increased magnetic field or its more homogeneous and symmetric nature. This was also used successfully with an enclosed HeNe laser head:

                           |_____| Rare earth magnet
         |                                            |
         |             HeNe laser head                |=====> Polarized HeNe beam
                           |_____| Rare earth magnet

    Use of Magnets to Generate Polarized HeNe Laser Beam shows acceptable locations for one pair of magnets along side a typical 1 mW HeNe tube. This placement was found to be effective but possibly not totally optimal - experimentation may be required. Under some conditions, a single magnet slighlty separated from the tube seemed more effective, possibly because the field was spread over a longer stretch of bore.

    As far as I could tell, with this dual magnet configuration, the output beam characteristics were similar to those of a polarized HeNe tube. However, additional and/or more powerful magnets might be necessary with other tubes.

    Output power did not appear to be affected significantly. A measurement done later on a Melles Griot 05-LHR-911 HeNe laser head showed that when the polarization effect was most complete, the output power decreased by about 5 percent. A polarizing filter would nearly totally block the beam at one orientation and have minimal effect 90 degrees away from this.

    I do not know about the stability or reliability of this scheme but the only other effects seem to be to increase the required input starting/operating voltage and/or magnitude of the negative resistance of the tube slightly (current dropped by about 10 percent with the magnets using an unregulated power supply) and possibly to shift to point of maximum beam power to a higher tube current (5 mA instead of 4 mA for one tube - but this could have just been my imagination as well). With that 05-LHR-911, the operating voltage at 5 mA increased from 1,500 V to 1,550 v with one set of magnets and to 1,600 V with two sets. And, the laser would not stay on in a stable manner with the magnets very near the anode (cable) end of the laser head, but I didn't think to try and adjust the current setting to see if that would help.

    As a side note, output power *increased* by about 5 percent with a magnet some distance from the laser head, possibly due to the Zeeman splitting suppressing IR losses, but of course there was no effect on polarization.

    Where the capillary of the plasma tube is exposed as with many older lasers, and the magnets can be placed in close proximity to the bore, their strength can be much lower. Some commercial lasers (like the Spectra-Physics model 132) offered a polarization option (-01) which adds an assembly consisting of several ferrite magnets glued between steel plates that screws in place alongside the tube with the pole pieces (the steel plates extending beyond the magnets) above and below the tube bore. I performed some tests using a near-mint condition SP-132 (from around 1973) and the original magnet option, the extinction ratio and power stability are not as good with the magnets compared to the more common approach using a Brewster plate or Brewster window tube though. The Spectra-Physics specifications only claim an extrinction ratio of about 30:1 compared to 500:1 or better for a laser using a Brewster plate or window. I doubt that magnets are used for polarization in any modern HeNe lasers.

    Since it is possible to control the polarization orientation with permanent magnets, the next step would be do this with electromagnets. This would permit polarization to be dynamically controlled. Adding a fixed polarizer would provide intensity modulation without any connection to the power supply or expensive electro-optic devices. Hopefully, by using multiple sets of coils distributed along the side of the HeNe tube, a lower field strength would be adequate. Liquid helium cooled superconducting electromagnets would definitely add to the cost of the project. :-) Perhaps, someday, I will try this out.

    Magnets and Mode Plots

    Several years after doing this initial experiment, I recorded the mode behavior of a common barcode scanner HeNe tube with different magnetic field configurations. The laser itself is the same one used to construct the system described in the section: Sam's Very Simple Stabilized HeNe Laser, mainly because it already had photodetectors mounted for the two polarization orientations. It was also known that the particular tube was nicely behaved with respect to mode sweep, with no tendency to flipping. Two sets of magnets were used. The plots were obtained for each magnet set over about 8 minutes as follows:

    The following two sets of magnets were used for these tests:

    1. Three pairs of medium strength ferrite magnets positioned on each side of the tube resulting in a fairly uniform transverse field over about half the length of the tube roughly centered between the mirrors. These magnets are 2" high by 5/8" wide by 1/4" thick, placed 5/8" apart side-by-side. Plot of Spectra-Physics 088 Mode Behavior in a Moderate Transverse Magnetic Field shows the results.

      An aluminum frame with the 3 sets of magnets was placed over the tube in each of the 4 possible orthogonal orientations for about 1 minute. The strength and configuration of these magnets results in the beam being somewhat polarized, but with significant double frequency ripples in intensity of both modes. Surprisingly, the preferred polarization orientation is orthogonal to the orientation of the magnetic field! When the magnets are rotated 90 degrees, the effects essentially swap polarizations.

      Note how the total power is significantly higher than with no field for the first two magnet orientations. There is also a the difference in shape of the modes depending on orientation of the magnets, ranging from pulse to sinewave.

      Orientations of the moderate strength magnetic field other than horizontal or vertical produce intermediate effects, but with low stability and even some flipping behavior.

    2. A single pair of very strong rare earth magnets similar to those in the setup described above, but somewhat longer along the direction of the tube (about 2-1/2 inches). Each magnet was 1-3/8" high by 1-3/4" wide by 1/4" thick located toward the anode-end of the tube. Plot of Spectra-Physics 088 Mode Behavior in a Strong Transverse Magnetic Field shows the results.

      The super strength pair of magnets was placed over the tube, again for about 60 seconds in each orientation. In all cases, the polarization preference was very strong and lined up with the magnetic field. The suppression of the mode orthogonal to the direction of the magnetic field was nearly perfect. But, the total output has declined for all but the second orientation (Ver-N) and quite dramatically for the last one (Ver-S). That reduction is at least 20 percent and quite obvious simply observing the brightness of the spot on the photodiode.

      It may be possible to find a preferred orientation of the high strength magnets where the total output power is maximized with good stability. Ver-N seemed particularly good with total power at least equal to that without magnets.

      A closeup of the first case is shown in Plot of Spectra-Physics 088 Mode Behavior in a Strong Transverse Magnetic Field (Hor-N). However, note from the plot that although the S-Mode (blue) is quite close to 0, the P-Mode (red) and total power are rather lumpy. Of course, this is probably not the ideal magnetic field configuration to force linear polarization being only a single pair of magnets. And, the polarization ratio is probably only about 50:1, not the 500:1 or 1000:1 of a normal linearly polarized HeNe laser.

      Orientations of the strong magnets at other angles aligns the polarization with them, but, but with various amount of a reduction in total output power.

    The individual polarized modes and total power were captured for the plots. The two orthogonal polarization orientations (shown in red and blue) were detected using the waste beam and photodiodes that were already present in this laser. A trans-impedance op-amp buffer converted their uA-level outputs to the +/-10 V range of the data acquisition system. The total power (shown in dark green) used the main beam with a photodiode and resistor load. The scale factors of the three signals are fairly close but not perfectly matched. However, even accepting errors in the scale factors, there are additional unexplained discrepancies between the sum of the modes and total power, which should be equal. This is especially evident for the high strength magnets comparing the dominant mode to total power (since there is almost no power in the other mode). I'm not positive of the cause but suspect some interference effects in the detector optics for the waste beam. While not actually destabilizing the laser, multiple reflections could explain the variation in mode amplitude, though why it seems worse for one of the modes is a mystery. But the ripple remains with the optics channel when swapping the photodiode electrical connectors. And, rotating the pickup assembly so it is at a slight angle definitely makes a difference. I have seen some high frequency noise of the laser output power possibly due to plasma oscillation in the tube with the high magnetic field. This may be resulting in some differences in response through the electronics. It looks like the total power is fairly well behaved, but the waste beam power has the irregularities. So, perhaps the preamp needs some attention. But I did try increasing the time constant of the op-amp buffer by a factor of 10 with no noticeable change in the response. Hmmm. With the magnets centered between the mirrors, the discrepancy between total power and the sum of the mode power was even worse. When pushed toward the anode-end of the tube, it seems to quiet down, though that might have just been a coincidence. I'm going with the optics interference explanation for now. :)

    Magnet Tests with a Polarized HeNe Laser Tube

    In an effort to determine how strong the effect of a magnetic field actually is on the polarization, I later tried the same scheme as described above but with a small polarized tube - the type with an internal Brewster plate. Not surprisingly, the effect of the Brewster plate is much more pronounced than that of even very powerful magnets. Here is the output power after warmup with different magnetic fields:

    1. No magnetic field: 0.75 mW.
    2. Strong transverse magnetic field (aligned with polarization): 0.70 mW.
    3. Strong transverse magnetic field (orthogonal to polarization): 0.50 mW.
    4. Weak magnetic field (almost any orientation): 0.80 mW.

    In all cases, the polarization was unchanged and output power was at least as stable as without any magnetic field. Thus, even the strong magnetic field was insufficient to overcome the losses of the Brewster plate at the (wrong) orthogonal polarization orientation but did reduce the gain at the (correct) aligned polarization orientation enough to cut output power by 33%. (For this short tube, lasing would probably have been killed entirely if forced to have its polarization orthogonal to the correct orientation.) These results are not unexpected - except perhaps for (4) - I do not know if the increase in power was simply a result of the usual Zeeman splitting effect suppressing the IR wavelengths or something else. A noticeable increase in output power due to Zeeman splitting is usually associated with long high power HeNe tubes, not the 0.5 mW tube used for these tests.

  • Back to HeNe Laser Testing, Adjustment, Repair Sub-Table of Contents.

    Repairing Leaky or Broken HeNe Tubes

    Gas Fill Problems with HeNe Tubes

    HeNe tubes can fail due to slow leakage through soft (Epoxy) seals (not found on modern tubes), or actual damage or defective manufacturing leading to a crack or break. Helium loss through diffusion is a special case and may be remedied easily without major effort or investment. See the section: Rejuvenating HeNe Tubes. Other gas fill problems will require a very non-trivial amount of work and access to sophisticated. Minor structural damage may be repaired but then a refill will be needed. See the section: Repairing and Refilling a HeNe Tube at Home? However, if your tube is broken in half, you should probably just salvage the mirrors (for possible future use) and move on to other things!

    The major HeNe laser manufacturers and laser repair companies may offer regassing services for larger more expensive HeNe tubes (high power internal mirror tubes or those with Brewster windows designed to operate within an external resonator). Figure on $500 or more to regas an HeNe tube, and more still if there is physical damage (assuming they will bother with it at all).

    Whether the cost of such an operation can be justified is another matter. For a high quality research laser it probably makes sense as the tube alone may cost several thousand dollars or more - if a replacement can be obtained at all. Even a basic HeNe tube with Brewster windows may cost over $600 (being much less common and thus much more expensive than the internal mirror variety). However, for small sealed internal mirror HeNe tubes, low cost replacements are readily available at perhaps 1/10th to 1/4th the cost of a regassing service (even cheaper if you are willing to use a surplus tube).

    However, where the tube has high mileage and died from use and age, it may not simply be a matter of regassing. The following is from the Melles Griot FAQ Page:

    "While regassing can provide some extension of the output performance in some gas lasers like the CO2, argon and the higher powered side arm HeNes (which have external optics), it is not recommended or provided for smaller internal mirror coaxial tubes. Typical end-of-life failure for a HeNe tube is cathode sputtering. This occurs when the protective oxide layer on the cathode is expended through continuous bombardment by the laser discharge. There is no cost effective way of regenerating this layer. When the oxide layer is expended, the discharge itself vaporizes the "raw" aluminum and deposits this material, in its vapor state, on other surfaces such as the optics and the bore."

    So, while refilling may help some, the sputtered aluminum coating will remain on critical surfaces. A careful visual inspection of the bore and mirrors may reveal whether a suspect tube is worth saving - a black or metallic film could indicate that serious sputtering has taken place. However, I've also seen tubes where discoloration in the bore, at least, had no noticeable effect on performance.

    Rejuvenating HeNe Tubes

    These techniques may be used on HeNe tubes that are marginal due to loss of helium though diffusion or a slight amount of contamination from air leakage due to age. Where the tube lases weakly, a helium soak or reactivation of the getter may help. If there is no lasing at all, many other causes are possible.

    The best way to determine if loss of helium or slight contamination is your problem is to check the spectrum of the discharge. See the section: Instant Spectroscope for Viewing Lines in HeNe Discharge.

    Helium Soaking

    Where just the helium (remember how slippery those He atoms can be!) has leaked out, there may be an alternative to the dumpster or a major refurb. HeNe tubes which do not lase well or at all due to loss of helium can sometimes be rejuvenated by soaking them in helium at normal atmospheric pressure for a few days or weeks. You don't need a pressure chamber or any other fancy equipment - a few helium filled party balloons and a garbage bag will do just fine.

    However, there could be other causes like misaligned mirrors or excessive tube current (due to a defective power supply). Check for these possibilities first and confirm loss of helium with a spectrometer capable of actually measuring the relative intensity of the spectral lines if possible. From my experience, just viewing the discharge with a diffraction grating will not reveal a low helium condition unless it is extremely severe - as in almost none remaining. (I've yet to actually see this. If anyone has a HeNe tube with certifiably low helium, please send me mail via the Sci.Electronics.Repair FAQ Email Links Page. I'd be interested in testing it.)

    The point to realize is that it is the partial pressure of each gas inside and out that matters. Neon is a relatively large atom and does not diffuse through the tube at any rate that matters. However, helium is able to excape even when the pressure difference is small. For a typical HeNe tube at only 2 Torr (1/380th of normal atmospheric pressure), the partial pressure of helium in the tube is still much much greater than its partial pressure in the normal atmosphere. So, helium leaks out even though the total pressure outside is several hundred times greater. Conversely, soaking a HeNe tube in helium at 1 atmosphere will allow helium to diffuse into the tube at several hundred times the rate at which it had been leaking out. Thus, only a few days of this treatment may be needed if the problem is low helium pressure. Assuming that the desired partial pressure of He is 2 Torr, the ratio of age:soak-time will be about 380:1 or pretty close to 1 day of soak per year of the tube's age.

    Helium loss is most likely with soft-seal tubes - those with an Epoxy-type adhesive holding the mirrors or Brewster windows in place. However, it is also possible for hard-seal tubes using frit seals or optical contacting to lose helium though probably at a slower rate and rejuvenation will also take proportionally more time. Checking the intensity of the He lines with a spectroscope is really the only way to know for sure if He loss is the problem and to also monitor the soaking process.

    Almost any sort of helium supply will work for atmospheric pressure diffusion including welding supply grade and even the stuff sold for filling party balloons. (Note, however, that these sources are mostly the common isotope of helium, He4, not the light isotope, He3 which may be what was originally in your tube - see the additional comments below.) A party tank of helium may be as little as $15 or $20 or just buy a few prefilled balloons and empty their contents into an air-tight plastic bag containing the HeNe tube. However, make sure what you are getting is really helium and NOT hydrogen!! In addition to the flammability issues, any significant H2 that makes its way into your HeNe tube will make the situation worse - probably terminal. Also note that as much as 50 percent of what is in those party tanks may actually be air, nitrogen, and/or some other unidentified gas, so the process may take somewhat longer (approximately by 100 divided by the percent of actual helium) though most of these contaminants won't hurt the tube.

    The required amount of effort hardly seems worthwhile for a $15 1 mW HeNe tube but it is something to keep in mind for other more substantial and expensive types.

    Note that there are a few types of tubes that won't benefit from helium soaking even if they have certifiably leaky seals. Those are tubes where the seal is between the interior and another sealed chamber as with some older Aerotech HeNe lasers. In these, the leaky seal is on a Brewster window but the laser mirrors are attached with frit or Epoxy to an external sealed chamber which is filled with air. The only thing helium soaking will do is slightly increase the partial pressure of He in that external chamber which will essentially no effect on the internal gas fill. It might be possible to drill a hole in the metal end-plate or melt a hole in the glass of the external chamber though, hopefully without contaminating the Brewster window. Or, use a diamond saw to cut one end off entirely and install the mirror on an adjustable mount.

    (From: Mark W. Lund (

    I have rejuvenated HeNe laser tubes with low helium pressure. Since the partial pressure of 1 atmosphere helium is much higher than inside the tube you don't really need to use high pressure, or even increased temperature. I just put them in a garbage bag and blasted some helium into it from time to time. The length of time necessary in my case was a few days, but depending on the glass type, thickness, and sealing method this may vary. It would be good to test the power every couple of days so you don't overshoot too much.

    One warning: Helium has a lower dielectric strength than air, so don't try to operate the laser in helium as it may arc over.

    (From: Philip Ciddor (

    My information is very old, but may be helpful. Early 2 mW red tubes had about 2 torr of He, so soaking in 760 torr (1 atmosphere) of He for 1 day per year of life roughly restored the initial He pressure, since diffusion rate is proportional to pressure difference. I have no data on the gas mix in current green or IR tubes, but if you can find it, similar scaling may be feasible.

    (From: Sam.)

    Gas fill probably isn't all that different for non-red HeNe tubes so the same general recommendations should apply. However, since their gain is lower, nearly everything about near-IR (1,523.1 nm and 1,152.3 nm), orange (611.9 nm), yellow (594.1 nm), and particularly green (543.5 nm) HeNe tubes is more critical including power supply current and mirror alignment. So, it is important to eliminate other possible explanations for low or no output or other problems before blaming loss of helium.

    I cannot overemphasize the importance of carefully monitoring the amount of helium that has diffused back into the HeNe tube (by removing it from the bag of He and testing with a spectroscope periodically and for a laser beam) - once helium pressure goes too high, the only (non-invasive) way of lowering it is to wait a few years or decades. :-) If power is just low and you are trying this, put the tube in the helium soak for a couple of days and then check power output again. If it has increased, repeat this procedure a couple days at a time until power levels off or starts to decrease. If power decreases after the first soak, helium loss isn't your problem!

    If it's possible to wrap the tube such that only the seals are inside the helium and not the electrode connections (the glass envelope shouldn't leak at any rate that matters), monitoring of power can be done without having to remove the tube from the helium container or whatever.

    CAUTION: Apparently, most modern HeNe tubes are actually filled with the light isotope of helium, He3, rather than He4 which for all intents and purposes, is the one found in nature (99.9998%). He3 has a higher energy state which may be better for exciting certain transitions. Thus, helium soaking with common He4 could result in problems including reduced maximum power, greater frequency spread, reduced stability, or something else. As noted above, once the HeNe tube has been helium soaked, the effects are irreversible without waiting many years. The only practical way to determine what isotope(s) of helium your tube used is probably to ask the manufacturer - even a high resolution spectrometer won't help if the helium has escaped. For a common red HeNe tube, there is little to lose by using common He4 though results may not be optimal. However, if the tube is from a specialized research laser, it would probably be best to have a professional laser refurb company or the original manufacturer deal with it. You could make matters worse.

    WARNING: In addition to not attempting to operate the HeNe tube itself in a helium atmosphere due the lower breakdown voltage, there may even be problems with He diffusing into power supply components or ballast resistors and lingering there. So, if possible, remove the HeNe tube from its laser head or system enclosure for the helium soak. Or else, wait awhile (your guess is as good as mine) after dumping the helium before applying power.

    Reactivating the Getter

    Where some air has entered due to age (not an actual leak, that you can forget about unless you want to try refilling the tube at home - see below), it may be possible to reactivate the getter and absorb/combine/react with the unwanted molecules. However, don't expect miracles.

    Note that not all HeNe tubes have getters. For some that do, the getter may never have been activated in the first place (if the gas fill was already deemed pure enough after pinch-off). See the section: Gas Fill and Getter for info on the getter in a HeNe laser tube. And, if the getter was activated, the source of the active material (in the getter electrode) may have been totally depleted during manufacture so there may be no more remaining.

    This only has a chance of working if the gas pressure is nearly correct - not if it has changed by a factor of 100. The closest example I have of the effect of the getter on tube vacuum is for a typical TV or monitor CRT:

    (From an engineer at Philips)

    "A regular CRT-type getter can reduce gas pressure from about 10-6 Torr to its final value of 10-9 Torr IFF the gases can be gettered at all. H2, O2, N2, CO, and CO2 can be gettered. CH4 (Methane) can not be gettered but by heating, it can fall apart into C (a solid) and H2 that can be gettered. Noble gases can not be gettered either, so their gas pressure will determine the final gas pressure in a picture tube."

    Of course, for a HeNe or Ar/Kr ion laser, those inert gas molecules ARE the desired result! :) Unfortunately, since the typical gas laser operates at a pressure 1,000,000 times higher than a CRT (a few Torr), any effect of the getter on detectable contamination is likely to be minimal. How to tell? If the color of the discharge is more towards white or pink than it should be and there is still at least some evidence of lasing, the getter has a good chance of returning it to normal assuming all its active material isn't already used up. If the color is too orange, then the helium loss may be indicated and a helium soak may be all the tube needs. See the section: Helium Soaking.

    However, there is probably nothing to lose if the tube is unusable and you won't be going the entire route of refilling it. Heating the getter can be achieved in a variety of ways including (depending on design and what you have available): DC current, glow discharge, Sunlight and Fresnel lens, RF, and induction heating, even a microwave oven. See the sections starting with: Methods to Activate Neon Sign Electrodes and Getters. The Solar heater approach is low tech and known to work where there is no visible 'white cloud of death' (heating the white stuff (which is probably unavoidable with the Sun's rays) seems to release previously trapped stuff making the situation much worse). See the section: Simple Solar Heater.

    I've also tried using a 1 watt fiber-coupled laser diodes with a focusing lens to heat the getter but although an incandescent spot could be seen on the getter, there was no significant change in performance. Perhaps I didn't let it cook long enough. A 10 or 20 watt diode or YAG laser might work better. :) But a CO2 laser will not work since 10.6 um can't get through the glass.

    The idea is to drive off some of the material remaining in the getter electrode onto the walls of the tube. If nothing appears or it turns milky immediately, the getter probably isn't capable of helping much - though even in this case, try out the tube again - it may have helped just enough. Lack of results could also mean that the getter electrode hasn't been made hot enough or the material it contained had already been fully used up.

    Note: If you expect to try your hand at actually refilling a leaky tube, DON'T attempt to reactivate the getter - you may need it later!

    The same approach can be used with ion laser tubes if they are made of glass and you can locate the getter. Those that are of all ceramic construction may still have a getter, but it may need to be heated by a precisely controlled current flow between the cathode end-bell and filament or something equally obscure like that - not easily guessed! Also, since these tubes are generally much more expensive than HeNe tubes, it may pay to have it professionally refurbed.

    Once the tube has been revived (or perhaps even before you make the attempt), adding an additional layer of Epoxy/TorrSeal at the tip of the exhaust tube, mirror(s), and any other possible areas of leakage would be a good idea. This is particularly relevant for modern hard-seal tubes since they shouldn't really leak at all (at least on time-scales that humans can understand). Thus, any contamination generally means an actual defect at the frit seals or exhaust tube (tip-off). Soft-seal tubes leak by design :) but adding an additional layer of sealant at the mirrors, end-caps, tip-off, and other suspect locations can reduce this somewhat. At least it won't hurt - unless you accidentally glop it on the OC mirror! :(

    I've successfully revived a couple of Melles Griot HeNe laser tubes which had getter electrodes but no visible getter spots (which means the material is transparent). One was a hard-seal tube that must have been contaminated in some way since after treatment, it has worked essentially unchanged for over a year. The other was a green HeNe laser tube that had an Epoxy seal at one end. However, all attempts to revive Spectra-Physics HeNe lasers have failed miserably and generally made matters worse. Heating the "white cloud of death" material (including what's no doubt inside the getter ring) must release whatever it previously trapped.

    Repairing and Refilling a HeNe Tube at Home?

    If you are really really ambitious, have lots of time on your hands, have access to lab supplies, laser grade He:Ne gas mixture, and a high vacuum system in good condition, you, too, can refill an HeNe tube - at least as an experiment. Whether it could be sealed off and then expect to have a long life is another matter.

    First, any physical damage would have to be repaired. For example, if an overzealous attempt at mirror alignment resulted in a mirror breaking off at the frit seal, it would have to be reattached - in as precisely the same position as possible using new glass frit or Epoxy (though that will leak over time). If someone yanked on the anode wire on a large HeNe tube broke the metal-to-glass seal, that would have to be repaired - again with Epoxy or by actually heating the glass to fuse it together. However, the latter risks shattering the entire tube if you aren't experienced in glass working. If you don't know where the leak is, then you need to find it first. :)

    Once the HeNe tube is known to be gas-tight, the seal is cracked at the exhaust tube, it is put on a high vacuum system to pump it down and backfilled with pure He:Ne gas mix several times while baking out impurities.are very finicky about gas purity.

    For more information on this sort of endeavor, see the chapter: Amateur Laser Construction, the section: Home-Built Helium-Neon (HeNe) Laser, and the introductory chapter: Home-Built Laser Types, Information, and Links for relevant information. Good luck! :-)

    Regrinding or Otherwise Compensating for a Chipped Mirror

    So that wonderful HeNe tube you had fallen in love with (we all have our peculiarities!) got knocked loose from its mounting and fell face down on the granite countertop. Exactly what were you doing with your laser? :) Now, there is a chip in the OC mirror which extends into the area of the beam and the beam is but a shadow (well not quite but you get the picture!) of its former self. OK, it ouptus a unique pattern but not quite what you had in mind! (Note that for the most part, a similar accident with the HR mirror at the other end of the tube wouldn't affect anything but its external appearance.) Is there any hope?

    Well, assuming the chip isn't too deep, it is possible to grind it out and then polish the resulting surface to optical quality. To do this properly will require a means of holding the tube just slightly off of perpendicular (to add some wedge - see below) to a rotating platform on which various grades of wet grinding compound can be introduced starting with something coarse like 400 grit and going up in stages to 1,200 grit or more, following by lapping with optical rouge for the final polish. That should get you a reasonably decent result after considerable effort and cost. But don't expect it to be to 1/10th lambda!

    One thing you won't be able to reproduce is the anti-reflective (AR) coating present on most HeNe OCs. (Well, not unless you have access to some vacuum coating equipment!) That is the reason I suggest grinding it on a slight angle - the resulting wedge will divert the reflected beam away from the axis of the cavity and minimize instability and interference.

    I was given a cute little HeNe tube with such a chip in the OC mirror. Now, this certainly wasn't worth spending much of anything to repair (it was only a .8 mW, barcode type HeNe laser after all!). So, I decided to experiment using the minimalist approach: emery paper. I started with 400 grit to remove the chip and then 1,200 grit. I also deliberately attempted to grind the surface parallel to the actual mirror rather than with wedge to see what would happen. All this just by hand so the result is also somewhat convex rather than perfectly flat. I need to find some rouge to attempt the final polishing if I ever bother.

    Even without fine polishing, the beam was much much cleaner than it used to be (formerly being spread out off to one side in random directions!). Just for grins and giggles, I went back to 600 grit to see what effect an even more random ground surface would have on the beam. The interference patterns are really quite interesting - sort of like a stellar globular cluster - so I may just leave it the way it this way. :)

    Another alternative where the area of the beam just touches the chip might be to push the mirror mount side-ways beyond the restricted area. With care, it may be possible to shift it by as much as .5 mm which could be enough.

    Or, use some optical cement to glue a flat piece of glass to the mirror filling the voids. With the proper material that closely matches the index of refraction of the mirror glass, such an approach may result in a beam that isn't too terrible. :)

    Using a Microwave Oven to Evaluate and Revive HeNe Laser Tubes

    WARNING: These are dangerous procedures, at least for your laser tube! Attempt at your own risk. There is a good chance that the tube will be ruined totally as a result of the glass or glass-to-metal seals cracking. There is also a chance that the procedures will make the situation worse or that any apparent improvement will be temporary.

    Note that using a microwave oven is safe for just checking to see if the tube is gas-intact and has approximately the correct discharge color. In this case, only a second or two is needed so heating is minimal. See the section: How Can I Tell if My Tube is Good?.

    The whacko procedures below may be used to provide an idea of what is wrong with a HeNe tube as well as to at least partially revive them in some cases. The difference between evaluation and revival is basically in cooking time and how many times the procedure is repeated.

    Alternative sources of RF energy can be used in place of the kitchen microwave but may not be quite as convenient or as readily available. :)

    I have had some modest success in at least partially reviving some old soft-seal HeNe laser tubes with the power output from 4 of 6 weak tubes being improved significantly, though not to anywhere near the rated specifications. However, one tube was destroyed due to the glass cracking (the first one I tried, not having a feel for the safe cook time), and on another, the power went down slightly. To what extent these results are due to getter reactivation or other phenomena is not currently known. The effects of the microwaves (whether it be from the discharge or just due to heating) would also appear to be useful as a diagnostic tool for evaluating HeNe tube condition.

    Since the entire tube or whatever has to be inside the oven (don't even think about drilling holes in the side or door!), this stunt probably only applies to smaller helium-neon laser tubes and maybe the getters in receiving tubes if you remember what they are. :) Here goes:

    However, what you may find is that the power increases immediately after treatment but then decays back to its original value or below over the span of a day or so, or faster if the tube is powered. But, this may be repeatable, so if you just need a temporary boost, go for it! :)

    I would appear that the microwave treatment may do any or all of the following:

    I expect that none of these phenomena will be result in a substantial change in behavior for a healthy tube. Thus, microwave (or other RF) excitation/heating may represent a viable diagnostic tool for evaluating HeNe tube condition.

    See the section: Attempting to Revive Some Soft-Seal HeNe Tubes for some not terribly conclusive results from using this technique, additional discussion of some of the peculiar effects, and some tests with a more modest RF exciter.

  • Back to HeNe Laser Testing, Adjustment, Repair Sub-Table of Contents.

    Reports from Sam's HeNe Laser Hospital

    A HeNe laser might seem like a simple bit of technology, a glass bottle with mirrors at each end filled with electric sign gases. But that cannot be much further from the truth. They can suffer from all sorts of maladies due to improper design, manufacturing, calendar age, hours of use, abuse, and neglect. These include (in no particular order): physical damage, contamination and other gas problems, slight or total loss of mirror alignment, improper coatings on the mirrors, wrong mirror curvatures, wrong size bore, lack of wedge in the mirror substrates, failure of cathode can pickling, bad electrical connections inside the tube, running with wrong current or reverse polarity, and many more. And too few of these are covered by popular laser health insurance plans and warranties. The warehouses and dumpsters of laser manufacturers have been filled with all sorts of specimens, and many eventually find their way to laser surplus outfits and eBay! :)

    The following are some of the cases I've come across over the years. And some of them are real doozies like Oops! HeNe Laser Tube Meltdown, may it rest in pieces. :)

    Sam Succeeds in Aligning a LONG HeNe Tube

    Just to show that the alignment techniques in the sections starting with Problems with Mirror Alignment aren't some textbook exercises dreamed up by theorists, long HeNe tubes can be aligned from scratch using a minimalist approach. I did it once. (Well, actually twice but the other tube was pretty short so it doesn't really count.) How many data points do you need to prove something? :)

    I had a 30" HeNe tube sitting in my attic for about 2 years. It would start but not lase. (To power it, I am using an SP-255 exciter set at its minimum current of 7 mA with an 80K ballast resistance.) The lack of lasing is almost certainly due at least in part to mirror alignment problems. In fact, originally, one of the mirrors was obviously bent at a visible angle! I had tried to straighten them both the best I could when I acquired the tube but was unsuccessful at that time. I had used the basic reflection technique for mirror alignment but wasn't able to configure the setup stably enough to work on such a long tube.

    A few days ago, I decided what the heck, no darn HeNe tube is going to get the better of me! First, I tried using the beam from an argon ion laser (it's blue so would pass down the bore and hopefully could be centered). No dice. The beam diverged too quickly for the long bore and it was impossible to figure out exactly what 'centered' meant - there was no single easily identified best position and orientation. (I assume that when laser companies do this, they have additional optics to produce beam of optimal size and minimal divergence as well as a spatial filter to clean it up. I wasn't quite willing to go to that amount of effort!)

    I then contemplated building a light bulb and telescope rig as described in conjunction with the home-built lasers in Scientific American but concluded that such an approach wouldn't have any chance of working with a long narrow bore tube. I also attempted the method whereby the reflection of the discharge from the far mirror results in a slightly brighter spot exiting the near mirror but not knowing how far off the mirror alignment actually was, this proved impossible and even Sam's Super Cheap and Dirty Laser Power Meter) with its sensitivity boosted by using a 5 uA panel meter for the readout (about 2 uW full scale) could detect absolutely no change when tweaking the mirrors. Bummer. :(

    So, I decided to use the "Bore Sight" method described in the section: Major Problems with Mirror Alignment. Please refer to Bore Site Method of Internal Mirror Laser Tube Alignment for what should be fairly self explanatory diagrams of this technique if you don't want to read the feature length version. :) The Bore Sight Cards (BSCs) were screwed to the ends of my wooden "Big HeNe Tube Cradle" (a pair of V-blocks attached to a 1x4) and their 1/16" holes carefully lined up with the bore of the 30" Tube Under Test (TUT). With the TUT removed, the Alignment Laser (A-Laser, a 1.5 mW HeNe head) was placed on the platform described in the section: Simple Adjustable Optics Platform with its aperture about 2-1/2 feet from the nearer BSC and aimed squarely down the center of the two bore sights.

    The TUT was then placed back in the cradle in exactly the same orientation as before, first with the OC facing the A-Laser. A lever adjuster (read: big flat blade WELL INSULATED screwdriver) was used to tweak the mount at the OC end to center the doubly reflected spot precisely into the bore sight aperture. Note: Two reflections - First from the TUT mirror and second off of the OC of A-Laser - this actually increases the sensitivity to alignment error). Then, I turned the TUT end-for-end to do the same with its HR mirror.

    A weak beam appeared after the first attempt! I practically fainted. :) Then, I worked at boosting the power by additional mirror adjustment.

    If the tube dropped on the floor or blew up, I'd be disappointed, but I accomplished what I really believed would be impossible without a much more sophisticated alignment technique! This was TOO easy! :-)

    OK, it isn't perfect - At first I was only getting a maximum of 3 to 4 mW from this 30 inch tube (which should probably be producing 15 to 20 mW) and the power is constanting changing - going as low as 1 mW over a 10 minute or so period. The beam is pretty clean, just weak and variable. Even very slight finger pressure on the mirror mounts intensity or disappear entirely. Gentle pressure on the center of the tube, or the tube's orientation ("This Side Up") also affects it noticeably. And, "walking the mirrors" by applying equal pressure in opposite directions at both ends doesn't seem to help much if at all and these effects are inconsistent. In fact, at various times, the same amount and direction of mirror mount deflection may increase or decrease the output! The behavior has some similarity to normal mode cycling but where a HeNe tube is operating with insufficient gain and/or a limited number of available longitudinal modes. Thus, I conclude that at this point, the alignment is close enough that any further mirror tweaking, if needed, will be done with the tube mounted three-screw adjusters described in the section: Means of Adjusting HeNe Tube Mirrors.

    I acquired this HeNe tube along with a couple of other long tubes of pretty much unknown pedigree. They all appear to behave in a somewhat similar manner (but the alignment of the others was fine). Possible causes include any or all of the following (I welcome any additional suggestions):

    1. Need for magnets to suppress strong IR lines as mentioned above. Long HeNe tubes often need a series of magnets to reduce the gain of the very strong IR lines, particularly 3.391 nm, through the process of Zeeman splitting (see the section: Magnets in High Power or Precision HeNe Laser Heads). However, I don't know precisely when a tube is long enough for this to be a problem or under what conditions the use of magnets can be avoided. A shorter but otherwise similar 19", 12.5 mW tube operates just fine without any magnets. I also don't know how much of a boost magnets can provide when all is said and one (or I exhaust the World's supply of magnets!).

    2. Loss of helium as a result of diffusion through the glass isn't out of the question (though a rough check of the spectrum doesn't show anything amiss). Given the age of these tubes (probably pre-1990), this is a possibility but I kind of doubt it to be a problem since that 19", 12.5 mW tube dates from the same era and if anything the effects of helium diffusion should be lower for a longer tube (about the same surface area to volume ratio, but a lower seal area to volume ratio).

    3. Poor or failed design. It's quite possible that these tubes were never quite right and were sitting in the back of the 'dead inventory' storeroom until discarded. One was definitely much older than the others, so they all weren't from the same batch, but perhaps the same recipe. The erratic behavior (especially) could be explained by mirrors that were too long a radius or flat (it isn't easy to tell for such a long tube). Incorrect OC reflectivity would result in a reduced maximum output.

    4. Or, just the fact that I neglected to issue the special HeNe laser long tube chants and incantations (which I have unfortunately lost). Now, this is something I definitely need to consider, though I don't know what options are still available at this point! :)

    My initial guess was that assuming this (and the other long tubes) aren't simply defective, is that they need a wad of IR suppression magnets in strategic locations to boost the output power and mirror micro-adjusters to stabilize the output power.

    This tube looks exactly like any normal coaxial style HeNe tube, just a bit longer than most. I have a dead SP-124 laser (which is of similar length but with a side-arm tube and external mirrors) so I know what it does for magnets (See the section: Description of the SP-124 Laser Head) but with that design, the magnets can be placed next to the bore. With a coaxial tube, there is at least a 3/4" minimum separation meaning that the magnets would have to be much more powerful to result in an equivalent strength magnetic field inside the bore. And as far as I know, big cylindrical laser heads aren't any different than small cylindrical laser heads - no magnets. But perhaps this is incorrect. However, Melles Griot lists several 25 to 35 mW cylindrical laser heads in their catalog that are only 2 inches in diameter - leaving little room for powerful magnets!

    I did do some experimenting a bit later and found that a pair of really powerful rare-earth disk drive positioner magnets seemed to help a bit with maximum power now about 7 mW, but did little to reduce the fluctuations in power over time - up and down. However, a series of weaker ceramic magnets along the side of the tube didn't do anything good or bad. I then tried a series of 8 toroidal ceramic magnetron magnets with alternating N and S poles sitting under the tube and this boosted maximum power to a bit over 8 mW with just the right finger pressure on one of the mirror mounts. I expect that another bunch of these magnets above the tube would add another mW or so but kind of doubt this as a cure. I can't imagine that the laser heads these things were designed for required a couple dozen or more super strong magnets to function properly. Or, maybe there are very special locations for each magnet (part of the secret formula) allowing for fewer and/or weaker magnets to suffice. The use of the magnets did boost maximum power by 60 to 100 percent but getting another 200 percent boost in this manner seems unlikely!

    I do believe that the addition of the three-screw mirror adjusters will be enough to reduce the variations in power not due to mode cycling. With the tube in supported inside the aluminum cylinder from a dead 24" HeNe laser head (another of those 19" tubes, but this one was up to air), power starts off low (below 1 mW) when cold but peaks above 7 mW and remains above 6 mW without touching anything. Since slight finger pressure on either mirror mount will achieve 7 to 8 mW at any time, this suggests that it is indeed a matter of the pointing accuracy of the mirrors changing due to thermal effects.

    And with respect to magnets, I've now acquired an intact laser head with similar a Aerotech HeNe laser tube in it. Indeed, the thing is loaded with magnets surrounding the tube on 3 sides over most of its length. So, it's quite possible they are essential to achieve any sort of stability and to reach a reasonable output power. However, after installing this tube in that head, the performance isn't all that much better than with my cobbled together collection of magnets. My conclusion now is that the tube was built with a defective recipe or the recipe wasn't followed, possibly with respect to bore size versus mirror curvature. The TEM00 mode may be too large greatly increasing diffraction losses.

    Repairing the Northern Lights Tube

    Thanks to my Solar powered getter heater, the power came up to 4.6 mW (from 2 mW). I used a $1, 7" x 10" plastic Fresnel lens reading magnifier focusing Sunlight on both the front (the actually chemical) and the back of the steel or whatever U-channel getter loop. After a couple of these treatments, the discharge was almost uniform and the correct color. Only knowing that there was a problem would anyone notice the slight change along the bore. The power at this point peaked at 3.25 mW. Unfortunately, the Sun moved away from my HeNe tube reprocessing area (i.e., back yard) so further progress had to wait until it returned.

    Knowing that this lens would be of high enough quality and of adequate size, I built an adjustable mount for it so that the getter can be positioned reliably at its focus. Not that I had too many doubts - it was quite effective at instantly vaporizing leaves and the occasional unfortunate bug. :) See the section: Simple Solar Heater for details. The next day, with my fabulous contraption in-hand, I gave the tube a few more treatments of several minutes each, focused on the inside (active area) of the getter. After the third or forth of these, the maximum power leveled off at 4.6 mW which leads me to believe that the contamination has been eliminated. The discharge color is now perfectly normal and uniform over the length of the bore. It turns out that the operating voltage has increased by about 100 to 200 V (estimated) between the contaminated and present state. In addition, the output now peaks at just about the correct 6.5 mA rather than 8 mA as it did before.

    A summary of discharge color versus power output for this tube is given below. I assume behavior will be similar for other tubes though the power outputs will differ in both absolute and relative terms.

    Thus, even a very slight anomoly in discharge color can indicate that output power is likely to be much less than might be possible with a little 'cleanup'.

    Interestingly, there is still absolutely no evidence of a getter spot so I assume my procedure doesn't actually result in a significant amount of material being ejected from the getter. Other possibilities are that the active chemical is perfectly clear in both its original and 'used up' state or that it is designed to be retained within the getter structure.

    Some final mirror adjustments at both ends (together using the 'walking the mirrors' technique - see the section: Walking the Mirrors in Internal Mirror Laser Tubes) and the tube is now producing a very respectable 5.25 mW. I pronounce it cured. :)

    Followup: I retested this HeNe tube after a rest of several months. It appears to be unchanged or perhaps even improved a bit - output quickly climbed to 5.25 mW and was still increasing when I powered down. So I wonder its problems were not due to an air leak or residual air but to some sort of internal contamination. Another indication of this is that the discharge color variation was opposite of what I have seen with soft-seal HeNe tubes. It was correct at the anode but tended toward pink/blue at the cathode.

    Additional followup: After more than a year, I can detect no loss in power. Thus, an air leak is unlikely as the original cause of the malady. I can only conclude that it was from a manufacturing goof.

    Strengthening a Weak Siemens HeNe Tube

    Further treatment required removing this tube from its cylindrical laser head. This wasn't that difficult as the the end-caps came off reasonably easily due to the brittle glue and after drilling out a pair of pop-rivets. The 12 RTV Silicone blobs were readily accessible and succumbed to my roofing flashing aluminum blade. Checking the alignment at the anode-end showed that it was also optimal in relation to the current cathode-end alignment. I thought that the discharge color might have been a bit on the pink side so I performed several Solar heating treatments on the getter but with absolutely no reaction of any kind.

    I was running out of ideas. Normally, I would not expect the alignment at both ends to have changed but after a comment from Daniel Ames ( I decided to do some more fiddling with the mirrors at both ends of the tube. While applying pressure to the anode-end mirror mount with a piece of wood (dry and well insulated!) I pushed on the cathode-end mirror mount in the opposite direction (this retains parallelism and is equivalent to 'walking the mirrors' for an external mirror laser). Guess what? I found that there was an orientation where this would result in significantly increased power. So I took a chance and bent the anode-end mirror mount by carefully calculated amount. In other words, at random. :) Well, actually by an amount that was approximately sufficient to result in the decrease in power when pushing on the it previously. Then, I adjusted the cathode-end mirror mount for maximum power.

    I am now getting about 1 mW (compared to .35 mW when the patient arrived) without any of the special Siemens chants (those should help, right?). However, I don't think mirror alignment will go much beyond the 1 mW barrier. I suspect that the gain of the tube is still somewhat low and that the slight misalignment at both ends resulted in a much more dramatic drop in power than it would have when the tube was new. I doubt that the alignment changed much by itself (the tube was inside a sealed laser head so I know that it hadn't been touched by anyone else).

    Attempting to Revive Some Soft-Seal HeNe Tubes

    I recently acquired about 2-1/2 dozen soft-seal HeNe tubes in varying stages of decay. Specifically, these are the Spectra-Physics Model 084-1 HeNe Laser Tube, a type commonly used in early barcode scanners. These use soft (Epoxy) seals for the fixed (totally non-adjustable) mirrors bonded to the tube end-plates. Most of the glass part of the tube is wrapped in thick aluminum foil (probably for thermal stabilization - this is common with even newer Spectra-Physics HeNe tubes such as their models 88 and 98), has an attached 100K ohm ballast resistor stack in heat shrink tubing, and rubber end-caps to more or less protect against shock and damage. (More details can be found in the section: An Older HeNe Laser Tube.)

    I performed an evaluation on each one just long enough to determine functionality and initial power output, if any. The 30 some odd tubes came through as follows:

    These HeNe tubes have getter electrodes and associated getter spots. All the weak or non-lasing tubes showed noticeable deterioration of the getter spots with varying degrees of white or brown deposits. (In fact, 1 of the dead tubes was missing its OC mirror totally and 2 of the others were cracked with interiors that looked as though they had been stored in salt water or something else that resulted in crusty deposits and actual etching of the glass, cause unknown.) The good tubes have a spot which has mostly the normal metallic black appearance.

    I decided to try my solar heater getter reactivator first. This proved to be a big mistake. :( Since there is no way to aim the solar beam to the getter electrode without passing through the powdery stuff, it gets heated the most and apparently releases all the old trapped gases that it had been accumulating over the years (probably 20 or so). Tube #1 (below) went from a pink discharge and no output (but probably very near threshold) to not being able to start at all. :(

    My next thought was to get back to my getter heater project and finish the coupling coil - but that sounded like too much work! Perhaps, if I had been more patient, those renegade gas molecules would have been reabsorbed but I didn't want to wait. So, I decided to try reactivating the getter of tube #1 by putting it in a microwave oven. Hey, what the heck - with a half dozen otherwise useless HeNe tubes, I could experiment! :) Unfortunately, I tried pressing my luck too far by leaving the tube to cook for just a bit beyond well done - and the glass cracked (what can you do with a capillary attached to a mirror?). If I had gone a little easier on it, the outcome would likely have been positive. It took a couple of hours to build up the courage to try the others (with shorter bake times of a few seconds and checking for hot spots after each one). However, the results were mixed and I'm now somewhat confused. The patient status list follows:

       Patient                 ----------- Power Output (1) -----------
        Number     Original     After Treatment    2 Days    1 Month
             1      0.0 mW                  NA - Cracked (2)
             2      0.0 mW          1.5 mW         1.7 mW    1.4 mW
             3      0.1 mW          0.6 mW         0.3 mW    0.3 mW (4)
             4      0.5 mW          0.5 mW         0.5 mW    0.5 mW (5)
             5      1.0 mW          0.8 mW         0.7 mW    0.7 mW
             6      1.2 mW          1.1 mW         1.0 mW    0.8 mW
             7        --            1.7 mW (3)       --      1.2 mW

    1. The power output of each tube is listed for its initial test, immediately following microwave treatment, and for subsequent followup tests the next day. In all cases, these data are for the maximum power after a 10 to 20 minute warmup or where its value had obviously stabilized.

    2. Patient #1 died from excess abuse in microwave before treatment protocol could be determined. R.I.P. (Rest In Pieces - Organs including mirrors and capillary may be useful for transplants).

    3. Patient #7 was thought to be DOA since it wouldn't start. It was put in the microwave oven just to confirm. Since the light show unexpectedly looked fairly normal, treatment was quickly aborted. However, as a result, no initial output power reading is available. (I expect the starting problem was unrelated to the tube's condition.) The decrease in power after 1 month may simply have been back to its pre-nuked value.

    4. Patient #3 has had additional treatments. See the section: Followup Experiments with a Low Power RF Source on Patient #3.

    5. Patient #4 was retested on a return visit and found to be outputting a TEM10 beam. The cause was possibly a speck of dust on the inside of the OC mirror. Some gentle tapping has at least partially dislodged or moved this flake (it can still be seen and won't go any further) with a resulting TEM00 beam and power output increased to .7 mW. The of the still low power (for a good SP084-1, it should be at least 1.5 mW) may still be due to this cranky bit of dust as it can still be seen in the internal beam.

      While patient #4 was on the treatment table, the RF exciter was turned on with absolutely no effect. Since there is no evidence of gas contamination, this isn't surprising.

    As noted, rated power of these tubes is probably about 2 to 3 mW. From this data, it would appear that the tubes in the worst shape are likely to benefit the most.

    Something that can be seen from the data and appears somewhat peculiar is that cooking certain tubes just long enough so that the microwave induced discharge glow reaches full brightness resulted in a substantial increase in output (when powered in the normal manner). However, after a few minutes (or maybe a day or so), the output power would decay back to its original value (or below as with tubes #5 and 6 - though the original values may be suspect and the decay may have happened regardless of whether the microwave treatment was attempted). Tube #3 peaked at about double its final value but still retained a 3-fold improvement compared to its condition upon arrival. In any case, I now believe that whatever is going on isn't strictly related to the getter - maybe also a combination of the heat resulting from the microwave treatment releasing trapped helium and/or neon from the walls of the tube (low gas pressure originally), helium deficiency due to diffusion through the tube walls/seals, or a phenomenon that is totally independent (more discussion below). Since this behavior can be repeated at will for those tubes that exhibit it - a quick shot in the microwave and you get a nice, but temporary boost in power output, which may have it uses. :) (Patient #3 has agreed to some additional experiments to determine if extended operation or other more advanced treatments can actually clean up contamination.)

    The result with tube #2 was impressive (in a relative sort of way, and more so since it appeared to improve further with a day's rest) but I have no idea why. I imagine that at least with respect to the getter, some of these tubes probably had no available un-activated getter material remaining in the getter electrode, nothing to activate. Tube #2 must have had a wad of the stuff hiding somewhere. :)

    (From: Consulting Laser Physician Daniel Ames (

    About Sam's microwave HeNe/getter soup recipe:

    From what you have described above with the 6 patients (data for patient #7 wasn't available at the time of the consultation. --- Sam), I can only surmise about the results showing power peaking and decaying. If the tube was tested within only a few minutes after being removed from the microwave oven, then I would suspect one or more of the following to have ocurred:

    1. He and Ne were released from the aluminum cathode and some from the anode metal, plus maybe some from the inner glass walls.

    2. O2 and other impurity gases were absorbed by the getter.

    3. Output power could start out higher than before the getter baking treatment, but here is what I suspect might have caused the output power to decrease after a few minutes: The aluminum cathode most likely absorbed more of the microwave energy than any other part of the tube and therefore it also became the hottest. As it began to heat up, it released some He and Ne but the treatment only lasted for what - 2 seconds approximately? Successful outgassing of metals in a vacuum even as high as 2 to 4 Torr requires more than just 2 seconds. I suspect that one of two things caused the power drop off:

      • The cathode remained hot enough for a few minutes and continue outgassing He and or Ne and possibly raised the tube's He. and/or Ne partial pressures to (above) the desired pressures and caused a power drop off.... Maybe.... Before and after discharge voltage and current tests should confirm this, either way, see below. Or:

      • The tube's gas temperature was elevated by the microwave energy which caused the He and Ne partial pressures to elevate too. If the tube originally (before treatment) was He deficient, than the elevated He pressure could actually cause the output power to temporarily increase until the tube cools back down to a normal operating temperature and the partial pressures of both the He & Ne have decreased again. One way to test this could be to wrap the tube with a thermal insulating material (e.g., Fiberglas) to trap the heat generated by normal use in order to raise the gas pressures, then monitor the output power.

      • The Weird Science Theory: If the tube was immediately powered up with its normal power supply after being removed from the microwave oven, could it be possible that the He and Ne gases are behaving in a similar fashion to food when it is cooked in the microwave oven - I.e., the gas molecules are caused by the microwave energy to vibrate and possibly just like nuked food, its molecules are still vibrating for a few minutes after the tube is removed from the M/W energy and immediately powered up by its normal power supply and behaving differently until they stop vibrating???????

    What about measuring and comparing the operating voltage and current on tubes #5 & 6 above with the reading from tube #2, since #5 & 6 actually dropped in output power below that of their respective (originally) observed power. This could give us a clue as to whether tubes #5 & 6 are actually higher in pressure or lower than tube #2.

    It would probably be easier on the glass tube and it's geometry if it was powered up and thus heated up to normal operating temps (just) before subjecting it to the intense heating of the metal parts of the tube by he microwave. The aluminum cathode will expand in diameter in the microwave, the metal anode too, so by allowing the normal power supply to heat up the glass and metal parts first at a normal rate of expansion, then it should have a better chance of survival in the microwave.

    HeHeHe..... and I thought this would be a 1 paragraph reply..... hehe :) I'll submit my usual bill for services. :)

    (From: Sam.)

    I could believe partial pressures increasing or He being released (I've been more convinced that He depletion may play a part though it isn't obvious from the discharge color). However, the metal parts of the HeNe tube actually remain quite cool. This could mean that any effect on the getter may actually due to the glow discharge and not the actual microwave heating though on tube #1, the getter glowed orange hot after a couple seconds once the tube had cracked - I suspect it doesn't get heated nearly as well with the surrounding gas competing for microwave attention. The glass between the cathode and anode of the tube gets hottest (which is what cracked with patient #1) but the cathode itself doesn't appear to get very warm at all.

    I really doubt any molecular vibration effects apply here - those sorts of phenomena have time constants measured in small fractions of a second. The behavior seen with patient #3 was on the order of 20 minutes.

    My current feeling is that the odd behavior is due to a combination of heating and release of gases from the tube walls and that the fundamental problem is one of low gas pressure but not a particular lack of He or Ne. I do expect to measure tube operating voltage and current producing maximum output (what of it there is). I may also attempt a helium soak starting with the tubes having the lowest output power (though none of the tube's spectra appeared to be obviously abnormal).

    Followup Experiments with a Low Power RF Source on Patient #3

    A few weeks after the original microwave revival experiments, patient #3 returned for some more extensive tests.

    Operating a HeNe tube is supposed to result in the scavenging of residual gas molecules due to the cathode acting as a sort of getter. So, I decided to perform a very scientific experiment on the most bedraggled of my assortment of Spectra-Physics 084-1 HeNe tubes - the one with the lowest output power and most off-color discharge - patient #3.

    I started by operating patient #3 for hours on end at 5 mA. Early in these tests, the output power would fluctuate quite substantially - dipping to as low as 0.1 mW at times. After perhaps a total of 24 hours of actual running time over the course of several days, the power has tended to stabilize somewhat, remaining over 0.25 mW at all times and peaking at 0.4 mW with an average of about 0.35 mW.

    However, additional operation hasn't resulted in any substantial improvement beyond this point. A few of observations:

    Thus, I concluded that since gas-metal reactions at the anode electrode are minimal, further improvement wouldn't be likely in any case since none of the rogue gas molecules were bumping around at the cathode where they might be taken out of circulation. Reversing polarity would sweep them to the other end of the tube but (1) running a HeNe tube with reverse polarity will quickly damage the anode mirror from sputtering and (2) the molecules will again congregate at the wrong end of the tube and stay there!

    Based on these observations, some other treatment would be required - something that would facilitate reactions at the getter and/or cathode but which wouldn't damage the mirrors.

    Using the microwave oven approach, that tube could be temporarily boosted to 6 times its original output power with it remaining more or less at a 3X improvement. So, I decided to do some additional experiments similar to these but under more controlled conditions.

    I repeated the glow discharge treatments on patient #3 but using a flyback based high frequency RF exciter instead of the microwave oven. With this approach, both the location of the discharge and the power level could be selected at will. In particular, the power could be set low enough that the discharge could be maintained indefinitely without fear of physical damage to the tube due to overheating.

    For the RF exciter, I adapted the circuit described in the document: Simple High Voltage Generator. The new schematic, with the high voltage rectifier removed is shown in Flyback Based RF Source and the major parts in ASCII, below (shown attached to a modern HeNe tube):

       +Vcc     Q1   +----------------+                               A||
         o           |                 )::                           .-''-.
         |       B |/ C                )::                           |\  /|
         |  +------|    2N3055         )::                          || || |
         |  |      |\ E             5T ):: +------------------------|| || |
         |  |        |                 )::(                         || || |
         |  |       -_-                )::(                          | || |
         |  |                          )::(                          |G|| |
         +--|-------------------------+ ::(                          |_||_| LT1
         |  |   Q2  _-_                )::(                          | || |
         |  |        |                 )::( Secondary (HV) winding   | || |
         |  |    B |/ E             5T )::(                          | || |
         |  |  ----|    2N3055         )::(                          |    |
         |  |  |   |\ C                )::(                          | C  |
         |  |  |     |                 )::(                          |____|
         |  |  |     +----------------+ ::(                          '-..-'
         |  |  |                        :: +--------------------------+||
         |  |  -----------------------+ ::
         |  |                       2T )::
         |  |               +---------+ ::
         |  |               |       2T ):: T1 - Flyback transformer from B/W or
         |  +-------------------------+         color TV or computer monitor.
         |                  |
         |            R1    |    R2
                      110        27   _|_
                      5W         5W    -

    With no high voltage rectifier, the output is radio frequency AC at between 10 and 20 kHz. This was applied between the cathode mirror mount and a 2" strip of aluminum foil wrapped around the tube to provide capacitive coupling for the return path without involving the anode-end mirror mounts (and thus avoiding the possibility of sputtering). There is absolutely no glow inside any part of the bore or near either mirror mount. In addition to allowing capacitive coupling through the glass of the tube, the AC would also assure that the gas molecules wouldn't get stuck in one spot. This circuit produces a nice glow when powered from only about 5 VDC at 1 A or so it runs cool. The visual effect is similar to that of a plasma globe operated at low pressure and as with those gadgets, the glow could be influenced by touching the glass of the tube.

    The physical connections to one of our patients is shown in RF Treatment of SP084-1 HeNe Laser Tube. Note that this way of exciting the gas in the HeNe tube will not cause the tube to lase as there is no high intensity discharge in the bore.

    CAUTION: If you try this, take care not to use too much voltage or the glass may be punctured! Spectra-Physics HeNe tubes have nice thick glass walls so the risk is quite low but don't press your luck - it isn't voltage but power transferred to the plasma that should matter so really high voltage isn't required. In fact, I'll be trying a coupling coil instead of capacitance through the glass next.

    The most effective position for the aluminum foil wrap to have any effect on tube performance was about midway between the end of the cathode-can and the anode mirror mount. This resulted in the glow discharge bathing the getter and end of the cathode. A test with the foil wrapped around the cathode area of the tube resulted in minimal effect despite the close coupling and nice glow.

    As with the microwave oven treatment, the RF also resulted in a dramatic increase in output power for patient #3. In fact, although my records are non-existent, I believe that this resulted in even more of a boost to over 0.8 mW. Of course, I could run the RF discharge for a long time (several minutes in this case so far) compared to a few seconds for the microwave treatment (before there was risk of overheating and killing the tube). But, as before, the output power still decayed back to its original value over the course of a half hour or so.

    Accompanying the power increase was a distinct improvement in discharge color. Recall that originally, the discharge was somewhat pink and charged to a somewhat blue color at the anode with almost a neutral white in the funnel next to the anode. The new color was much more normal though possibly a bit on the orange side indicating an excess of neon or lack of helium (as before with the microwave oven treatment). In fact, the funnel discharge color was distinctly orange - more so than is typical of healthy HeNe tubes. Another change was that the tube's operating voltage declined by up to about 100 V when compared to its value with the off-color discharge. (This is the opposite effect observed with the "Northern Lights" tube - see the section: Repairing the Northern Lights Tube. One thing that has been confirmed is that heating plays little or no role in the power boost - the RF approach results in very little heating of any part of the tube.

    Next, I set up the RF exciter to run at the same time as the normal HeNe power supply so I could monitor the beam power while tickling the gas outside the bore. With this configuration, output power could be maintained at a much higher level, though not at the absolute maximum that could be achieved by 'off-line' RF treatments.

    So, it appears as though maintaining a modest glow discharge outside of the bore can be used as a means of life support for these marginal soft-seal HeNe tubes. Too bad about the additional high voltage (the foil) that needs to be well insulated. :) If only HeNe tubes had gas return channels from the anode to the gas reservoir! (A helical capillary longer and narrower than the bore would prevent the normal discharge from taking that shortcut.) Then, there would be a steady flow of gas and even without the RF, there would be no concentration of contaminants in the bore or near the anode. With the RF active, there would be continuous cleaning and instant purifying action!

    More to follow. :)

    Treatement for the Yellow HeNe Laser Tube with a Warped Bore

    This is the upper tube in Three HeNe Tubes of a Different Color Side-by-Side. The OC (and anode connection) is at the left with the cathode terminal and getter visible below it. No attachment is made to the OC mirror mount on the right. This may be made by PMS/REO based on its style though I don't know for sure.

    The fact that gentle tapping affected the behavior suggested that something was loose inside. And, even pointing the tube up in the air at various angles would occasionally result in at least a weak output beam - perhaps the tube would be useful as an inclinometer. :)

    At first, there was no visible indication of loose parts - its general condition is quite good. However, upon close examination, the bore is supported at the OC-end of the tube by a cup affair which had a set of fingers that look sort of like the pedals of a tulip and these were actually loose around the bore. Either the tube had been used to hammer nails, or the mirror mount next to the cathode can had been accidentally used as the cathode connection causing local heating. Since most modern HeNe tubes use the mirror mounts for both power supply connections, a natural mistake is to attach the negative of the power supply to the cathode-end mirror mount. While this would result in the tube appearing to operate normally, there will be serious overheating of the mount and possible sputtering of the OC mirror. The overheating could cause the petals to relax and loose their grip but any sputtering overcoat on a low gain yellow OC mirror would almost certainly result in *no* output. Since there are signs of life, this scenario is therefore unlikely.

    In fact, pressing laterally on the HR-end mirror mount - not to deflect the mirror but to actually move the entire bore slightly by flexing the glass of the tube - would result in a strong good quality beam. Interestingly, even careful adjustment of the mirror alignment at both ends - but without this external force - would only produce a weak beam with much less power than possible with the added deflection. The OC mirror mount could be easily rocked without affecting anything else. However, for the HR mirror mount, I had to construct a Melles Griot style three-screw locking collar for this test to be able to make slight adjustments in the alignment without permanently bending the mount. Otherwise, any effect would be a combination of the bore being moved and the mirror alignment with respect to the bore changing.

    It appears as though the HR mirror was correctly aligned as just changing this relationship would only result in lower maximum output though it was possible to reach a compromise where the tube produced a steady beam by also tweaking the OC mirror alignment. However, this was less than 1/2 the possible power available by just the bore movement technique.

    My theory is that the bore is actually slightly warped - though I can't tell by looking at it. If it were just improperly positioned, realignment of both mirrors should have resulted in a strong beam equal or nearly equal to its original performance. Given that this didn't happen, I am forced to the conclusion that the lateral deflection not only moves the bore but also unwarps it to some extent. Another indication of a bore problem is that just adjusting the mirrors tends to result in a TEM10 rather than the expected TEM00 beam. However, as the lateral force is applied, the beam starts out TEM10 and then the two sub-beams merge to form what looks like a TEM00 beam though I haven't confirmed that this is actually so. With the cathode can obscuring most of the interior, it is impossible to see if there are other internal problems. It needs to have an X-ray or CT scan. Is there medical insurance for sick lasers? :)

    To deal with the chronic condition - there is after all no practical way to actually go in there and really fix the problem - I intend to construct a mount for the tube that will also have a lateral force adjustment. Some experimentation (actually quite a bit of it) has revealed that the optimal force seems to be low enough that there is minimal risk of breaking the tube, though I'd be happier with some other solution.

    So, I mounted the tube in the head cylinder from a Melles Griot 05-LHR-151. The HR mirror mount was covered with a rubber boot, around which I placed a plastic ring with a 4-40 tapped hole on one side. A strategically placed hole through the head cylinder allows a screw to thread into the ring and by very careful adjustment, pull the HR mirror mount to one side ever so slightly. With a bit of experimentation, the optimal orientation was determined and marked. After a slight detour where the HV arced to my adjustment screw, it is now stable after warmup at about 1.4 mW at 594.1 nm, which isn't bad for a tube of this size. I will have to add a prominent CAUTION sticker to alert anyone that they should not attempt to tighten that magic screw as bad things might happen.

    Did I mention that the yellow 594.1 nm wavelength is my favorite. :)

    Awhile later, a complete yellow PMS LHYP-0021 (yellow 594.1 nm, 0.2 mW spec) laser head came in with a similar condition - it would only lase at a particular orientation, but at least this was reliable and the output power is decent (0.5 mW) considering its ratings. However, being a polarized laser, it would be desirable to be able to set it up without regard to such quirks! I don't think the bore was actually warped for this laser, but rather that it had been whacked in shipping and that petal assembly had been bent so that it wasn't clamping the bore firmly in the correct position.

    Reviving a Spectra-Physics Model 130B Antique Laser

    The SP-130 may have been the most solidly constructed of any small gas laser in history! See: A Typical SP-130 (Note original manual). The case, which is also the support chassis for the tube, external mirror mounts, and power supply is built of precisely milled aluminum panels. Everything fits together like a fine watch (if you remember those!). Versions of this laser were produced as early as 1965 (that is the date on one of the diagrams in my original "Spectra-Physics Model 130 Gas Laser Operation and Maintenance Manual", the one in the photo, above.) More information can be found in the section: Description of the SP-130 Laser.

    This is the third sample of the Spectra-Physics 130B laser that I have acquired. I don't know if there ever was an SP-130A but the SP-130 may have been an earlier version using a tube with a heated filament instead of the more modern cold cathode design.

    SP-130B #1 initially started and had a discharge that was weak though approximately the correct color, but died on the operating table - cause unknown. The discharge winked out, never to return. All indications are that the tube is up to air except that the getter hasn't changed to the "white cloud of death" appearance.

    Apparently, it didn't really die but would not stay lit and was so hard to start that it I never succeeded in restarting it again. Even testing with an Oudin coil appeared to confirm it was up to air, but there was probably a very faint glow, not visible under normal room lights. I gave this laser to a friend of mine who thought he could talk a friend of his into regasing the tube. He managed to get it started exactly once, like me. More on this in the next section.

    SP-130B #2 (the actual laser in the photo, above) was DOA with an up-to-air tube and some prior dissection attempts including cut wires. The mirrors were also totally ruined, possibly from poor storage conditions or careless handling or both.

    Which brings us to SP-130B #3. This one started and ran fine but the discharge color was initially red/blue, along the lines of the example labeled "Moderate - no output" in Color of HeNe Laser Tube Discharge and Gas Fill. These are normally hopeless and terminal but I figured it wouldn't hurt to run the laser for awhile just in case a miracle occurred. In fact, over a period of several hours, the color did gradually change eventually approaching something reasonable, at least in the bore. (Normal is defined as "salmon" or white-ish red-orange and more of an orange color in the expanded areas.) The color in the expanded areas was not as orange as would be normal but was fairly close. But there was still no output.

    Next step: Check mirror alignment and clean optics. First, I removed the HR mirror and used a working HeNe laser on an adjustable platform to check OC alignment by passing its beam down the bore and looking at the reflection back to its output aperture. This appeared to be slightly off center, so a bit of tweaking was in order. Then, I replaced the HR and adjusted it to also place the reflection squarely back into the alignment laser's output aperture. Still no output.

    During this time, I also attempted to clean the optics as best I could knowing that the mirrors might be soft-coated and in that case can't be cleaned with anything stronger than breath-fog. :) The mirrors and Brewster windows were cleaned without incident but the Anti-Reflection (AR) coating on the OC mirror didn't survive so there would be slight ghost beams if the laser was to work at all. The sticky tape method of mirror glass retrieval recommended in the SP-130B manual also removed the coating. :(

    Next, I decided to actually consult the manual with respect to alignment - what a concept! :) Their procedure is even simpler than mine: Using the curved mirror set, just tighten both mirror mounts down so they are flush with the case. The machining is precise enough that this should produce a beam. I only have a curved OC, the HR is planar. So, I tightened down the OC mirror mount and checked it with my HeNe alignment laser - at least as good as doing it my other way.

    Doing the same with the HR mirror mount didn't produce a beam, but when I loosened it slightly, I could jiggle the mirror just enough... And, for the first time in perhaps 20 years, I detected a few coherent photons in a flash from the OC. After somewhat more tinkering and letting the system bake, it was doing between 10 and 40 microwatts depending on the setting of the current adjust pot. Maximum output is at the full clockwise position which suggests that there is still gas contamination or possibly just low helium - it doesn't peak as expected at some intermediate value. After cleaning the Brewster windows (at least they probably won't disintegrate like the AR coating if looked at the wrong way!), output power has exceeded 0.25 mW, not up to spec (0.75 mW) but still a lot better than 0.0 mW and a bit amazing considering the age of this laser.

    So, this patient will be held in intensive care for some time to determine if any more cleanup takes place. I also suspect that a shot of helium would be beneficial. Given that air (probably) has leaked in, helium has likely leaked out. Also, the blue-green portion of the spectra of the discharge appears a bit weak - that is mainly from the helium. What I don't know is the age of the tube (it was probably a replacement) but it is probably at least 10, possibly 20 years old. When I do get around to a helium soak, I'll probably start with 10 days (1 day/year of life) to be on the safe side and see if that helps. It's bad form to overdo it by much - you can't reverse the process except by waiting 1 year for each day of extra helium!

    However, it's been over 6 years now and there has been no noticeable change in performance. The laser was obtained in November 2000 and it is now January 2007. I run it for a few seconds almost daily and that seems to be enough. If not powered for a of couple days, there will be no output initially, but it will come back in less than a minute. So, I rather doubt the tube is leaking in significantly but simply has some internal contamination. There could be some helium diffusion through the glass. I haven't even cleaned the Brewster windows in several years.

    Restoration of a Spectra-Physics Model 130B Antique Laser

    A friend of mine, Phil, managed to sweet-talk a major laser service company to regas the almost impossible-to-start plasma tube in this laser. OK, he actually bribed them with a dead but intact large frame argon ion laser and some other goodies. I never expected them to come through, but they did - after 6 months or so. It wasn't a full refurb but just a "chop and fill" and the expected life may be rather short - perhaps 100 hours. But that's plenty given the only use is likely to be for Phil to turn it on to show people that a (second) working SP-130B really exists!

    They actually sent me two lasers - the regased SP-130B and an SP-130 "parts" unit with an up-to-air tube that had been cluttering their back room. The latter is actually in rather good condition other than the tube and missing trim strips. In particular, the mirrors were good, which came in handy.

    The laser service company was unable to get a beam after installing the regased tube. This wasn't entirely surprising as I knew the OC mirror had been damaged when I removed the plasma tube to inspect it and dinged the mirror surface with the Brewster tip. I thought the central portion of the mirror was in good condition, but perhaps not enough of it. However, I doubt they tried too hard in any case. Even a chop and fill operation entails a fair amount of work. So, by the time they installed the tube, there was probably little enthusiasm for a complete alignment.

    My initial attempts to get this patient to lase were also unsuccessful. I did install the OC mirror from the "parts" SP-130, which after cleaning, appeared pristine. Alcohol had no effect on the surface finish so I assume it to be hard-coated. I also checked the Radius of Curvature (RoC) of both HRs and OCs to confirm they agreed with the SP specs (planar and 30 cm, respectively).

    So, having failed to detect a single coherent red photon, I decided to order up a single pass gain test to confirm that the gas-fill was done properly. No sense wasting a lot of time trying to obtain lasing if the gain is too small!

    For this, I used an SP-117C stabilized HeNe laser since its output power would be quite constant after it had warmed up and locked. The SP-117C was placed on an adjustable platform and aimed down the bore of the SP-130B with both mirrors removed. By turning the SP-130B discharge on and off, the change in transmitted power could be measured easily. The contribution from the discharge light was also checked with the SP-117C beam blocked. The result was about 1 percent - certainly enough to lase with the normal OC mirror with a reflection coefficient of around 99 percent, since the round trip or two pass gain would be 2 percent. (This doesn't account for losses from the Brewster windows, but if reasonably clean, this should be well under 0.5 percent total.)

    Without the front mirror in place, the condition of each of the Brewster windows was also checked for scatter inside and out using the alignment beam. They were cleaned to minimize scatter from the outer surface. There appeared to some excessive scatter from the *inside* of the rear (HR-end) Brewster window, but probably not enough to prevent lasing.

    Next, the HR mirror was replaced and aligned to direct the reflection of the SP-117C beam directly back to its output aperture. (This retroreflected beam might destabilize the SP-117C and cause it to lose lock, but would not really matter since it was only being used for alignment at this point.)

    And then, the OC mirror was installed and aligned so the reflection from its surface also went directly back to the SP-117C output aperture. Still no lasing even after loosening the locking screws of the mirror mounts, first at one end and then the other, and jiggling. :)

    The OC (front) mirror is curved - 30 cm RoC - and alignment should be less critical than the planar HR (rear) mirror. So, it was fine tuned and then the locking collar on the real mirror tube (not the mount, but the slide). In addition to allowing the mirror distance to be changed, there is some unavoidable "slop" in the alignment. Finally, a flash! I had been close to giving up suspecting that Brewster scatter was too much, but once the flash was detected, it was quite easy to fine tune the rear mirror to get a sustained beam, then dust off the Brewsters windows to increase the power, and walk both mirrors to peak it. At least as an initial attempt.

    As noted, the discharge color was more orange than normal and very bright. It occured to me that perhaps the "very bright" part was not only a result of overfill, but also excessive current due to a lower operating voltage than is present at the design pressure. Sure enough, running on a Variac, the peal output power occured around 90 VAC instead of 115 VAC. The knob on the front of the SP-130B was always set at the minimum current. Aside from eliminating an odor of overheated electronics, running at the optimal current increased output power by about 30 percent. In fact, turning the knob up killed lasing entirely at normal line voltage.

    So, as it stands now, the output is about 0.57 mW after warmup at optimal current. I expect this can be improved with additional cleaning of the Brewster windows and mirrors, but I'll leave that to Phil. The rubber boots also don't seal very well but are better than nothing for now. A new set would be a good idea.

    I built an input voltage reducer from an HVAC control transformer so the input to the SP-130B is now about 75 to 80 VAC and the knob actually peaks the output power about 3/4 of the way up. The transformer is wired with its dual primary windings in series to produce about 59 VAC from the centertap, and 17 V is added to this from its secondary (now running at 1/2 the normal voltage). I wanted to put this inside the SP-130B case but there simply is no room. A purely resistive voltage reducer would dissipate significant power and even for that, there is no room.

    Conclusions: This laser is now operational. At present, the output power is somewhat low but this is almost certainly in part due to the need for additional optics cleaning, and also due to being overfilled. While not quite as totally authentic vintage as SP-130B #3, it is close. The glasswork is very unobtrusive so the only tell-tail indication of rework is the dead getter since there was no way to open it up without admitting air, and no way to install a new one without more time and effort than could be justified.

    Thus, I know now of 2 working SP-130/B lasers in the Universe! :)

    Two Melles Griot HeNe Laser Heads with Terminal Sputtering Disease

    These were probably nice high power laser heads at some point in the past but now were clearly in deep trouble. (Thankfully, someone else had already removed the HeNe tubes from the aluminum cylinders so the diagnosis could be made a lot more easily.) At first I thought the gas fill was contaminated somehow (because of the funny white-ish color) and even went so far as to try activating the getter with my Solar furnace with no change at all.

    The key symptom which didn't register at first but is obvious in retrospect were several silvery metallic spots around the tube next to the cathode end-cap/mirror mount. On many Melles Griot tubes, the cathode has a set of 4 holes punched through it equally spaced around its periphery. Normally, it is possible to vide the interior of the cathode and end of the bore through these holes. Not now. What the metallic spots must be are deposits of aluminum on the glass due to very serious sputtering taking place inside the cathode. Inspecting the bore as best I could (until an autopsy can be performed), it would also appear as expected that there are similar deposits on it near the end inside the cathode can. Whether the sputtering was simply from normal end-of-life when the cathode can pickling (oxide) gets used up, from some manufacturing defect, or from abuse, I do not know. The laser heads had closely spaced serial numbers so it's possible they were from a bad batch, or just from a set of lasers shipped to the same customer and used under similar circumstances.

    Unfortunately, prognosis is poor and salvaging the organs for transplant (e.g., the mirrors) may be in their future. :)

    Four Melles Griot HeNe Laser Heads with Broken Bores

    All four of these laser heads must have suffered some terrible trauma though there was no external evidence of bruises, scratches, scrapes, or dents. Perhaps an entire rack of HeNe heads had dropped to the floor. :( The location of the breaks were also interesting. On these tubes, the bore is supported at three places: the fused glass at the anode-end; the main spider about mid-way, and another spider which is part of the cathode. The breaks were between the two spiders, right at the main spider on the one sample that was naked, probably the same place on the others though I haven't extracted them yet. It's also possible that the double spiders in close proximity resulted in too much stress or a peculiar resonance under the wrong conditions. The bore is made of very thick glass and its extension into the cathode isn't that long. However, these are the frosted variety (inside and out) which I imagine to be weaker than those of similar size which is made of polished glass.

    By tweaking the mirror mounts while the tubes were oriented optimally, I was actually able to get one sample up to 2.75 mW which was stable as long as the tube wasn't moved or rotated. The others peaked at 1.0, 0.6, and 0.25 mW respectively.

    Except for the 2.75 mW tube, the others are destined for my organ, err, mirror bank. I'll probably pull the 2.75 mW tube from its cylinder and keep it as a sort of curiosity and warning to any other HeNe heads that might be tempted toward recklessness. :)

    Spectra-Physics Model 120 HeNe Laser Head with Terminal Gas Leakage Disease

    With a diagnosis of terminal gas leakage disease, the only course of action is a tube transplant. Fortunately, I had another good tube (in a resonator) for this purpose. Either the bare tube or the entire resonator could be replaced. I chose to remove the entire resonator and install my spare intact rather than swap tubes since it is slightly lower risk but a replacement tube can be installed in about 5 minutes without requiring anything more than a touch-up of mirror alignment.

    The transplant went smoothly with the patient making a spectacular recovery. :)

    CAUTION: Don't be tempted to touch any of the coarse mirror alignment screws (the ones at 120 degrees around the mirror mount flanges - their setting is very critical and if you lose the beam, alignment from scratch will probably be needed. Use the pan and tilt screws (in the end-plates, on horizontally either side of the mirror mount flange) for all alignment. These shift the center of the bore with respect to the curved mirrors. If you can't get a a beam, the tube is bad, the Brewster windows or mirrors are dirty, or someone else messed with the coarse adjustment screws!

    Spectra-Physics Model 120 HeNe Laser Head with Moderate Gas Leakage Disease

    This SP-120 has a getter electrode but no obvious getter spot. Since every other SP-120 tube I've ever seen had a very noticeable metallic getter spot if still good, or the "white cloud of death" spot if beyond hope, I can only assume that for some reason or just lack of quality control, the getter in this tube was never fired - that may be an option if needed.

    The laser came in with no signs of lasing at any reasonable current setting but after 10 minutes of a steady 6.5 mA drip, coherent red photons started appearing in small quantities. Patient's chart of accumulated treatment time:

          Arrival  0.2 hour  3 hours   13 hours  24 hours  34 hours
          0.0 mW    0.1 mW   1.7 mW     4.0 mW    4.4 mW    4.6 mW

    The output power of 4.6 mW is less than 65 to 75 percent of what a new SP-120 will produce at a current of 6.5 mA. Presently, the tube will output 5 mW at 7.50 mA and 6 mW at 9 mA. But I don't know the recommended maximum current for the SP-120 and 9 mA was still not the peak, rather the limit of my power supply. In any case, 6.5 mA is always a safe value for this size HeNe laser. Although the other SP-120 tubes I've tested also peaked at a current higher than 6.5 mA (I don't recall what it was), as noted, their output was still much greater at 6.5 mA than the patient. Though 5 mW output at 7.5 mA might actually meet spec, treatment will continue for a few more days. :)

    For more on reviving soft-seal HeNe lasers, see the section: Care of HeNe Laser Tubes.

    Spectra-Physics Model 907 With No Output

    The SP-907 is the OEM version of the SP-127/107 laser tube and resonator with an overall length of more than 38 inches and a nominal output power of 35 mW. This patient came in with a cut power cable, broken cathode-end ballast resistor tube (only really affects appearance), and no power supply. The SP-207 (both linear and switchmode versions) is the recommended exciter but I don't have one. So, I had to adapt my SP-255 to the task.

    First, I just connected it as best I could with alligator clip leads to see if the laser would do anything. It didn't even flash with the input voltage cranked up to 140 VAC on a Variac. (The SP-255 is a linear power supply so boosting the input would boost the starting and running voltages as well.) I wasn't particularly surprised as the SP-907 tube is about 50 percent longer than the SP-124 for which the SP-255 is designed.

    On a hunch, I grounded the frame as it was not grounded originally. Then, exactly once, it started and continued to run until I backed the Variac down below 110 VAC or so. However, while lit, there was no sign of red output. The discharge color looked reasonable - perfect in fact - so this confirmed that the tube was gas intact and had no serious leakage. (A small getter spot was also present and looked reasonable as well. I don't know if the rest of the getter spot turned clear when used up of if this small spot was all there was.)

    But I couldn't get the tube started this way again no matter how long I held my breath. :)

    As a test I wired the tube backwards since with reverse polarity, the starting voltage is often somewhat lower though the operating voltage is higher. With this arrangement, it would occasionally flash but that's about it.

    Next, I returned the wiring to the correct polarity and applied some RF from my flyback HV widget via a strip of aluminum foil to the bore trying a few different places. When in contact with it relatively near the anode-end of the tube, the laser would flash on momentarily with the Variac to the SP-255 cranked all the way up but would never "catch".

    By accident, I did find out one interesting thing: If left alone for an hour or more, applying the full 140 VAC to the exciter suddenly without slowly turning the knob up on the Variac would result in it starting. But, only if allowed to sit for that hour (or longer). Hmmmm... Maybe it likes the output to climb quickly from near 0 V to its starting voltage, this somehow coupling via the tube capacitance and initiating the discharge. To verify this, I took a 400K ohm resistor and carefully discharged both the power supply and laser tube capacitance. And, presto! The tube started even without the hour's wait. In fact, it would now start at 125 VAC or sometimes even 115 VAC after only the time it took to apply the resistor.

    Great! So, I added a 200M ohm bleeder resistor across the power supply output and attached a nice Alden cable to the laser head. This enabled it to start and run reliably but it might require 125 VAC for starting after which it could be backed off to 115 VAC while running to reduce stress on the SP-255 pass-bank. Later, I added an external pod with a stage of boost circuitry to increase the SP-255 starting voltage. (See the section: Enhancements to SP-255.) This eliminated all starting problems and the need for the Variac. I set the operating current at 10 mA which should be enough (11.5 mA is nominal but I'd rather run it a bit low until later).

    I did dust off the Brewster windows - at least they are accessible after pulling back a rubber boot (unlike the SP-120 where it's impossible to clean them in place). No change.

    At this point, it is almost certain that the major problem is mirror alignment. I emailed the person I got it from and asked: "Before I attempt to align this beast, do you know if the mirrors have been touched?". Reply: "Well, maybe someone attempted to peak the power and totally lost alignment." "Duh, thanks for telling me." :)

    My first approach was to use the "bore sight" method of mirror alignment because I felt there was no way to get a HeNe alignment beam cleanly down the bore. The "bore sight" method allows all alignment to be done by reflecting from the mirrors externally, using a pair of cards with small holes positioned at the tube's axis to align the alignment laser to the tube. (See the section: Major Problems with Mirror Alignment, earlier in this chapter.)

    I used my trusty little 05-LHR-911 HeNe laser head on an adjustable platform to align its beam through the cards, which had previously each had a hole drilled precisely at the location of the center of the SP-907's mirrors. This worked reasonably well for the OC-end and confirmed that the OC mirror was way out of alignment - by 1 or 2 whole turns of the 1/4-28 adjustment nuts! (There would be no lasing on this long a resonator if the nut was off by even 1/10th of a turn!) So, someone really messed things up. :(

    However, I didn't count on what I found next: The outer surface of the HR mirror is coarse-ground (frosted), not polished, so there is no way to reflect a beam from it which this method of alignment requires. Why did SP do that? :(

    So plan A didn't work.

    Plan B is to do everything from the OC-end starting with removing both the HR and OC mirror mounts (just 3 nuts each so at least that's easy) and start by getting as much of my HeNe alignment laser beam through the bore, then installing the HR and aligning for a reflection back from there, and put the OC in and do the same.

    I fabricated some precise micrometer (80 tpi, the mirror adjusters from a large ion laser) adjustment plates and attached these (2 screws at one end, 1 screw at the other) to the laser head. This will provide the degree of control I need to align the tube's bore with the alignment beam. Providing fine pitch screws centered at each mirror of the laser being aligned rather than on the alignment laser results in a much more intuitive setup since there is almost no interaction between adjustments.

    Finally we have lasing!

    What a pain. In addition to the mirrors being all out of alignment, there are adjustments on bore straightness which were also messed up and it was was impossible to get any resemblance of a clean beam down the bore from my HeNe alignment laser. But, with a bit of careful tweaking, a spot was detected on the wall acting as a screen that was clearly from the alignment beam. Then, I replaced the OC mirror mount and aligned its back-reflection to coincide with the HeNe alignment laser's aperture, with dancing interference patterns. Finally, replacing the HR mirror mount and after a few minutes of gentle rocking, flashes where detected. :) Once a stable position was found for the HR (just sitting on the rods), the OC mirror was carefully adjusted to maximize power - still probably less than 1 mW. Then, the HR mirror mount nuts and washers were installed and carefully adjusted to tighten up the mount, never losing sight of the beam! Finally, I walked the mirrors to peak power. I will say one thing, these mirror adjustments are very smooth and repeatable with little backlash even though the entire range of lasing is probably less than 1/10th turn on the nuts.

    Note that I didn't follow the original Plan B procedure exactly taking the short cut of using the OC back-reflection to align it first rather than attempting to get a clean return beam back down the bore from the HR. Fortunately, it was successful.

    This SP-907 currently peaks at 18+ mW but will probably do 25 mW, maybe more, when run at the optimum current (it's still at 10 mA) with a proper cleaning of the Brewster windows - which is still a pain since they attract all sorts of stuff as soon as they are cleaned, and my operating suite isn't exactly a Class-100 clean room. Power typically drops way down just pushing the rubber boots back in place because that dislodges dust and guess where it goes! :) The mirrors could probably also use some cleaning but I'm not inclined to tackle those just yet.

    I have since done this same thing on a totally non-lasing SP-907. Even with only a foot or so between the alingment laser and the OC, it was enough to get flashes when jiggling the HR.

    Melles Griot GreNe with No Output

    This green HeNe laser tube came from a self-contained rectangular Melles Griot laser, "GreNe" model 05-SGR-871, about 24 inches long with an internal brick power supply (which appears to work fine). The tube is interesting in that it has a frit (hard) seal at the cathode-end but an Epoxy (soft) seal at the anode-end. This was probably done to reduce thermal stress (in the frit oven) on the very delicate OC mirror. In fact, I am in contact with the person who may actually have worked on the design or manufacturing of this laser at Melles Griot. :)

    Normally with a green (or other "other-color") HeNe laser tube having a discharge color/gas fill problem, there is little hope of recovery. The gain is so low that even trace contamination results in no output at all. However, for some reason, I got the feeling that this one was close enough to warrant some effort.

    Regardless of treatment options, the tube had to be removed from the chassis. This required unscrewing two aluminum mounting blocks, unscrewed some nylon set-screws, and pealing away at the black RTV Silicone holding the tube in place. This accomplished, Mr. GreNe was moved to my diagnostic facility (e.g., my adjustable HeNe laser power supply, tapped ballast resistor, and current meter).

    Initially, the tube operating voltage was about 20 percent low and variable - getting even lower as the tube warmed up. The color was obviously wrong but I suspected that there was still some hope. It was very pink but not deep red or blue.

    So, the first treatment procedure was to run the tube for awhile to see if that alone would result in at least some recovery. And, each time the tube was powered-on, the discharge color showed some definite improvement, though after running for a few minutes, it would tend to return to its former condition.

    However, after a total of about 8 hours of a 6.5 mA IV drip over several days, a few green photons started appearing for about 30 seconds shortly after powering up. During that time, I gently pressed on the cathode-end mirror to determine if alignment could be improved. It seemed fairly decent though I would tweak it later. Successive power cycles (with a cool-down period) appeared to result in somewhat more green output and for a longer time.

    I then applied several radiation treatments to the getter from my solar heater. I just set up the tube so a part of the getter ring was at the focus of the solar heater (about 1/4" focal spot from a 7"x10" or so Fresnel lens) and let it bake for a few minutes. Probably a total of 1/2 hour in a half dozen sessions of that around noon on a cloudless day, powering up in between to check condition. :) After a few of those, there is no further improvement. That is the basically the same thing I did to a contaminated red HeNe tube over a year ago (see the section: Repairing the Northern Lights Tube) but that was hard-sealed (it is still doing fine).

    I tweaked the alignment of the HR (cathode-end) mirror using the three-screw adjuster that was already there. That increased the output by about 10 percent. The adjustments were not super critical (as would be the case with a tube having marginal gain) and were repeatable. The beam is TEM00 and nice and circular.

    Following the solar treatments, there was a sustained green output between 0.4 mW (when first powered) dropping to about 0.32 mW steady state with very little power variation due to mode sweeping. The operating voltage has stabilized, probably close to its spec'd value, changing only very slightly during warmup. The discharge color now looks normal for a red HeNe tube, maybe a bit more saturated red than usual but it is stable and hasn't changed with additional getter treatments. (The color may be normal. If I recall correctly, the discharge color of my green 1-B HeNe laser tube looks similar.) The color in the expanded section of the bore near the anode is close to a normal orange. I suppose with the Epoxy seal, helium has likely leaked out in addition to air leaking in. Low helium pressure might explain both the discharge color (if it's really incorrect) and somewhat low output (a modern 05-LGR-170 tube is rated at 0.8 mW but see below). After running for a few more hours, the power has stabilized around 0.4 mW with little change during warmup. This probably means that additional benefits from doing anything with the getter will be negligible.

    However, I'm going to run the tube for a few more days. The power still appears to be climbing - very slowly but steadily. Though at this rate, it may be a few years before the tube achieves rated power. If that doesn't help after a few days, I will perform a helium soak. It should be a simple matter to enclose the anode-end only in a plastic bag filled with helium and even be able to power the tube to check progress. There is little risk of overfilling doing this for a couple weeks (the manufacturing date of the laser is 1988 and this is almost certainly the original tube) - 1 day for every year of age. For now, I have reinstalled the tube in the laser case using the set-screws but no RTV Silicone so it can be removed if needed. According to my contact at Melles Griot, it's possible that this laser had a minimum power spec of only 0.2 mW. Mr. GreNe is already doing twice that. :)

    Followup: After a year or so of occasionally turning the laser on for a few minutes to check that it still worked, I must have missed a couple months with the result that there was no green output at all. However, letting it run for a several hours restored it to nearly the same health, without needing any getter treatments. So, indeed the recommendation to run a soft-seal HeNe laser tube periodically is confirmed!

    Melles Griot Long GreNes with No Output

    These are rather long green (543.5 nm) HeNe laser tubes, possibly Melles Griot 05-LGR-191 or -193 (or their predecessor). At first I thought they were possibly designed to be multimode since the bores appear quite wide compared to even other TEM00 red tubes of similar length and the operating voltage is also rather low. I'd expect the rated output power to be several mW, possibly as high as 5 mW if this is the case. Should they turn out to be 05-LGR-193 tubes, the rated power would be 2 mW minimum. This is about the highest power of any currently manufactured green HeNe laser. However, I've been told that older tubes that look similar to these might only have a rated output power of a few tenths of a mW. And, these are probably rather old.

    The next test was to check mirror alignment. Using the "Instalign" procedure described in the section: Sam's Instalign(tm) Procedure for Internal Mirror Tube Mirror Alignment revealed that someone must have attempted to break the legs of these tubes. The cathode-end mirrors were so far out of alignment that the pointing error could be seen with the naked eye and the reflected spot of the alignment laser wobbled by several degrees as the tube was rotated. This is far beyond what the locking collars could correct, so they were removed and a steel plate that fit in the restricted region of the the mirror mount was used to gradually restore the mounts to something approaching correct alignment - where there was no detectable wobble in the reflected beam. This is somewhat hard to determine due to the multiple reflections but the smallest spot is from the outer planar surface and this was used as an initial guide. Note that since there might be some wedge in the mirror, this technique alone may not be sufficient to achieve close enough alignment. The same was done with the anode-end mirror, but it appeared to be much closer to proper alignment, possibly because the butcher, err, surgeon who had these tubes previously didn't want to mess with the high voltage!

    The patients were retested but still found to be lacking in any green output.

    The next step was to use the alignment laser to shoot a beam down the length of the bore. Fortunately, these particular tubes have a very wide bore for their size and doing this was not difficult. To actually optimize alignment, the reflection all the way back from the far mirror was used - just visible as a tiny dot of light when centered. There are actually multiple reflections but gently rocking the far mirror while observing the reflected pattern revealed when the far mirror was fairly well aligned. The tube was then turned 180 degrees and the same thing repeated.

    The patients were again retested but still found to be lacking in any green output. Rocking either mirror didn't have any effect. So, one of the patients (designated Patient #1) was selected for extended 6.5 mA therapy and put on the power supply for several hours.

    Finally, pressing on one of the mirror mounts resulted in a flash of green light! Some quick work with the steel plate, and then with the locking collars and it could be somewhat sustained, though still very weak. And, almost *everything* affected the output power. Usually, the beam was TEM00, but both TEM01 and TEM11 modes were observed at times. Mirror walking succeeded in improving the situation somewhat, but not dramatically. Adding magnets also increased the output power, though probably not by enough to justify the effort required to install them permanently. The output power appears to peak with a current between 6.0 and 6.5 mA. Patient #1 was allowed to run for several more hours. While no dramatic improvement has taken place, the output is much more consistent and 50 to 100 microwatts of green photons can be maintained indefinitely.

    The other tube, Patient #2, was retweaked for alignment several times before it finally started lasing, but with somewhat better results. After mirror walking, up to 0.4 mW of output power could be maintained consistently.

    While the output of even the healthier of these tubes is likely below its spec'd value, getting *any* green tube to lase can be quite a challenge. Not being mounted in any sort of thermally controlled enclosure (like a cylindrical laser head) doesn't help the situation as any low gain HeNe laser tube will be subject to significant power fluctuations if left in the open. Although I've seen multimode output from these tubes when the mirror alignment wasn't optimal, they are probably in fact designed to be TEM00. The beam diameter is approximately 1 mm but it's difficult to measure precisely to the 1/e points and it could be slightly smaller and consistent with the 05-LGR-191 or -193. These tubes are now probably fairly stable, if tired, but still good enough for a cool demo.

    Melles Griot Yellow Laser Head With No Output

    This laser head actually has a Coherent label, but is obviously made my Melles Griot since it is physically identical in every way, shape, and form, to an 05-LYR-173. It was purchased on eBay by a friend of mine. The eBay listing said it was working, even having a stated output power of 4.2 mW. But the laser head was received in its present dead state. A replacement was sent to him which worked fine and the seller didn't want this one back. So, I was asked if I wanted it or should he throw it away! Throw it away! Geez! :) What a question. It arrived packed in the original foam cushioned Coherent box. You could drop that out of an airplane and the laser head wouldn't even feel it. But the power supply brick was also in there when shipped from seller to buyer, so maybe that was bouncing against the head during the entire trip!

    Since the operating voltage and current behavior are normal, that only leaves a few possible causes. It could be a high mileage head which is now simply not lasing. Or, the mirror alignment could somehow have been knocked out in shipping, despite the padding. Possibly the bore shifted position slightly which would result in symptoms similar to misaligned mirrors.

    So, after removing the front end-cap, modest finger pressure was applied to the mirror mount. And, presto - a flash of yellow! Some quick work with my custom HeNe laser mirror adjusting tool and what's this? 5 mW? Is that accurate? How can that be? A bit more effort and allowing for complete warmup and, can you believe: 5.7 mW at 6.5 mA and more than 5.9 mW at 7 mA. This has got to be the liveliest yellow HeNe laser head for its size I've ever seen. The beam is identical in every respect to that of the yellow laser described in the next section - single transverse mode TEM00, and single line (594.1 nm only, confirmed with diffraction grating as well as somewhat calibrated monochronometer). What's more, its stability in terms of power fluctuations with warmup is better than many similar size red HeNe lasers climbing smoothly after starting out at about 4 mW when cold showing less than 2 percent p-p mode sweep variation in between.

    The sensitivity of output power to finger pressure on the mirror is higher than that other physically identical (but not as lively) yellow laser head. But once adjusted, it appears to stay put and has remained unchanged for several months. Perhaps the curvature of the mirrors is different and/or they are of higher quality.

    Even someone who had worked for many years in the HeNe laser division of Melles Griot and built "other color" HeNe lasers couldn't believe it, insisting the laser must be multimode or multiline or something else, arguing that the gain at 594.1 nm isn't that much different than at 543.5 nm (green) and green laser heads of similar size don't generally exceed 2 mW. So output power nearly triple that is at least somewhat unusual. I've heard of 7 or 10 mW yellow lasers, but they were typically much longer. However, since the CDRH sticker rating is 10 mW, I guess that sort of power was expected. Otherwise, it would have been listed as only 5 mW. I asked if Melles Griot was holding out the good stuff for Coherent since the highest power yellow HeNe laser they sell - the 05-LYR-173 - is only rated at 2 mW. :)

    However, it turns out that exact laser used to be listed on the Coherent Web site. And from the listed specifications, it was clear that the laser head is identical to the Melles Griot 05-LYR-173 and therefore must be manufactured by Melles Griot. (Coherent no longer sells any HeNe lasers.)

    Awhile later I did check the operating voltage of this laser head and it is definitely lower then the spec'd value of the 05-LYR-173. In fact, it's even lower than the operating voltage of the 05-LYR-171 which is rated at only 1 mW. For the genuine 05-LYR-173, the operating voltage is listed as 2,590 V. However, for this Coherent yellow head, I measured about 2,500 V. That 90 V may not sound like much of a difference but it's unusual to measure a lower operating voltage than the spec'd value at rated current. (A similar measurement of a low mileage 05-LYR-171 matched its specs exactly at 2,520 V. It's also possible the 2,590 number is wrong as the tube in the 05-LYR-171 and 05-LYR-173 appear to be physically identical but possibly the latter has a 78K ballast instead of 68K ballast since that is what's recommended for the bare tube.) In any case, even if the difference in voltage is only 20 V, perhaps the Coherent tube uses a slightly modified recipe that pushes the envelope on TEM00 performance, possibly at the expense of mirror adjustment stability. My testing wasn't in depth enough to determine if the beam diameter and divergence are eaactly the same for both lasers. However, from the fact that the model is only listed as 2 mW, I rather doubt there is any difference. Possibly it was simply from a batch of really lively laser tubes!

    The cause for the initially dead state is also a unknown. The mounting orientation of the laser head has little effect on output power and some gentle tapping evokes no response. This is a relatively new laser (manufacturing date of 2000) so it should have the molded-in-glass bore supports which can't deform like the older thin metal spiders. Thus, it's not clear how the bore could move from any physical shock. The mirrors are the normal metal tube extensions with a narrowed section for adjustment, with no locking collars. There is no way a mirror could just decide to move on its own from any amount of physical abuse that wouldn't totally destroy the tube. I was told that the power supply brick was included in the same box without padding and might have whacked the head during shipping, but that sounds unlikely as a cause.

    So, this too will remain a pleasant mystery for now.

    Melles Griot Yellow Laser Head With Variable Output

    The output power varies smoothly in what appears to be close to a sinusoidal manner with a period many times longer than the normal (much lower amplitude) mode sweep variation. It is very obviously related to thermal expansion since the period gets longer and longer during warmup.

    There is no obvious cause. With the front end-cap removed, pressing gently on mirror shows that it is well aligned whether the power is min or max or anywhere in between. Applying magnet therapy to check for the presence of the competing 3.39 um IR line has absolutely no effect.

    This laser head was surplus and came from a source that made it available as a laser known to not meet listed specifications but considered suitable for some purposes and therefore not "recycled". The output power at all times is still well above the rated minimum for the 05-LYR-171 of 1.0 mW but the variation in output power is four times the acceptable mode sweep specification of 10 percent.

    It appears to be brand new, a factory reject due to the power variation problem. The magnet test pretty much rules out 3.39 um competition, though the longer period of the variation is more consistent with this cause than anything else considered up to this point. The mirrors appear to be well aligned and stable. Low gain would result in a variation at the mode sweep rate which is much higher than what was observed. The mode sweep variation of a few percent is superimposed on the much larger long period variation and appears normal.

    Tests that were performed include the following:

    Based on all these expensive tests (only partially covered by HeNe laser health insurance), the diagnosis is that the improperly made HR mirror is resulting in the power balance shifting between the front and rear. Until this set of tests were performed, IR (3.39 um) mode competition was at the top of the list, but it has virtually been ruled out by the mode cycling and power variation period ratios, and the fact that magnets have little effect on output power.

    One final test was performed to confirm the diagnosis: I touched a cotton swab soaked with alcohol to the HR mirror about the time when the power of the main beam was near its maximum value. The power of the main beam instantly dropped by between 10 and 20 percent (from about 2.6 mW to 2.2 mW). The power recovered to near its previous level very quickly as the alcohol evaporated. Since the room temperature alcohol also reduced the temperature of the mirror substrate, that also played a part, but I didn't realize at the time to check more carefully. At the time, I also didn't fully realize that this result was totally consistent with the etalon explanation. More careful experiments could have been performed but I wasn't excited about the mess that might be created by using index matching fluid. :)

    The results of these tests would suggest a cure of sorts: A wedged optic could be attached to the outer surface of the HR mirror using index matching optical cement. The output power variations would then be if not eliminated, greatly reduced, with the output power of the main beam beaing about 2.4 mW and the output power of the waste beam being about 1.46 mW. The output power would be stable, but no where near the top end of the range. Even though this would greatly exceed the laser's specifications, I'm not sure whether this would be a preferred result, expecially since the laser head would no longer be very interesting. Another similar boring cure would be to add an additional HR behind the whimpy one, perfectly aligned so that all power gets reflected back into the cavity. I don't know whether this could be made stable but have tried some experiments. While the results with some specific mirrors were encouraging, these will probably remain as interesting experiments. As I found out, due to the low gain of yellow compared to red, the mirror to be used must be very selective in order to prevent any red lasing (mostly at 632.8 nm). For details, see the next section. With the most effective mirror, the output from the front varied from above 3.5 to 4.3 mW of pure yellow (594.1) with a much reduced waste beam of under 0.3 mW. A suitable flat mirror glued to the existing HR might work as a permanent stable solution since that surface is already (too well) aligned. But again, the result would be a boring (but rather lively) laser. :)

    Unless the patient decides on one of these options, no further treatment is recommended. The chronic problem is not likely to get any worse or any better but will continue to be monitored. I have modified the rear end-cap to allow the second beam to exit without requiring repeat surgery each time tests are to be performed. As I found out, a vital organ (ballast resistor) was in the way when simply drilling a hole through the end-cap's center. A transplant was successful with the new ballast resistor clipped directly to the HR mirror mount.

    Awhile later, I pulled the rear end-cap off of a healthy 05-LYR-173, a laser head virtually or totally identical to the 05-LYR-171, except for a power rating of 2.0 mW instead of 1.0 mW. They are probably the same laser head, but selected based on power before printing the label. :) Sure enough, the waste beam from the 05-LYR-173 was only about 35 microwatts (uW) and changed little during warmup even though the output power from this particular laser head was about the same (4.1 mW) as the total power (front and rear) from the sick 05-LYR-171. In fact, the waste beam power may have even decreased slightly as a percentage of total power as the laser warmed up. And, the kicker is that there were multiple very obvious ghost beams due to wedge clearly visible even with only 35 uW in the waste beam!!! Case closed. :-) :-)

    The next section includes plots of the sick yellow laser head as well as a normal one. More on all this including additional discussion and plots of normal red laser heads can be found in the section: Power Variations Due to Lack of Mirror Substrate Wedge.

    So in conclusion, if only the coating on the HR mirror had been incorrect with low reflectivity at 594.1 nm, the output power of the main beam would be reduced and the output power of the waste beam would be increased, but they would both be stable, and the laser head would still exceed listed specifications (1 mW minimum output power). If the HR optic had no wedge but its reflectivity was correct, there would probably only be minimal output power variation because the reflected waste beam would be much lower in power. The laser head would also likely exceed listed specifications. However, this "perfect storm" of manufacturing defects resulted in the variable output power behavior and a fascinating study showing that even small back-reflections can have a dramatic effect on HeNe laser power stability.

    It's gratifying to have been able to deduce what was going on with this laser based on simple measurements. Although, there might be a record of this problem in a file somewhere at Melles Griot, it's also possible that the head was simply put in a bin labeled "reject, variable power" and left at that. It must not be all that common as my contact at Melles Griot who had worked as an engineer in the HeNe laser division for many years wasn't able to nail it.

    Output Power Plots of Yellow HeNe Laser Heads

    Here are some plots of the power output of this sick laser before attempted treatment, as well as a healthy similar model laser for comparison. The sophisticated diagnostic instrumentation system consists of a pair of $2 photodiodes feeding two of the analog inputs of a DATAQ Chart Recorder Starter Kit attached to my ancient 486DX-75 Kiwi laptop running Win95. Each photodiode is reverse biased by 30 VDC from a +/-15 VDC power supply with a variable load resistor to set the calibration. The output is taken between the junction of the resistor and the photodiode, and power supply common (0 VDC). The resistor can be set so the maximum optical power will be near +10 V to make good use of the +/-10 V range of the A/D converters.

                   R1     PD1
     +15 VDC o----/\/\----|<|----+
                  100            |
                                 \<----------+----+---o A/D Input (+/-10 V range)
                                 / R2        |    |
                                 \ 25K       |    /
                    R3           |       C1 _|_   \ 200K ohms (Zin of A/D module)
     -15 VDC o-----/\/\----------+     1 uF ---   /
                   68K                       |    \
                                             |    |
       0 VDC o-------------------------------+----+----o A/D Ground

    Since the input impedance of the A/D is not infinite but about 200K ohms, and the negative power supply is -15 VDC, the voltage for zero optical power isn't necessarily -10 V but may be somewhat lower or higher with the calibration for this laser. (It was quite close though in most cases.) This doesn't matter for the purpose of the plots since the DATAQ display software allows for offset and gain adjustments but would have to be factored in if precise power measurements were important. A simple op-amp buffer stage could easily be added to provide proper gain and offset adjustment. The calibration factors for all plots except the two polarized ones has been adjusted so they would look about the same relative to actual output power, about 4.5 mW full scale.

    The capacitor across the input is intended to minimize noise pickup. The resulting filter rolls off at around 0.6 Hz. For reasonably well behaved HeNe lasers, even during the initial warmup period, this bandwidth is more than adequate. The sampling rate for all the plots is at least 10 Hz to allow for averaging since the A/D seems to have an uncertainty of about 2 LSBs. Even with lasers where external mirrors have been added (described below) that are mounted with inadequate vibration dampening, nothing fundamental will be missed though some of the more rapid fluctuations may be reduced a bit in amplitude. This would be most noticeable near the start of the warmup period. Where greatest accuracy was considered critical and more rapid fluctuations were expected (as with the external mirror therapy), the tests were done without any filtering.

    Sick Melles Griot 05-LYR-171 yellow HeNe laser head:

    Plot of Variable Melles Griot 05-LYR-171 HeNe Laser Head During Warmup shows the power variation over the course of 1/2 hour or so from a cold start. The total (random polarized) output power of each beam is being monitored. Plot of Variable Melles Griot 05-LYR-171 HeNe Laser Head Near End of Warmup shows only the last cycle of the warmup period expanded with the detailed small power fluctuations due to normal mode cycling more visible. The outputs from both ends of the laser head are superimposed with equal calibration. Note how the slow power variation from the front and rear are out of phase but the much more rapid (normal) variations due to mode cycling are in phase. The peak output power for the main beam (blue) and waste beam (red) are around 2.8 mW and 1.8 mW respectively, near the end of the run. I didn't bother waiting, but from the increasing period of the variable power cycles, it's possible that only 1 or 2 additional complete cycles would take place before the outputs stabilized based on the balance of heat generation and heat loss. There is a reason that these plots have not been expanded to fill the available height. This will be obvious later. :)

    As can be seen, for this laser, the power output from the main beam when it is minimum and the power output from the waste beam when it is maximum are almost the same, so this also means the that the transmission factors (1-R) are then similar. Based roughly on the data, we have:

                          Power in     Power in    Relative T  Relative T
            Condition     Main Beam   Waste Beam     of HR       of OC
            Min (phi=90°)  2.78 mW      1.28 mW       0.46        1.00
            No Etalon      2.39 mW      1.46 mW       0.61        1.00
            Max (phi=0°)   1.88 mW      1.75 mW       0.92        1.00

    The case of no etalon (wedged or perfectly AR-coated HR) was estimated based on the equation and the data. Note that for this low gain laser, the total power (main and waste) is not constant, so I used the average.

    The slight shift in power balance as the laser warms up between the main and waste beams when the main beam output power is a minimum may be attributed to the increasing temperature of the HR mirror. (I have virtually eliminated the obvious possible cause - differing gain and offset for the two A/D input networks.) The temperature rise results in expansion of the dielectric stack causing a slight wavelength shift in its reflectivity peak, and thus a change in the reflectivity at 594.1 nm. Since the HR mirror is near the anode ballast resistor, it may see a temperature rise of 50 to 75 °C. To shift the power balance by the observed 5 to 10 percent would not take much of a change in the HR's reflectivity. With the peak reflectivity already off to one side, it's on the slope of the reflectivity function so another 0.1 nm or even less of a shift might be enough. Assuming the temperature coefficient of a typical dielectric stack at 594.1 nm to be one half that of the mirror substrate, for the temperature rise of 70 °C used in the calculation above, the wavelength shift would actually be about 0.2 nm. But I picked the "one half" arbitrarily. One Web site showed a coefficient over 5 times greater but that was for a dielectric filter, not a mirror, and I don't know if the same materials would be involved and am trying to find some hard numbers on this. But even 0.2 nm could indeed be enough.

    The next plots show what happens if a polarizing filter is placed in each of the beams so that only one set of the (usually) orthogonal polarized modes contribute to the output power. Plot of Variable Melles Griot 05-LYR-171 HeNe Laser Head During Warmup (Polarized) shows the entire warmup period and Plot of Variable Melles Griot 05-LYR-171 HeNe Laser Head Near End of Warmup (Polarized) is a closeup of only the last full cycle. (The scale factor for these two plots is arbitrary - the two orthogoanl polarized components must add up to the total power in the first set of plots.) As above, the front and rear beams have opposite phase for the slow varying power fluctuation and the same phase for the much more rapid mode cycling. But note the wild fluctuations that are often not at all similar from cycle to cycle in output power on a short time scale, not even close to normal for a well behaved HeNe laser. It's not clear exactly what this means but the gyrations probably include polarization flips in addition to simply modes falling off one end of the neon gain curve and appearing at the other end. The back reflections from the HR mirror's outer surface are probably the primary cause of this instability.

    Healthy Melles Griot 05-LYR-173 yellow HeNe laser head:

    As a comparison with a similar model laser that behaves in a normal manner, see Plot of Melles Griot 05-LYR-173 HeNe Laser Head During Warmup, Plot of Melles Griot 05-LYR-173 HeNe Laser Head Nar End of Warmup, Plot of Melles Griot 05-LYR-173 HeNe Laser Head During Warmup (Polarized), and Plot of Melles Griot 05-LYR-173 HeNe Laser Head Near End of Warmup (Polarized). The 05-LYR-171 and 05-LYR-173 are virtually, if not totally physically identical, but have output power ratings of 1.5 and 2 mW, respectively. The model number is probably assigned after manufacturing based on performance. These plots were made using the same setup. The output power of this laser is slightly over 4 mW after warmup. Note the complete absence of any slow periodic variation in output power - the trend is almost perfectly monotonic indicating good mirror alignment and little or no wavelength shift for the mirror coatings. And, the polarized plot is relatively well behaved as well with none of the rapid sudden wild random gyrations present in the variable output power laser. The closeup shows that while the mode cycling power variation exhibits a complex waveform, it repeats almost exactly from cycle-to-cycle. For more on how normal lasers behave, see the section: HeNe Laser Output Power Fluctuation During Warmup.

    External Mirror Therapy for Variable 05-LYR-171 yellow Laser Head

    After completing the battery of extensive diagnostic tests, the patient volunteered for some low risk experimental therapy to reduce the power variations. In an attempt to get around the problem of the improperly manufactured HR mirror, an external mirror would be added behind it. The rational being that by reflecting back the wasted 594.1 nm photons into the laser cavity, there would be the hope of increasing the power out of the front and and reducing the power variations. For this purpose, an adjustable mirror mount was attached to the highly stable 1"x4" wooden plank/V-block setup used for testing HeNe lasers, just beyond the HR-end of the laser head. Some of the results were unexpected, though predictable in hindsight. The following were tried in the order in which they appear:

    The circulating optical power between the internal HR and external mirror was quite impressive with most of the mirrors, possibly as much as 0.2 WATT or more for the Ultimate Broadband. But it was mostly red! If the external surface of the 05-LYR-171's HR mirror were cleaned more carefully, the power might go even higher but the laser would likely be even more unstable. Nope, not going to do that.

    Since I have filter capacitors resulting in a time constant of about 0.2 seconds on the two inputs to the A/Ds, a low pass filter is formed with a 3 dB cutoff of about 5 Hz. Higher frequency events will be reduced in amplitude and might not show up on the plots. However, I've examined the raw data (the full 60 samples per second) and it appears that 5 Hz is an adequate bandwidth to capture whatever is going on because there is little activity anywhere near this frequency. However, I will probably do a test run with the capacitors removed to be sure.

    Note that since the external mirror mount and laser head are not rigidly attached to each-other, some creep between them may contribute to the complex behavior. No, I don't intend to perform any in-depth analyses of any of this either! However, this does offer a method of pushing the power variations to higher frequencies. Vibrating the mirror with a loudspeaker in close proximity or a piezo transducer (e.g., a piezo beeper from a dead digital watch!) would result in an output power that was the average of the long term trend. The reason is that the short term fluctuations are caused by um-sized changes in the distance between the external mirror and the HR mirror of the laser tube. I did exactly this experiment by sitting a loudspeaker next to the external mirror driven from a function generator at around 700 Hz, which must have been close to some resonance because much less audio power was needed to produce the desired effect. For that same 30 cm RoC red HR, when the output power was averaged over a time span even as short as one sample period at 60 samples/second, most of the rapid fluctuations disappeared. Of course, they are still there but at a much higher frequency. But for some applications, that would be acceptable.

    It's still possible that by using a suitably selective yellow HR mirror or one which kills the red wavelengths more effectively, that the laser could be even better tamed. In the meantime, the patient has opted to learn to live with this peculiar malady (without permanent external mirror therapy). While the benefits of using the green OC are quite impressive, the cost is not yet covered by HeNe laser health insurance plans. However, that option still exists and will be tested if a promising mirror comes along.

    Assorted Red Laser Heads With Variable Waste Beam Power

    For 10 or 15 years, barcode scanners used mass-produced red HeNe laser tubes often designed for least cost. Wedged optics are likely more expensive to manufacture, not because of the grinding and polishing, but because the change in thickness limits the size of the raw glass blanks that can be used. So, apparently, many if not most of the HeNe laser tubes destined for barcode scanners lack wedge in their HR mirrors.

    As a followup to the yellow tube studies, above, it was decided to check out some common red tubes which exhibit similar waste beam behavior. With proper high reflectance coatings on the HR mirror, there will be no dramatic power variations in the main beam, but the lack of wedge will affect the waste beam power with some small corresponding inverse variation in main beam power..

    Although these tubes are in generally excellent health, in the interests of science, a battery of tests were performed on each to characterize their rear-end waste beam variability. The results are summarized below:

    Both show varying degrees of waste beam variability. The cause of differences in amplitude is not known but may be due to very slight amounts of less than perfect flatness or wedge of the outer surface of the HR mirror that are present, though not added intentionally.

    And a couple of control subjects with wedged HRs:

    A subsequent study of several 05-LHR-006, all with the 50-03400-014B part numbers (including tho two subjects above) show quite a range of variation:

                       Main Power     Waste Power
        ID #     Plot   Average      Min (Avg) Max     Ratio   Comments
       161079            1.5 mW    19.2 uW   41.5 uW   2.16    No Wedge
       145573            1.5 mW    15.7 uW   31.1 uW   1.98     "    "
       117048  006-1m1   1.5 mW    15.0 uW   27.0 uW   1.80     "    "
       116076  006-2m1   1.4 mW    29.0 uW   44.0 uW   1.52     "    "
       164091            1.4 mW    16.0 uW   20.3 uW   1.27    Tiny Wedge
       156144            1.4 mW         (18 uW)        ----    Small Wedge 
       257644            1.5 mW         (30 uW)        ----    Normal Wedge
       374746            1.0 mW         (17 uW)        ----    Large Wedge

    These measurements were made using the quick low cost approach using a blow dryer to heat the HR mirror. There was no noticeable ghost beam for the first 4 subjects when the waste beam was projected on a white card at any distance from the HR. The next subject had just detectable fuzz off to ne side a few feet from the HR, with visible ghost beams at increasing angles for the final 3 subjects. With any sort of decent wedge, the variation of waste beam power due to HR subtrate etalon effects - if it exists at all - is swamped by the normal mode sweep and thus cannot be determined reliably. So, only the average waste beam power is given for them. The more extensive time consuming and expensive total warmup test will be required to obtain min, max, and ratio values.

    However, the residual waste beam power variation for tubes with HR wedge (including the last Siemens LGR-7641 run) are likely due to etalon effects in the OC mirror! A blob of 5 minute Epoxy on the 05-LHR-006 tube's HR made essentially no change in the amplitude of the ripples. Had it been reflections from the wedged surface, they should have gotten smaller. See the section: Power Variations Due to Lack of Mirror Substrate Wedge.

    Conclusions: The only obvious manifestation of the rather dramatic rear-end variability are some slight ripples in the main beam corresponding to the power stolen by the waste beam. For these red tubes, it's a very small percentage and well within their acceptable specifications. No treatment is required unless any of these tubes are to be used in an application where power stability of the waste beam is critical. But such adventures would be strongly discouraged since a perfect remedy is likely impossible, or at least very expensive and time consuming, and definitely not covered by any laser tube health insurance plan.

    Case Study of 145 Melles Griot 05-LHR-640 HeNe Laser Tubes

    This work was underwritten by the International Society of HeNe Health (ISHH). :)

    These 145 Melles Griot HeNe laser tubes were all volunteers. The standard 05-LHR-640 has a rated output power of 0.5 mW and are very small - about 5 inches in length by 1 inch in diameter. They look like normal Melles Griot tubes with the glass bell at the anode-end, not the barcode variety with the metal end-cap (though Melles Griot still calls them barcode scanning tubes). And the beam exits from the cathode-end like most normal tubes. This batch are all OEM tubes, from some sort of scanner based measurement system.

    Six of the 145 tubes didn't survive travel, another one had a gas-fill problem, and the records for two more were lost somehow and didn't make it into my statistics. :) So, there are only 136 entries for working tubes.

    Functional testing was done using a Melles Griot 05-LPM-379 power supply set at 4.5 mA using my HeNe laser diagnostic unit. (See the section: Ballast Resistor Selector and Meter Box.) The ballast resistance was set at 54K. (For some reason, possibly due to the long wire lead to the tube anode, 54K was more stable for tubes with a higher operating voltage than the next choice of 81K.) Both tube voltage and tube current were monitored continuously. Output power was also measured continuously using a laser power meter with the listed value being approximately 0.05 mW lower then the peak (from mode cycling) after 5 to 10 minutes with the precision only maintained to the nearest 0.05 mW. If anything, some of these output power values may be lower than after complete warmup since the output power had not necessarily stopped increasing in the time allotted by the ISHH for each test. During this time, mirror alignment was checked by gentle sideways pressure on the cathode-end mirror mount and if a noticeable increase was detected, the mirror alignment was optimized for peak power. About half the tubes benefited from this treatment (free of charge), many quite significantly.

    Interestingly, the output power when new for these (and other very small) tubes may be nearly 3 times the 0.5 mW spec'd value! Even the sickest of the tubes that survived would still meet output power specifications.

    The chart below is sorted by tube operating voltage. The spec'd voltage for the 05-LHR-640 is 880 V but none of the tested tubes came anywhere close, with even the best of them that looked and acted brand new were almost 40 V higher. While it's possible my measurements are off somewhat, I don't think it is more than a few volts either way since I've double and triple checked the calibration. I have also tested some Melles Griot laser heads that I know have seen little use and the voltage for those was within 1 percent of the listed values. So, possibly the voltage specs I have for the 05-LHR-640 are incorrect, or even with modest use, the voltage climbs considerably. However, these are actually OEM laser tubes, manufactured for a specific application. So, perhaps the operating voltage spec for them is slightly different than the standard tube.

    Even though the voltage readings in comparison to 880 V may not be significant, the relative voltages should be consistent. As can be seen, the highest output power on average is not surprisingly associated with the lowest tube voltage. As the voltage increases - which is how these tubes change with use - the output power on average declines somewhat, but not dramatically. Also on average, increasing amounts of the dreaded brown crud appear in the bore for tubes exhibiting a higher operating voltage, with zebra stripe striations indicating plasma oscillation for many of them.

    Group 1:
     ID#   Vop   Pout  | ID#   Vop   Pout  | ID#   Vop   Pout  | ID#   Vop   Pout
       1   918   1.10  |   2   924   1.20  |   3   925   1.30  |   4   925   1.40
       5   926   1.30  |   6   928   1.10  |   7   929   1.25  |   8   929   1.25
       9   931   1.00  |  10   931   1.15  |  11   932   1.10  |  12   932   1.10
      13   935   1.20  |  14   935   1.30  |  15   935   1.35  |  16   936   1.00
      17   937   1.10  |  18   938   1.20  |  19   938   1.20  |  20   939   1.20
      21   939   1.20  |  22   940   1.00  |  23   941   1.20  |  24   942    .95
      25   942   1.20  |  26   942   1.35  |  27   943    .95  |  28   944   1.10
      29   946    .90  |  30   946   1.25  |  31   946   1.25  |  32   948    .95
      33   950   1.10  |  34   954    .95  |  35   954    .95  |  36   954   1.00
      37   956   1.00  |  38   957   1.00  |  39   958   1.05  |  40   959    .90
    Group 2:
     ID#   Vop   Pout  | ID#   Vop   Pout  | ID#   Vop   Pout  | ID#   Vop   Pout
      41   959    .90  |  42   960    .85  |  43   960    .90  |  44   960    .95
      45   961   1.00  |  46   962    .95  |  47   964    .95  |  48   965    .80
      49   965   1.00  |  50   967    .60  |  51   967    .90  |  52   968    .85
      53   968    .90  |  54   969    .85  |  55   969    .85  |  56   969    .85
      57   969    .90  |  58   969    .95  |  59   969   1.00  |  60   970    .70
      61   970    .80  |  62   970   1.00  |  63   971    .50  |  64   971    .70
      65   971    .85  |  66   971    .90  |  67   971    .95  |  68   971   1.00
      69   972    .90  |  70   973    .75  |  71   974    .85  |  72   975    .70 
      73   975    .80  |  74   976    .70  |  75   976    .75  |  76   976    .75
      77   976    .85  |  78   977    .60  |  79   977    .90  |  80   977    .90
      81   977    .90  |  82   978    .80  |  83   978    .80  |  84   978   1.15
      85   979    .80  |  86   979    .85  |  87   981    .85  |  88   983    .75
      89   984    .75  |  90   984    .80  |  91   984    .80  |  92   985    .75
      93   985    .90  |  94   986    .65  |  95   986    .75  |  96   986    .85
      97   986    .90  |  98   987    .75  |  99   987    .90  | 100   988    .75
     101   988    .80  | 102   989    .65  | 103   989    .65  | 104   989    .70
     105   989    .90  | 106   990    .90  | 107   991    .85  |  
    Group 3:
     ID#   Vop   Pout  | ID#   Vop   Pout  | ID#   Vop   Pout  | ID#   Vop   Pout
     108   992    .70  | 109   992    .70  | 110   992    .75  | 111   993    .60
     112   993    .65  | 113   993    .65  | 114   993    .70  | 115   993    .80
     116   993    .85  | 117   994    .55  | 118   995*   .70  | 119   996    .55
     120   997    .60  | 121   997    .65  | 122   997    .75  | 123   999    .55 
     124   999    .80  | 125  1001    .75  | 126  1003    .70  | 127  1004*   .65
     128  1006    .75  | 129  1007    .50  | 130  1007    .65  | 131  1007    .70
     132  1009    .60  | 133  1009    .85  | 134  1010    .65  | 135  1011    .55
     136  1011*   .60  |

    The chart is divided into groups in ascending operating voltage, which likely translates roughly into hours of use. Group 1 has tubes I consider to be in essentially new condition with low operating voltage and very little or no brown crud in the bore. As can be seen, on average, they also have the highest output power, over twice the 0.5 mW rating. Tubes in Group 2 are definitely quite healthy but have seen some use. Their operating voltage is somewhat higher and there is usually some brown crud in the bore. But their output power still averages nearly double the rated value. Most of the tubes in Group 3 have substantial brown crud in their bores but are still usable, even those marked with a "*", which were hard to keep running and would start flashing after a couple minutes using my test setup, which has a rather long run from the final ballast resistor and the tube anode. However, when operated with the ballast close to the anode, they were stable. The output power of virtually all of these are still well above 0.5 mW. Out of 137 intact tubes, only one didn't work at all (not listed) due to a bad gas fill. It wasn't end-of-life as the bore was in pristine condition and the operating voltage was quite low (851 V, dropping as the tube warmed up). The tube was either leaky or had not been processed properly when manufactured with the discharge color being excessively pink with reduced brightness.

    Here are the stats for all working tubes:

    Finally, I used MS Excel to generate a plot of these data and trends. See Melles Griot 05-LHR-640 HeNe laser Tube Health.

    Many of these tubes are available for low cost adoption. Please see: Sam's Classified Page.

    Some Hughes 3184H Semi-Antique HeNe Laser Heads

    The Hughes 3184H is a really old HeNe laser head which consists of a two-Brewster HeNe laser tube inside a gold-colored cylinder with mirrors in the semi=adjustable end-places as shown in Hughes Model 3184H HeNe Laser Head and Hughes Model 3184H HeNe Laser Head Construction.

    These had apparently been stored in a damp, if not absolutely wet basement for around 30 years. The test dates written on some of the laser heads are from 1973! As described in the section: The Ancient Hughes HeNe Laser Head, these laser heads actually consist of two-Brewster HeNe laser tube inside a massive gold-colored aluminum cylinder, with somewhat adjustable end-plates which contain the mirrors.

    I had tested two 3184H laser heads in the past that were actually newer than this batch (1976 and 1979) but neither lased. However, when these eighteen were originally tested by someone else, 4 were found to be in at least somewhat working condition. I was happy to trade a bunch of mostly non-laser electronics items for these laser heads, mainly for the 2-B tubes inside.

    I've now tested all 18 and found that 4 of them lased without doing anything and 3 others run at an operating voltage close to that of the ones that work - about 1,600 V across the red and green wires (at 6 mA) - but produced no beam. All of the rest started, but have an operating voltage at least 200 V higher and show the tell-tail reddish discharge color indicating air leakage. All of the bad ones have serial (ID) numbers below 1000 and are not listed. With a bit of mirror tweaking - quite a bit for most - and in some cases B-window cleaning, all the heads with the more or less correct operating voltage now lase, with those outputting at or above 3.00 mW being within measurement error of their power back in 1973. I believe that with careful cleaning and alignment, they would further improve.

    All are listed below:

             Power    Operating
       ID#   Output    Voltage   Comments
      <900   0.0 mW   >1,800 V   None of these 10 even come close to lasing.
      1014   2.6 mW    1,614 V
      1015   0.5 mW    1,585 V   Started out very weak, more below.
       ""    1.8 mW    1,593 V   Peaks, then declines.
      1038   3.5 mW    1,610 V
      1040   3.1 mW    1,629 V   Produces 4.1 mW on test rail, see next section.
      1042   2.5 mW    1,616 V
      1043   3.0 mW    1,614 V
      1050   3.2 mW    1,682 V
      1075   2.0 mW    1,602 V   Produces 3.8 mW on test rail, see next section.
      3506   0.0 mW    1,720 V   Newer (1976) has pink/red discharge color.
      4197   0.0 mW      --      Newest (1979) is nearly up to air.

    The low operating voltage of ID# 1015 suggests gas contamination. In support of this, besides the low output power, is the behavior of output power with warmup: first the output power increases to as much as 1.8 mW and then declines over the course of a few minutes. This is repeatable but the peak and minimum output power, as well as the operating voltage, has so far been steadily increasing so there may be hope yet. The first entry is the initial behavior after warmup, prior to cleaning and alignment. The second entry shows the peak power and highest operating voltage after cleaning and alignment, and a few power/warmup cycles. A portion of the power difference (but none of the voltage difference) is due to the cleaning and alignment but most is probably a result of gas cleanup. Hopefully, after awhile it will increase to a respectable output power and be stable there. Presently, it starts out somewhat low, peaks, then declines to about 1 mW. The operating voltage also declines to below 1,575 V when it is hot.

    The other laser head with a somewhat low operating voltage, ID# 1075, may also benefit from similar treatment, but probably not as dramatically.

    The second from lsat is one of the laser heads I acquired a year or two ago. It's from 1976 and lights up but won't lase. The discharge color is now pink-red thought I think it may have been more correct before (but it never lased).

    The last one is the other one I've had for awhile. It used to have a weak discharge color and incorrect operating voltage, but now only pulses weakly and won't start.


    All the Hughes 3184H laser heads that were not totally dead were asked to return for followup tests to determine how their condition has progressed after about six months. They are compared in the chart below.

           <------ Power Output ------>
      ID#  May 2005  Nov 2005  May 2006  Comments
     1014   2.6 mW    1.7 mW   0.24 mW   After running for ashile.
     1015   1.8 mW    0.9 mW   0.0  mW   Pink discharge.
     1038   3.5 mW    3.3 mW     NLA     Not optimized on return visit.
     1040   4.1 mW    4.1 mW   4.5  mW   On test rail.
     1042   2.5 mW    1.8 mW   1.34 mW   After running for ashile.
     1043   3.0 mW    3.1 mW     NLA
     1050   3.2 mW    3.1 mW     NLA     Not optimized on return visit.
     1075   3.8 mW    3.4 mW   4.0  mW   On test rail, unchanged

    (NLA: No Longer Available, moved to a different location.)

    It would appear the heads that produced near spec'd power originally are essentially unchanged after 6 months, but the ones that were marginal to begin with have declined significantly. Power cycling therapy is being performed free of charge on them but they probably won't come back completely. The slight decline of ID#s 1038, 1050, and 1075 is probably not statistically significant since the current was not optimized for maximum power and only limited time was allocated for each test.

    In particular, ID# 1015 would only produce momentary flashes when first powered up. Suspecting alignment or contamination (since it has been opened), the mirrors were removed and it was placed on test rail. It required some running time and a few power cycles to get back to even the 0.4 mW.

    Hughes 3184H HeNe Laser Plasma Tube Test Rail

    I have now constructed a setup which has a sturdy mount for the laser head, initially had a pair of adjustable Parker-Daedal 1 inch mounts (similar to a Newport MM-1 but not quite as precise) for the mirrors.

    This enables the health of the actual tube to be assessed as well as permitting the B-windows to be more conveniently cleaned. Version 1.0 is shown in Photo of Hughes Model 3184H HeNe Laser Plasma Tube Test Rail - V1.0 and Closeup of OC-End of Hughes 3184H Test Rail. Do you know what the two spots visible inside the cylinder are?

    To minimize dust collection on the Brewster windows, the laser head should be oriented either with the Hughes label on the side (which puts the B-windows vertical) or on top (which puts the B-window facing down). Even so, cleaning of the B-windows will be needed periodically to peak power. This can make a huge difference, especially if a clump of dust decides to settle in a bad place - and dust tends to be attracted to the high intracavity beam. As long as there is lasing, anything on the B-windows will be obvious.

    Note that the 3184H laser head is not symmetric. In fact, it appears that the bore is significantly wider at the OC-end compared to the HR-end to more closely match the mode volume of the hemispherical resonator (flat HR and 30 cm RoC OC, spaced just under 30 cm apart) and it's much wider overall than would be required for the beam diameter. As confirmation, the spot on the HR (in the original laser) is very small, almost a single point. Thus, interchanging the OC and HR without also changing their RoCs will result in mediocre performance. So, something along the lines of a hemispherical arrangement really is the one that should be used, though it's possible that a long radius hemispherical resonator might be even better.

    After realizing that a symmetric configuration with a pair of 60 cm mirrors was far from optimal (due to the tapered bore), I switched to using a flat HR and a 60 cm OC (from an SP-084, 99%R at 633 nm). These result in decent output power (see below) but produce a beautiful doughnut (non-TEM00) beam profile. The original hemispherical configuration would force a TEM00 beam even with the relatively wide (but tapered) bore. Eventually, I may try to optimize the mirror RoCs for a TEM00 beam, but that would be so boring. :) I can't use the OC mirror from a 3184H laser head because its 30 cm RoC is too short for the extended length cavity. A 40 cm RoC OC might be acceptable though.

    The easiest way to perform mirror alignment on this sort of setup is to *never* lose the alignment of *both* mirrors at the same time! Otherwise, an alignment laser will probably be needed, though I have been successful in restoring lasing on an intact 3184H laser head by careful trial and error, simply using the mechanical alignment of the end-plate with respect to the head cylinder as a guide. Though, maybe it was more a matter of luck.

    So, start with a 3184H that is lasing. Secure it in the mounts and adjust the external mirror beyond the OC so that all the reflected spots converge. Then remove the 3184H's OC. The alignment should be close enough so that when the end-plate is removed, if it isn't already lasing, slight rocking of the mirror in Y while very slowly adjusting the alignment in X will restore a lasing condition. Unfortunately, the same procedure can't be used with the HR because of its fine-ground outer surface. For that one (as well as the OC should its alignment be lost later on), I use a 60 cm RoC HR or OC mirror mounted in a mirror cell such as shown in Simple Mounting Cell for Salvaged HeNe Laser Tube Mirrors, or any other similar mounting scheme. This can simply be held against the end of the 3184H cylinder and easily adjusted by hand until the tube lases (assuming the other end is aligned). Then, the adjustable mirror can be tweaked until the spots converge as above. Once the 2-B tube has been tested, the Hughes end-plates can be reinstalled. The OC can use its outer surface reflection as a guide but the HR will simply have to be carefully adjusted until flashes occur. Buts as long as alignment of both ends is not lost, all of these procedures are really quite quick and easy.

    In preliminary tests, head ID# 1075 that was only doing 2 mW with its mirrors produces over 3.5 mW on this rig. I'm not sure how much of the improvement is due to being able to clean the Brewster windows very well and how much is due to the different mirror RoCs, more modern mirrors, or cleaner mirrors.

    I have since replaced the original mirror mounts with some Newport mounts (U50-A) to improve the precision and repeatability. While the Parker-Daedal mounts are perfectly acceptable for beam steering, their behavior inside a laser cavity left something to be desired. The U50-A has about 3 times the resolution with their 100 tpi screws, and less interaction between X and Y. I "machined" a pair of 1/2 inch to 8 mm adapter rings for the typical HeNe laser mirrors that would be used most often. For Version 2.0, I've also added some aluminum plates under the rail to add some stiffness, as even gently touching the mounts resulted in major fluctuations in output power. A sturdier resonator frame with a pair of my home-built mirror mounts would probably be even more stable but this was quick and easy. :) While probably unrelated to these changes, the output power of head ID# 1075 is now up to 3.8 mW.

    Of course, as with any external mirror laser, keeping the Brewster windows and mirrors clean is a challenge. They are oriented vertically in this setup and the mirrors stay fairly clean. But the Brewster windows tend to collect all sorts of stuff (technical term!) even after a short period of time. Only occasionally do I clean the mirrors, and then only using the drop-and-drag method with lens tissue and pure isopropyl alcohol. I usually just use new cotton swabs to gently dust off the Brewster windows, with the scatter as a guide to cleanliness. When the scatter from the outer surface is less than the scatter from the inner surface, that's the indication that they are about as clean as possible.

    Next, I installed head ID# 1040 in place of head ID# 1075 on the test rail. It was fairly close to optimal alignment after just dropping the tube in in place and tightening the clamps. Then, some mirror walking to peak output power. While originally doing 3.1 mW, it now produces up to about 4.1 mW. So, a bit livelier but probably within the normal manufacturing variation from head to head, and not due to a problem with head ID# 1075. I'd expect the remaining working heads to behave similarly.

    After this, I had to try a Coherent, Inc. mirror called an "Ultimate Broadband". This mirror does indeed appear to be highly reflective through the entire visible spectrum. Being 1/2 inch in diameter, it fit right into the Newport U50-A mirror mounts. Indeed, the super mirror performed even better than my original HeNe HR. The output power peaked at about 4.5 mW after initial alignment, though I couldn't get more than about 4.3 mW consistently (probably due to my lack of ability to keep the four exposed optical surfaces clean in my optics/laser lab, err, basement!). I don't really know how much of the power increase, if any, was due to higher reflectivity of the super mirror, and how much might have been due to its RoC of 1 meter, compared to flat for the original HR.

    I later installed a pair of Ultimate Broadband mirrors in hopes of getting something other than boring 632.8 nm (red). But although lasing was strong and stable inside the cavity, there was no evidence of any other wavelengths in the output beams which were weak as expected - these are HR mirrors after all! I also installed the mirror set from a defunct PMS/REO LHYR-0100 yellow HeNe laser tube but also as expected, could not get anything at all probably due to the losses through the Brewster windows, not present in the original internal mirror tube.

    Hughes 3176H HeNe Laser Head with 3509H Power Supply with Low Output

    The Hughes 3176H appears to be the successor to the 3184H despite its lower model number. The case is identical in every respect except that instead of flying leads, it has a normal coax with Alden connector, but still coming out the side. There is also an additional plate on the output end of the laser head with a beam shutter (but that would fit the 3184H as well). The tube has changed slightly though. The HR-end is the same glass stem and glued or optically contacted Brewster window. But the OC-end has an angled metal stem with a glued Brewster window.

    (There was also a 3178H which is only 8-3/8 inches long, so 1 to 2 mW but otherwise similar to the 3176H with the Alden cable coming out the side. The one I have seems healthy, and after aligning the front mirror, it outputs 0.94 mW at 5 mA and 1.0 mW at 6.5 mA. A more careful alignment of both mirrors may boost it some more. Even though the laser head has a standard Alden connector, its ballast resistance is too low to run on a conventional power supply. I assume the internal construction is the same.)

    Before surgery, the laser was tested for output power behavior. In addition, the mirror mount adjustments were carefully tweaked and found to be near optimal.

    Next, the laser head was installed on the test rail described in the previous section and the front mirror was removed and the external OC was fine tuned. Peak power increased to 1.9 or 2.0 mW, though the original mirror appeared to be spotless. Despite some scatter on the Brewster window, cleaning that didn't help very much. In fact, there appears to be some scatter on both the inside and outside surfaces. Despite repeated cleaning with alcohol, the minimum scatter on the outside can never be reduced to the level that is generally present, and no where as low as for my good 3184H. Removing the rear mirror, cleaning the Brewster window, and fine tuning the external HR made little difference. The scatter off the HR Brewster window is also excessive. Walking the mirrors also didn't help much.

    So, with optimal conditions on the test rail, maximum power is about 2.1 mW but that isn't sustainable and it goes down to 1.8 or 1.9 mW after fully warming up. Replacing the head's mirrors - first the HR and then the OC, resulted in a drop of output to close to the pre-surgery levels. The HR had little effect - the OC caused the most reduction in power. I tried an OC from a 3184H but that was no better. The power reduction may be attributable to the slight difference in resonator geometry between the external mirrors and head mirrors. The external mirrors are set up to be marginally TEM00. In fact, optimal output power occurs with a slight doughnut beam (TEM01*). The Hughes laser would have been designed for TEM00, which can result in a slight loss of power. So, it was probably very near optimal in terms of cleaning and alignment even after almost 30 years. Quite amazing.

    The tube is definitely tired. I can drop my good one onto the rail and get 4.3 mW without trying too hard. Whether it's related to the scatter on the B-windows or just abused gas I don't know.

    The output power increases with increasing current beyond where it's running, but that could be a gas problems or just suboptimal resonator due to the scatter.

    Finally, the power supply was adjusted to 6.5 mA after the well hidden pot was found recessed inside the side of the potted HV module against the sheet metal divider inside the power supply. I'm not sure that 6.5 mA is correct - perhaps the 6.8 mA was set to achieve rated power at the factory! But, 6.5 mA should be a less stressful amount of current for an gracefully aging laser! :)

    Conclusions: The patient has been asked to return monthly for followup output power tests to keep track of its health trends. If the output power remains about the same, then the long term prognosis is good. However, if there is an obvious rapid decline over the course of several months, this may be bad news with no cure. No further surgery is anticipated.

    Transplant Surgery for Two Sick Spectra-Physics Model 117 Stabilized HeNe Laser Heads

    The SP-117 is a stabilized HeNe laser which produces a single longitudinal mode output with a nominal frequency of 473.61254 THz stabilized in frequency to a slope of the HeNe gain curve. More information on this and the SP-117A (its successor) may be found at: Description of the SP-117A Stabilized Single Frequency HeNe Laser.

    I acquired an SP-117 laser head, and separately, an SP-117 controller and second laser head (shown in Spectra-Physics Model 117 Stabilized HeNe Laser System). This controller is in pristine condition but quite old, with IC dates codes of 1983 or earlier. I assume the laser head that came with it is of similar vintage. My other SP-117 laser head is dated 1985 and is also in pristine condition, or so I thought. :)

    Unfortunately, the laser tube in the laser head that came with the controller was dead (its organs shown in Spectra-Physics Model 117 Stabilized HeNe Laser Head Components. The tube starts, but the discharge was that sickly white-ish color indicating end-of-life. The tube in my other laser head works fine. However, the controller would not stabilize using it - the output power kept fluctuating in the normal way it does when any vanilla-flavored HeNe laser is warming up, but the Stabilized indicator came on almost immediately, meaning the detection circuitry for the Stabilized indicator wasn't seeing any changes in the voltage level. Poking around inside the SP-117 controller, I found a couple of test points that seemed to be the two sampled signals. One was bouncing up and down but the other one wasn't moving.

    As a test, I took the sensor assembly from the head with the dead tube and plugged it into the controller. Shining a laser pointer on each of the photodiodes evoked a nice strong response from the associated test point so it seemed like there was either something wrong with one photodiode in the head that lased, or some internal adjustment wasn't set correctly. for the sensitivity of the photodiode channel that wasn't working. Since I have no idea of what the adjustment pots do (there are 3 of them, unmarked with any function information), I simply swapped the photodiode assembly from the dead head into the one that lases. It's just two screws and a connector, with no alignment required.

    This resulted in the laser now behaving much better. The voltage on the test points oscillated from about 6 to 8 V during the warmup mode sweep, and after about 18 minutes, the Stabilized indicator started flashing. In another couple minutes, it came on solid for awhile. But some fine tuning of the internal adjustments must be required, because one of the photodiode channels could be seen slowly varying up and down by a few tenths of a volt, and occasionally, the Stabilized Indicator would flash for a few seconds. There was never a mode hop, but the laser's frequency was changing by more than an acceptable amount on the slope of the gain curve. Around 20 minutes later, it finally settled down.

    There is also somewhat of a mystery in that all tests of the supposedly defective photodiode found no problems initially, more on this below. Perhaps whatever is/was wrong with the photodiode is what caused this laser head to be taken out of service before it was run into the ground - these stabilized lasers typically being left on continuously for years to avoid the annoying warmup delay. :)

    I found a healthy 088-2 (greater than 2.8 mW. in my HeNe laser tube collection and have now installed it in place of the dead tube. The axes of the modes of the 088-2 were determined using a polarizer - one of them happened to line up with the exhaust tip-off, so keeping track of that was easy. There were two initial concerns:

    1. The number of modes that oscillate for this tube. Since it's physically identical to the dead tube, I doubt this is an issue. The longitudinal mode spacing is about 640 MHz so it would be theoretically possible for 3 modes to fit in the approximately 1.5 GHz doppler broadened gain curve. However, when stabilized, one of the modes will be locked on a slope of the gain, offset from the center. The second mode would be somewhere on the other slope and this would make it virtually impossible for any additional modes to see enough gain to oscillate. So, during warmup, there might be three modes at times but when stabilized, this is unlikely.

    2. The waste beam from the HR mirror is intense enough to generate a satisfactory signal for the controller. Without pulling apart the working laser head, there is no way to tell the output power of the waste beam. However, again, being physically identical to the dead tube, it is unlikely SP would have made too many changes - if any - specifically for the SP-117 laser. And, it is still an HR mirror with no AR coating, so that's some confirmation.

    The only damage during disassembly was to the RTV silicone blobs holding the original tube in place which will need to be replaced somehow someday, and the loss of 1 or 2 wraps of the aluminum foil, but I doubt that matters as there are 8 or 10 more.

    Reassembly for testing was straightforward. At first, this head had the same photodiode problem as discussed above, but unplugging and replugging the photodiode connector to the photodiodes cured that - at least temporarily. Then, it appeared as though the signal level was very low (concern number 1), but that turned out to be caused by misalignment of the tube with the very small hole in the beamsplitter assembly due to lack of a stable mounting arrangement. Once carefully positioned, the signal level was nearly as strong as with the good SP-117 laser head, and indeed after the normal delay, the Stabilized indicator came on, twitched a couple times, and then settled down. So, there seems to be no problem stabilizing after the transplant. In fact, this laser head has a larger voltage swing for the photodiode signals (about 2 to 6 V, though that perhaps be fixed by an adjustment in the controller) and might even be more stable than the other one. Some adjustment of the controller is probably in order eventually, but the hospital staff does not currently have the necessary skills. A cable to the photodiode channel test points was added so they could be monitored without having the controller open.

    However, there was one final glitch. After mounting the tube somewhat more permanently - with several layers of electrical tape and putting everything back together, the photodiode problem reappeared. In fact, it seems that whenever power was cycled, the connector would have to be removed and replaced to get it out of some latch-up state. In that state, the output is actually negative - opposite of normal. On a hunch, since the connector can be installed in either of two positions oriented 180 degrees from each-other and effectively swapping the two photodiodes, I reinstalled it the other way. And, lo and behold, it now appears to always power up properly. There is no good reason why this should have worked, so the mystery remains. Whether it's something funny about the photodiode or some problem in the controller is not known either. But for now, the system is operational. Since the polarization orientation of the tube was selected arbitrarily, swapping the photodiode in this manner has no more real effect on behavior that matters.

    The entire detailed tube replacement procedure is provided in the next section.

    Conclusions: Both patients appear to be doing well. The controller may need some adjustments so periodic checkups will be performed. Overall though, the surgery would appear to be a success.

    SP-117 HeNe Laser Tube Replacement Procedure

    The following assumes the use of a similar SP-088-2 HeNe laser tube, but any tube of similar length and output power with cathode-end output should work, assuming it can be stuffed in somehow.

    It is best to take a photo at each step for reference. While the position and orientation of most of the parts is obvious, there are some where a photo might be helpful. See Spectra-Physics Model 117 Stabilized HeNe Laser Head Components, though it may differ from yours in the details.


    1. Remove three sets of 4 setscrews each to free the internal laser assembly from the aluminum head cylinder. The setscrews are at the cable-end, a few inches from the cable-end, and a few inches from the output-end. The later two sets may be concealed under some adhesive strips. The output bezel does not need to be removed. It is best to completely remove the setscrews because they need to be turned out so far that some will likely fall out on their own and disappear to where all those missing socks are hiding. :)

    2. Mark the orientation and then remove the output polarizer (if present) by loosening the single setscrew and pulling it free. This step isn't essential but will minimize the chance of damage to it and allow easier access to the end of the tube.

    3. Remove four screws holding the black output optics assembly in place, pull it free, and set it aside. Unclip the cathode wire from the tube. (Note that there may be two different lengths of flat-head 4-40 screws. The shorter ones secure the aluminum tube frame.)

    4. Cut the cable ties and unplug the connectors for the heater (typically red/red/orange or just red/red wires) and cathode return (green wire). Make a note of the connector positions if not obvious.

    5. Remove four screws holding the tube assembly to the black beam sampler assembly, pull the tube assembly free, and unclip the anode wire.

    6. Make a note of how the heater shroud is oriented relative to the tube assembly.

    7. Use a thin strip of metal to slice through the 8 blobs of RTV silicone holding the tube in place as close to the tube surface as possible. This should permit the tube, along with the heater shroud, to slide out of the cylindrical aluminum frame leaving the RTV in place. If this is done carefully, the existing RTV blobs will both center the new tube and hold it snug enough for alignment.

    8. Peel the heater from the tube. Take special care not to damage the printed heater traces.

    9. If the replacement tube doesn't have an aluminum foil wrap, it will need to be removed from the old tube, or replaced with enough kitchen foil to build up to about the same thickness. It may be possible to peel the old wrap free but it is easier to use a sharp blade to slice through just the first 1 or 2 layers since these are glued in place and very difficult to peel off. The loss of those layers won't matter.

    Replacement tube preparation:

    1. Power up the replacement tube on a separate HeNe laser power supply or the one in the SP-117 controller (the control connector will have to be plugged into the SP-117 controller to complete the interlock circuit).

    2. As it is warming up, monitor the output through a polarizer to determine the orientation that results in the maximum fluctuation of power due to mode cycling.

    3. Mark this orientation on the tube to use during reassembly. Power down and disconnect the power supply. Discharge the residual capacitance!!! :)

    4. Install the aluminum foil wrap (if not present) approximately centering it and making sure the foil is at least 1/2 inch from the anode end-cap. It should be smooth and tight and secured in place with thin tape.


    1. Install the heater shroud so that the red wires line up with one of the polarization axes of the tube. If there was an orange wire stuffed under the outer wrap, replace it. I have no idea of its purpose, and it is completely lacking in some SP-117 laser heads and all SP-117A laser heads as far as I can tell. Use some clear tape to secure the heater shroud in place if needed.

    2. Install the tube/heater in its aluminum cylinder, along with the green cathode wire routed through to the output end of the tube and clip it in place. If the original RTV is present, the tube assembly will likely be approximately centered and snug enough to stay in place until being secured permanently with additional RTV.

    3. Reinstall the black output optics holder.

    4. After attaching the anode clip, reinstall the tube assembly into the beam sampler taking care that the anode wire doesn't block the hole to the beam sampler. Make sure the wiring channels line up. Reattach the connectors for the heater (red wires) and cathode (green wire). Take care that the green wires line up in the two halves of the connector!

    5. If the original RTV could not be saved, use whatever is handy to center the tube temporarily to confirm correct operation. Wads of electrical tape and bicycle inner tubes work. :) If the original RTV blobs are present, then at most, some thin plastic shims will serve to secure it temperorarily and center the tube.

    6. Remove the photodiode bracket. Power up the laser and confirm that the waste beams both are fairly well centered in the direct and right-angle apertures. Adjust the centering if necessary. Replace the photodiode bracket making sure the grounding strap is in its original position to press against the cylinder cover when it is replaced.

    7. Allow the laser to warm up while monitoring the photodiode test points inside the SP-117 controller. (I don't recall what they are on the SP-117, but on the SP-117A, these are TP3 and TP6.) They are at the corner of the PCB near the center of the case. The two channels should be roughly equal in minimum and maximum excursion. This typically ranges between 2 or 3 V and 6 or 7 V. The exact voltages aren't critical but they should be set to be about equal. Fine tune the orientation of the tube to maximize the voltage swings. If the behavior of the two channels differs greatly or one is missing, double check the tube centering and that none of the photodiode wires have broken off. Note: Prevent ambient light from reaching the photodiodes as this will affect their behavior.

    8. Install the laser tube assembly into the head cylinder. This is necessary to prevent ambient light from reaching the photodiodes and to provide the correct thermal environment for the laser to stabilize.

    9. The laser should stabilize after the usual 15 to 20 minute warmup (less if the tube had been run for awhile and is still warm) with minimal flashing of the Stabilized indicator. The two test points should be approximately equal after stabilization with minimal hunting or drift.

    10. Assuming this is satisfactory, use some RTV silicone to secure the tube permanently (8 locations). Only use a small amount of RTV so that the blobs are similar in size to what was there originally. DO NOT use a rigid adhesive as the tube needs to be able to expand freely. DO NOT use hot-melt glue as the tube gets hot enough during normal operation to soften it!

    11. Reinstall the output polarizer (if originally present) initially orienting it based on your alignment mark. If having pure polarization is critical, it will have to be fine tuned using a scanning Fabry-Perot interferometer, diode detector monitoring the approximately 700 MHz beat signal, or by some other means. (On later versions of the SP-117/A, this step can be postponed until later, as the polarizer is accessible from the front of the laser head once the bezel is removed.)

    12. Reinstall the completed assembly into the laser head cylinder with the control cable coming out of the rear plate aligned with the output aperture label or shutter lever. Secure it with the 12 setscrews that were there originally. Adjustment of these setscrews can be used for centering and alignment of the output beam.

    The SP-117 and SP-117A laser heads are nearly identical but there are some subtle differences which will need to be taken into account. Some photos of the disassembly process won't hurt in getting the thing back together properly. :)

    Melles Griot 05-LHP-925 Won't Start

    The seller of this laser claimed it "powered up". We know that "powered up" doesn't necessarily mean much in an auction listing but could it be that this laser indeed did run for the seller but somehow is not comatose?

    There are only two reasons for a current of this magnitude. Either the tube is nearly up to air and can't sustain any significant current, or there is electrical leakage. Healthy HeNe lasers will have essentially no current until they start.

    So, the next test was to pull off the anode (cable-end) end-cap and use a hand-held Oudin (Tesla) coil to check for gas integrity. (Reliably exciting the gas in the tube through a long cable with an Oudin coil might not be reliable and would thus be inconclusive if there was no glow.) Indeed, the tube lit up a healthy orange red color with even the minimum output from the Oudin coil. So, it's probably not a gas problem, although it could still be a hard-start tube.

    Next, the head was powered via a jumper cable, bypassing the original Alden cable and end-cap entirely. This still didn't work.

    The only remaining source of electrical leakage is the start tape which runs from the anode the entire length of the tube. It's supposed to be well insulated, but these may cause problems if an attempt is made to run the laser in an excessively humid environment and there is any defect in the adhesion of the Mylar insulating tape.

    So, I performed a starttapeotomy by cutting the start tape off of the anode mirror mount and stuffing its tail down inside the head cylinder. This went smoothly without complications. The laser then started right up on the 05-LPM-915 power supply, first time, every time. The output power after OC mirror tweaking isn't great, but acceptable for these sorts of surplus lasers - about 15 mW after warmup. (It's rated 17 mW.)

    The original buyer of this laser head is located in a very humid climate. So, possibly he was overly eager to test the laser right out of the box off the cool truck, and some condensation found its way inside the head.

    Conclusions: Patient is cured. In my opinion, the start tape on Melles Griot HeNe lasers causes more trouble than can be justified by the statistical improvement in start time that it supposedly provides.

    Spectra-Physics Model 155 Won't Start

    So, something in the HV power supply was probably shorted. First, the HV diode feeding the filter capacitor bank was lifted. This didn't help indicating that the fault was in the HV multiplier. Disconnecting the first 500 pF 6 kV capacitor feeding the multiplier allowed the AC from the transformer and DC on the filter capacitors to return to normal. After systematically lifting several HV diodes and capacitors, it was found that the problem was the first 500 pF 6 kV cap. It must have been shorted or shorting at very low voltage.

    Replacing the cap with a 1,000 pF, 15 kV part (the closest I had available!), seemed to fix the problem and the tube lit up. However, it soon became obvious that it wasn't going to start reliably at normal line voltage. Only when boosted to 125 VAC or more, did it start consistently. And, even then, it wasn't usually instantaneous. It doesn't appear as though the starting circuit is the problem as it produces nice juicy sparks to my Simpson 260 probe even after some time following being powered off. So, it's probably a hard start tube.

    So, a Sam-special start wire was added, soldered to the positive output of the power supply where it connects to the red anode wire, and wrapped several times around the narrow glass tube stem just beyond the cathode. This seems to start first time, every time, and instantly every time. In fact, there appears to be a pre-ionization flash when the power switch is flipped just before the main discharge goes on continuously. A start wire won't be useful if the tube is hard-running since it will tend to turn the tube off as readily as it turns it on and the result will be a flashing laser. But this tube stays lit down to something like 70 VAC. It's just hard starting.

    Conclusions: The original problem may have been the hard starting tube, causing stress on the HV capacitor and its failure. It's possible that this may have blown the fuse if allowed to be in the shorted state unattended, but the person who worked on it before me had replaced it. Now the patient seems healthy enough, easy starting, runs without problems, and has an output power of about 0.9 mW.

    Two Tortured LASOS LGK7512P Yellow HeNe Laser Heads

    Since Patient #2 at least started, some exploratory surgery was in order. The front end-caps on these laser heads are just press-fit into the cylinder. So, using my custom laser head end-cap extractor (a 3/8" thick aluminum plate with 4 holes drilled through it so that screws can be attached to the end-cap), the end-cap on #2 was wiggled loose and pulled off. Then, the problem became immediately obvious: The plastic "cup" in which the anode ballast resistance was potted in red rubber stuff had partially melted! Once this was removed, the beautiful yellow beam in all its glory reappeared! Why had the plastic melted? I have a theory and this will be discussed later. After using a variety of small tools including reamers and drill bits to clear out the melted plastic, the laser head could be reassembled. After warmup, the output power was measured at greater than 2.6 mW for this laser rated at 2 mW. Not too shabby and close enough to the 2.73 mW scribbled on a sticker as to be considered as good as new.

    Now on to Patient #1 which wouldn't start. Something inside seemed to rattle which is usually not a good sign. After removing the front end-cap as with #1, it appeared as though the ballast of this laser head was also somewhat melted, though not nearly as badly. I tested for continuity from the Alden connector to the anode of the tube and that was fine, as was the negative connection to the cylinder. At this point, I was fairly sure the tube was broken and up to air but before pronouncing the Patient #1 dead, there was the test of last resort: The Oudin coil. And sure enough, applying some nice high frequency high voltage sparks to the anode mirror mount revealed a healthy orange glow inside the tube. There's nothing wrong with the gas in this tube! So, the only other explanation would be that the connection between the cathode and cylinder/negative of the Alden was open.

    The rear end-cap doesn't come off nearly as easily as the front one. Working around the perimeter with a sharp blade along with removing the cable clamp enabled it to be pulled free. Sure enough, the wire for the cathode mirror mount was just dangling! And, there was a cathode ballast resistor which was clearly fried to a crisp. Curiouser and curiouser. With a trusty alligator jumper lead between the mirror mount and the cylinder, Patient #1 started up just fine. After installing a new resistor (guessing 10K, 1 W would be fine since 0 ohms worked fine), it also produces about 2.6 mW after warmup.

    So what happened to cause very similar failures of both these lasers? One possibility is that there was a power supply failure resulting in excessive current to the tube. The optimal current is 6.5 mA. If the current increased to, say, 8 mA, the power dissipated in the ballast resistors would go up by 50 percent. At 9 mA, it would be almost twice as high. Higher current wouldn't exactly be too good for the tube over time, but the additional heating could likely be much more immediately damaging to the ballasts. The crisped remains of the cathode ballast resistor didn't appear to be rated more than 1/2 W and would be dissipating over 0.4 W under normal operating conditions. The anode ballast resistors are potted in red rubber stuff so there is no way to know what their ratings are, but possibly that conformal coating helped to spare them the same fate, simply passing the excess heat to the plastic shell, which showed signs of melting in both patients, but more seriously in Patient #2. Patient #2 has a relatively recent date of manufacture (compared to now, which is August 2006): November 2005, so it may have been a replacement for Patient #1 (which is from September 2003) after it failed completely. When Patient #2 also failed with the blocked beam - and likely rather quickly - someone probably realized there was a problem with the power supply.

    Conclusions: Both Patients #1 and #2 seem to be fine now. The cathode ballast resistor in Patient #2 was not replaced since there are presently no symptoms, but it should be monitored to make sure that problems don't develop in the future. But perhaps a higher power resistor was used and it survived. Unfortunately, there is no way to determine the actual history of these lasers so the runaway power supply hypothesis will have to suffice for a cause.

    REO One-Brewster Tube - No Lasing

    This is the third sample of this system that I've acquired. (See the section: The Ohmeda Raman Gas Analyzer One-Brewster Laser.) The first two lasers were fine with healthy tubes. The tube in this one is obviously sick. Its complexion is decidedly too pinkish and the brightness of the discharge is also somewhat weak. In addition, it exhibited the "swirling lightning bolt syndrome", apparently typical of contaminated PMS/REO tubes with nearly full length cathodes. While most of the staff thought there was no hope with no possibility of successful treatment, something about this patient demanded that an effort at least be made. Extended running time on normal power is a last resort option for these cases. While the success rate is low, the cost is minimal.

    It seems that for these high-Q tubes, REO still uses a soft-seal for the B-window to minimize stress. Either that, or the optical contacting wasn't entire successful since the tube definitely leaked.

    So, the tube was separated from the rest of the assembly and simply allowed to run on its original power supply for 8 to 12 hours a day over the course of three weeks.

    Initially, the only indication that this was having any effect at all was a subtle improvement in the discharge color more toward the normal "salmon". However, the brightness of the discharge was still obviously low. Periodic checking for lasing with an SP-084 HR mirror was negative.

    Since there was some, if minor, improvement in complexion, it was decided to order some additional tests. A complete spectral scan using a Verity monochromator-detector was performed for this patient as well as another similar REO 1-B tunable laser tube (details to follow). However, the results of these scans were rather inconclusive. It was expected that there would be some sort of contamination or low pressure. But there was no indication of significant non-He or non-Ne spectral lines and in particular, little or no hydrogen. The ratios of He:Ne were also similar to healthy tubes. Perhaps, if the detector gain was cranked way up, H2 contamination at 656 nm would have become apparent - even a very small amount of H2 - something like ten parts per million - will kill lasing.

    The next test to be performed was for double pass gain. A Melles Griot 05-LHR-911 HeNe laser was sent through the Brewster window and reflected off the internal HR. This is the "probe" beam. A non-polarizing beamsplitter reflected a portion of the return beam to a laser power meter. By turning the REO tube's power on and off, the change in power of the probe beam could be determined. (The 05-LHR-911 had to be warmed up for at least a half hour so its mode sweep power variations wouldn't confuse the readings. And, it was confirmed that the bore discharge glow didn't affect the measurement significantly.) Here are very approximate results:

       Time on Therapy    Double Pass Gain    Discharge Appearance
           14 days            0.25 %          Still weak, but salmon color
           18 days            0.50 %          Swirling lightning bolts disappear
           20 days            0.75 %          Brightness improving
           21 days            1.50 %          Nearly normal complexion
           22 days            3.00 % *

    There was just enough improvement in the appearance of the discharge over the first two weeks to suggest that a positive outcome was possible even if continued testing with the SP-084 HR was unsuccessful. But once the double pass gain reached 1.5 % I was thinking: "This has to work now!". :)

    And, indeed, finally, it was possible to easily obtain flashes using the hand-held SP-084 HR mirror. So, the patient was put back together, first with the same HR mirror to allow for initial cleaning of the Brewster window. Then, the Brewster prism assembly was installed, aligned, and cleaned. The output power has continued to climb, especially for the orange (611.9 nm). (* With the intact patient, there is no easy way to measure the gain, which also no doubt would be seen to increase. For this length tube, 3 perecnt double pass gain would be my estimate.)

    And continueing with output power measurements:

                <------------ Internal HR ------------>
      Time on   632.8 nm   <-------- 611.9 nm -------->
      Therapy    Output     Output   Increase   Percent   Comments
       0 days      0 uW        0 uW                       Start - no output
      22 days    365 uW      583 uW   583 uW              "First Relight", cleaned
      23 days     ----     1,405 uW   822 uW   140.99 %   Increasing rapidly
      24 days    565 uW    1,604 uW   199 uW    14.16 %   Increasing slowly
      25 days    735 uW    1,735 uW   131 uW     8.17 %      "        "
      26 days     ----     1,792 uW    57 uW     3.29 %      "        "
      27 days     ----     1,817 uW    25 uW     1.39 %      "        "
      28 days     ----     1,860 uW    43 uW     2.37 %      "        "
      29 days    807 uW    1,903 uW    43 uW     2.30 %      "        "
                  ----     1,963 uW    60 uW     3.15 %   Re-cleaned B-window *
      30 days    830 uW+   2,016 uW    53 uW     2.70 %   Increased some more
      31 days     ----     2,041 uW    25 uW     1.24 %   Increasing more slowly
      32 days     ----     2,069 uW    25 uW     1.23 %
      33 days    864 uW+   2,097 uW    28 uW     1.35 %
      34 days     ----     2,115 uW    18 uW     0.86 %
      35 days     ----     2,110 uW    -5 uW     0.03 %
      36 days     ----     2,110 uW     0 uW     0.00 %   Stable
      38 days     ----     2,110 uW     0 uW     0.00 %   Stable after 1 day rest

    * This 60 uW increase in orange power was due to window cleaning. Any increase due to gas cleanup seemed to be minimal, if any, when the cleaning was done. However, the output power then increased another 50+ uW over the next several hours and the next day. + Indicates that the red power is actually slightly greater than the value shown since it was measured before the end of the "day".

    Note that a "day" is actually about 10 to 14 hours of run time. The laser sleeps when I sleep. :)

    The status for last full test is shown below:

                   Power from   Power from
       Wavelength  Internal HR  External HR
        632.8 nm      864 uW      147 uW
        611.9 nm    2,080 uW       29 uW
        604.6 nm        0 uW        0 uW

    The total power out both ends for red is much greater than that of the other two samples of this laser assembly which had no problems, though the exact split of power between the internal and external HR mirrors differs. But that is a mirror issue. The intracavity power isn't known, but based on measurements of one of the other lasers, it may be as high as 10 WATTs or even more. And, the total power for the 611.9 orange line is actually significantly higher than the other lasers. However, unlike those, there is no evidence of the 604.6 nm orange line. This, too, is probably a mirror issue since the requirements in the Raman analyzer are only for high circulating power for the red line.

    However, based on a guess as to the meaning of the "S" and "T" parameters (See the section: The Ohmeda Raman Gas Analyzer REO One-Brewster Laser.), the present performance for red may only be less than 1/4 of what's possible if this laser was new and had perfectly clean optics.

    Conclusions: The performance of this laser for 632.8 and 611.9 nm is now better than that of the other 2 known good samples. There is now little change from one day to the next. It's not known whether the contamination originated inside the tube, or from leakage through the soft-sealed B-window. The patient has been discharged (no pun...) but will probably need to have periodic run time to maintain good health.

    Update 1: At the first followup visit, approximately 1 month later, the 611.9 nm output power (the only wavelength tested) reached 95 percent of its post-treatement value within 1/2 hour, but didn't seem to want to climb higher, and then declined a few percent. Cleaning the B-window had no effect, though other surfaces could still have gotten contaminated. But after a couple more power cycles, it was up to 98 percent, close enough for government work. :) After a few more power cycles, it peaked at the original power but wouldn't hold it, so it was run for another 14 hours straight, at which point it recovered and maintained to slightly greater than post-treatment power. (The increase may have been due to the additional Brewster cleaning.) Continuous running is probably what should have been done. Forget that stuff about power cycling on basically healthy tubes! :) When the tube gets hot, it outgases contaminants from various surfaces causing the reduction in power. But only at that point can the weak gettering of the cathode have any effect. As long as the power only declines slightly, just continue running and it will recover.

    So, it would appear based on the results of this followup treatment, that the laser loses about 5 percent of its output power per month from being idle, and that running it a couple hours per day, or one day a week may be needed to avoid this decline. Another followup will follow in 1 month.

    Update 2: At the second followup visit 1 month later, the 611.9 nm orange power peaked at 2.01 mW, then declined to 1.98 mW after 1/2 hour. But after running straight for about 23 hours (no sleep), it had recovered to 2.10 mW. And after another 12 hours, had reached 2.14 mW, higher than the last visit. The patient was asked to return in 2 months.

    Update 3: After two months of non-use, the orange power after an initial warmup period of 1/2 hour was only 1.9 mW but recovered fully to 2.14 mW after about 50 hours of continuous exercise. The patient was asked to return in 6 months. I wish it would do this on its own though. :)

    Although running soft-seal HeNe lasers is recommended for continued health, and extended run time for a weak or zero-output tube may help sometimes, in my experience, only PMS/REO tubes have a good chance of recovering to like-new performance using this technique. However, there seem to be three types of sick PMS/REO tubes with a weak or pink discharge:

    1. Type 1: Running at normal operating current for 10s of hours will restore to near-new specifications. This has also been behavior of an old LSTP laser (not mine) and with the laser in an Ohmeda anesthesia Raman gas analyzer.

    2. Type 2: Running at normal operating current for 10s of hours may result in some improvement but continuation results in output power going down. An additional characteristic is that when turned on from a cold start, the output power will initially be much higher, perhaps near-spec, but then will die away within a few seconds to minutes.

    3. Type 3: No amount of run time has any effect on either discharge appearance or output power, which is probably 0.0 mW. My guess is that this is simply (2) if continued to run into the ground.

    Also, PMS/REO tubes seem to have one unique characteristic when they are gassy. Namely, that if one looks down the inside of the tube from the glass-end, there will be swirling white-ish streamers visible between the cathode and bore. And, it has been suggested that when the swirling clouds are present, it's a good indication that extended run time will result in a successful cure. What is the cause?

    Update 4: As expected, the patient negelcted to schedule a visit and it's been about 1 year since the last followup. Unfortunately, this meant that recovery will be lengthy and costly, if possible at all.

       Time on Therapy     611.9 nm
            Start          0.000 mW   
            30 min.        0.348 mW
            1/2 day        0.886 mW   
            1 day          1.240 mW   
            2 days         1.361 mW   
            3 days         1.450 mW   
            4 days         1.650 mW
            5 days         1.744 mW
            6 days         1.797 mW
                           1.912 mW Cleaned Brewster window
            7 days         1.968 mW Testing ended

    The only wavelength that is being monitored is 611.9 nm since this has been shown to track the 632.8 nm output quite reliably. Since the power seemed to be leveling off after 6 days, it was decided to do a Brewster window cleaning, which resulted in the output power immediately increasing to 1.912 mW, and then somewhat unexpectantly further increasing to 1.968 mW, within about 5 percent of the previous best value. Knowing that this patient will not stay clean over the long run, the more risky and expensive tuning prism cleaning was not performed. However, assuming a similar degree of contamination on each of two surfaces, that would almost certainly restore full power, and possibly more.

    REO One-Brewster Tube - Slightly Low Output

    Here's a case where it's a good idea to get periodic checkups. Had I not been treating the other REO laser, this patient would probably never have been evaluated at all. And, this classic case of low poweritis may have progress to the point where serious intervention would be needed, if it could be treated at all. As it is, running for a few days should be sufficient to restore it to perfect health. In fact, better than before since it had never been run long enough for complete gas cleanup to occur.

    The following time-line starts at the point of peak output, about 5 minutes after power-on:

                <------------ Internal HR ------------>
       Time on   632.8 nm   <-------- 611.9 nm -------->
       Therapy    Output     Output   Increase   Percent   Comments
      0.0 hours    ----     1,451 uW                       Start - Peak output
      0.5 hours    ----     1,259 uW  -192 uW    -13.0 %   Minimum output
      2.0 hours    ----     1,314 uW    55 uW      4.3 %   Initial increase
      3.0 hours    ----     1,317 uW     3 uW      0.2 %
      6.5 hours    ----     1,312 uW    -5 uW     -0.3 %

    Here, "hours" are real non-relativistic time; if it gets to days, they will be my normal 12 hours or so/day. :)

    If the treatment is interrupted for even a few seconds, there is a short term increase in power. If interrupted for a few minutes, the entire progression starting near peak power, declining to a minimum, and then recovering to the steady state power (about 1,312 uW), repeats. It's not clear at this point if simply running the laser for any amount of time will be successful. But it is almost certainly a gas contamination problem.

    Alert!!! Patient went into partial photon arrest during cleaning. I attempted to clean the B-window and then the Brewster prism. The power kept declining. I am not aware of any obvious problem with the tube and no damage to any of the optics surfaces or the external HR mirror. After multiple attempts at cleaning including completely removing the mirror and cleaning (carefully), output power on 611.9 nm is only about 1/10th of what it was before. There is no 604.6 nm at all. Long term intensive care may be required.

    After a frustrating lack of improvement from optics cleaning, exploratory surgery was called for. So, the Brewster prism/external HR assembly was removed and replaced with a 98 percent 632.8 nm OC mirror. On another similar laser assembly (Laser 1 from the section: The Ohmeda Raman Gas Analyzer REO One-Brewster Laser), this results in 5.4 mW of output power. However, on this laser, it maxes out at 2.8 mW. There is virtually no detectable scatter on the Brewster window and the mirror is clean (or at least as clean as it was when Laser 1 was tested). While the internal HRs on this laser and Laser 1 differ in their reflectivity, it is still very very high on both and thus the beam out of the internal HR should not be a significant factor. I can't absolutely confirm that the tube current is correct, but varying the input voltage to the power supply results in negligible change in output power implying that the power supply is regulating properly. Thus, the current has almost certainly not changed. So, where the missing power went is a mystery, but no doubt with 632.8 nm well below expectation, the 611.9 nm power will be very small.

    Here are some vital stats without the Brewster prism and REO external HR:

                   Power from    <------- External Mirror ------->   Intracavity
       Wavelength  Internal HR     Type     Reflectivity    Power       Power
        632.8 nm       8 uW      60 cm OC     98.0%       2,800 uW      0.14 W
         "    "       40 uW      SP-084 HR    99.966%       282 uW      0.85 W 

    Stay tuned.

    Conclusions: None at present.

    REO One-Brewster Tube - Very Low Output

    This is the resonator assembly from the unit described in the section: The REO One-Brewster Particle Counter HeNe Laser. This is a basic one-Brewster resonator with no tuning prism or other intracavity optical elements. But unlike the patients in the two previous sections, the bore discharge is clearly visible from an exposed section of the glass portion of the tube. So, the sickly complexion was immediately obvious. Knowing that PMS/REO HeNe lasers have a good chance of recovering with extended run time, that treatment approach was initiated. I'm actually rather surprised it lased at all.

    Since this unit has HR mirrors at both ends of the laser, even a perfectly healthy tube won't result in huge output power - output from both ends are waste beams. But since the tube is similar to ones in other similar systems, they should be higher than 5 uW and 15 uW! This laser has a manufacturing date of 1996 - over 10 years of age. However, when it was taken out of service is not known.

    Whoever salvaged the laser decided it would be creative to cut the power supply wires literally 1/4 inch from the Voltex HeNe laser power supply module. Attaching new wires was a real treat, especially attempting to insulate the one for the high voltage! Hopefully, the multiple layers of heat shrink tubing and electrical tape will be adequate. For now, it does work without unsightly incidents such as arcing or meltdowns! :)

    The external HR mirror was inspected and appeared perfectly clean, and the Brewster window was cleaned without any significant change.

    But sure enough, improvement in both output power and discharge appearance was evident very quickly, with the approximate measurements below:

        Time on   Output from    Output from  Intracavity
        Therapy   Internal HR    External HR     Power
         Start        5 uW          14 uW       0.33 W
        2 hours      13 uW          37 uW       0.85 W
        6 hours      20 uW          57 uW       1.30 W
       10 hours      40 uW         115 uW       2.60 W
       14 hours      60 uW         172 uW       3.90 W
       17 hours      72 uW         207 uW       4.68 W
       20 hours      86 uW         247 uW       5.59 W

    At this point, two things were done. First, 2 of the 3 IR suppression magnets were re-glued in what I thought were the same position as they were originally. (The original glue job was ugly!) And, an external OC was substituted for the internal OC to be able to determine intracavity power and the reflectivity (or transmission) of the REO HR mirrors. This all went smoothly and after the external HR was replaced, the laser seemed to be happy, with similar waste beam power as before surgery.

    However, a critical situation requiring emergency care developed! The waste beam power started declining slowly but surely until after several hours, it was down to about half of the last measurement above. There was no obvious explanation for this turn of events. Since the power had not changed after re-installing the external HR, contamination on its surface was unlikely. But the Brewster window was checked and cleaned with no change. Alignment was checked and found to be perfect. Tube current was unchanged at 5 mA. The only possible explanation other than an unlikely coincidence that the tube just decided to become end-of-life was that the position and/or orientation of the magnets was not the same and somehow, this resulted in the power falloff. I know that's a stretch but when one has eliminated all the likely suspects....

    The first thing I tried was to restore the magnets as best I could to what I thought was precisely the original position and orientation. This was based partially on photos of the unmodified assembly and partially on my recollection. But this didn't seem to make much difference and the decline continued.

    So, it was time for desperate action! The tube was pulled from the particle counter assembly and placed in my one-Brewster tube intensive care unit (1-B ICU) with the 99 percent OC mirror and no magnets. Its initial behavior was not promising. When powered on from a cold start after having been off for 24 hours, the output power would peak at over 3 mW within a few seconds and then decline over the course of a few minutes to less than 1.5 mW and appeared to be continuing its steady decline. Maximum output power was achieved with a tube current of around 6 mA when the laser was first turned on. But as the output power declined, the current for maximum power increased to beyond where it would be safe to run the laser. Once the output power (at the normal tube current of 5 mA) declined to 1.5 mW, power was turned off to avoid doing something irreversible. This cycle could be repeated (after waiting 24 hours). (The peaking and rapid decline in power was evident with the laser in the particle counter assembly, but was only a 10 or 20 percent difference, not the more than 2:1 as it is with the OC mirror.)

    So, an emergency conference of the department heads was convened. :) It was decided that there was nothing to lose by simply allowing the laser to run continuously as there were no other treatment options available. (Regasing would not have been covered by laser's health insurance plan.) And then, soemthing totally unexpected happened: The power bottomed out at around 1.3 mW and started climbing:

        Time in   Change/   <----------- Output from External OC ------------>
        1-B ICU   24 hours    5.0 mA    5.5 mA    6.0 mA    6.5 mA    7.0 mA
         Start               1.33 mW
        3 hours              1.60 mW
        6 hours              1.85 mW
        9 hours              2.12 mW
       21 hours              2.90 mW
       24 hours   0.73 mW    3.06 mW
       27 hours              3.18 mW
       29 hours              3.26 mW                                 3.71 mW
       33 hours              3.40 mW                                 3.87 mW
       45 hours              3.65 mW                                 4.11 mW
       51 hours   0.61 mW    3.75 mW                                 4.21 mW
       57 hours              3.85 mW
       69 hours              4.00 mW                                 4.47 mW
       72 hours   0.34 mW    4.05 mW
       79 hours              4.12 mW
       83 hours              4.18 mW                                 4.62 mW
       95 hours   0.26 mW    4.30 mW                                 4.80 mW
       98 hours              4.34 mW   4.58 mW   4.75 mW   4.85 mW   4.90 mW
      109 hours              4.42 mW    
      118 hours   0.21 mW    4.50 mW   4.69 mW   4.89 mW   4.98 mW   5.02 mW
      130 hours              4.54 mW   4.76 mW   4.92 mW   5.01 mW   5.03 mW
      142 hours   0.15 mW    4.65 mW   4.88 mW   4.98 mW   5.03 mW   5.07 mW


    1. Values for "Change/24 hours" were interpolated if an exact multiple of 24 hours was not available.

    2. The therapy current was increased to 5.5 mA from 5.0 mA after 130 hours to assure stabilily on the adjustable unregulated HeNe laser power uspply. The clean tube appears to have a higher negative resistance and higher dropout current using the original ballast resistance and the discharge was tending to go out and restart on AC line fluctuation.

    3. Originally, when the tube was very weak at 1.3 mW, but started at around 3 mW, the output power initially would peak with a current of around 6 mA. Once the power declined, the current for maximum power would go too high to safely run the tube, even for measurements. However, once the sustained power had climbed above about 3 mW, the current for peak power dropped to around 7.0 mA. This somewhat higher than common current may be correct for this multi-spatial mode wide-bore laser. The peak has been very close to 7.0 mA since 29 hours when the current is increased (from the therapy current of 5.0 or 5.5 mA) for measurements. There is an uncertainly of perhaps 0.1 mA but no statistically significant movement since then.

    I've never heard of IR suppression magnets affecting gas cleanup behavior but that is the only explanation that makes any sense. And the rapid repeatable drop in power was definitely a gas contamination issue, so that eliminates any issues with Brewster window cleaning or the external HR mirror. My hypothesis is that the change in magnetic field disturbed the location of the discharge. The magnets that probably were responsible were the ones at the cathode-end of the tube. There, the discharge is spread out where it hits the cathode, and that would be susceptible to being moved by the magnetic field. Originally, the two magnets located there had the same poles (N or S) facing the tube. However, when I re-glued one of those, I had opposing poles facing the tube so they would attract each-other and hold the magnets in place while the glue dried. This made no difference in terms of IR suppression, but could have had a big impact on pushing the discharge around. In the 1-B ICU, there is now no field, so it will be interesting to to see what effect replacing the magnets will have on behavior. If this is repeatable, there could be something significant and potentially useful going on, perhaps with implications for the treatment of other lasers.

    Why does discharge location influence gas cleanup? The areas of ion bombardment and heating changed and this must be affecting the "bad gas" atoms. So they are being trapped and released from the cathode surface. Or something. :)

    I do wonder if some of the very slow improvement near the end is actually due to the tube aging. This may have been a nearly unused laser. New tubes are often overfilled to maximize run time life, and the discharge seems a bit brighter and more orange than typical for common HeNe lasers.

    After completion of treatment, the tube was returned to its body. The intracavity power was found to have increased by over 92 percent to over 10 WATTs! This was at a tube current of 5 mA. The intracavity power reaches almost 12 W at around 7.0 mA. Power from a cold start now increases monotonically, initially at about 80 percent of the final value (at 5.0 mA).

    The patient will be monitored for awhile to confirm stability but any dramatic change is unlikely. The only difference between the 1-B ICU and the particle counter assembly is the (rotational) orientation of the tube. And, I'm not prepared to believe that gravity will have a detectable effect!!! :)

        Time on   Output from    Output from  Intracavity  Operating
        Therapy   Internal HR    External HR     Power      Current
       170 hours    165 uW         474 uW       10.7 W       5.0 mA
         "   "      182 uW         523 uW       11.8 W       7.0 mA

    This is now the total time from when the tube was first turned and was very weak. It is more or less continuous, running day and night for over 7 days straight except for a gap of 2 days between the first and second sets of data, above (before entering the 1-B ICU), and while the tube was being installed back into the particle counter assembly. At this point, the output power has essentially leveled off, within the measurement uncertainty.

    The amount of scatter off the Brewster window is rather small considering the intracavity power. So, I bet if the external HR was similar to the internal HR, perhaps 50 percent more intracavity power might be possible.

    The magnets do not seem to have any profound effect when momentarily installing them to check power, so they will be left off for now at least. We can do without relapses!

    Conclusions: If as is likely, this laser has been out of service for several years, then the treatment should result in decent performance being maintained without requiring frequent running to clean up gas contamination. This will require one or more followup visits to confirm. The next test will be to see if replacing the magnets results in this entire decline and restore cycle to be repeated. That seems likely. Since the increase in performance with the magnets is only a few percent, it may be best to simply not re-install the magnets at the cathode-end of the tube, or to move them a bit further toward the anode so that the effect is inside the bore rather than at the spread-out cathode discharge. Some futher testing may be performed in the future to determine which approach is best. But that won't happen for awhile. The first followup visit will be in 1 month.

    Update 1: In about 1 month, the laser was run for several hours at the normal 5.0 mA. It started at 131 uW (from the internal HR) and climbed to 160 uW after about 8 hours but didn't seem to be increasing any further. Since there are many opportunities for contamination to enter despite the various seals, a Brewster cleaning was ordered and resulted in 175 uW. Slightly more power might be possible. It's not clear how much of the improvement is due to the Brewster cleaning but almost certainly much of it. The next followup will be in 2 months.

    One interesting observation - not unique to this particular laser - is that when the boot is put back into place after cleaning the Brewster window, the output power will actually *increase* slightly over the next minute or so. The change is only 1 or 2 percent, but it is real and is not associated with a shift in alignment, nor probably to residual solvent evaporating or something like that as might be suspected. Rather, the reason is likely that dust particles that entered the (hopefully) sealed interior of the cavity when it was open to my non-clean room lab are settling out. Thus they are no longer producing scatter of the intracavity beam, and its associated power loss. Short of figuring out how to get the particle counter photodetector and electronics working, it may be possible confirm this by looking for optical noise in the waste beam to decline over time after the boot is put back into place. Something for the future!

    Update 2: As is typical with these patients, the next followup was not in 2 months, but more like 6 months. However, the laser must have been eating ealthy and exercising regularly as its performance after only 1/2 hour was very close to previous values - 170 and 483 uW.

    PMS One-Brewster Tube - No Lasing

    This patient came in a PMS LSTP-1010 5 color tunable HeNe laser, probably one of the most way-cool HeNe lasers ever produced. When operating properly, the output can be selected among 5 wavelengths: red (632.8 nm), orange 1 (611.9 nm), orange 2 (604.6 nm), yellow (594.1 nm), and green (543.5 nm). However, when I acquired this tube in 2002. It was already very sick and capable of only a few hundred microwatts of red continuously, and perhaps a flash of orange when initially turned on. It continued to decline from there. I had pretty much given up on it until attempting to revive the PMS tube described in the previous section. It was convenient to perform a spectral scan on this one as well, and the results were virtually identical. This provided hope that it too could be revived with extended run time. It is now 2007 and at least the appearance of its discharge hasn't changed detectably in 5 years!

    So the plan is to run this tube while checking its double pass gain periodically over several hundred hours if necessary. As before, the double pass gain will be monitored by reflecting a red HeNe laser beam up and back from its internal Brewster window and extracting a portion of the return beam with a beamsplitter. First, I used my trusty reliable Melles Griot 05-LHR-911. But that laser takes 2 hours to warm up to the point where the power variation due to mode sweep is slow enough to deduce a small change in reflected power when the PMS laser is turned on and off. So, I substituted a Spectra-Physics 117C stabilized laser which settles down in 10 to 15 minutes from a cold start. Might as well use it for something! :) Although I haven't figured out how to switch it to intensity stabilized mode from frequency stabilized mode (it's a jumper block and I haven't found any docuementation!), the total power is still quite constant.

    Its initial condition is that the double pass gain is around 0.75 percent. This is somewhat higher than I had expected, but with its internal OC mirror likely having a reflectivity of 99 percent (transmission of 1 percent) for 632.8 nm, no red lasing is even possible. And testing for other wavelengths won't be done until it does decent power for red.

    Even with the stabilized laser, the power readings still fluctuate enough to be confusing, so I constructed a simple passive circuit to take the difference of the difference between the incident and reflected beams, adjusted for equal gain. It would have been better to normalize this result automatically, but that would have required a divide somewhere which was more work than I really wanted! Another option would be to capture the measurements with a data acquisition system and do the calculations with a C program or MATLAB. For now, the passive circuit will do. :)

      Time      Gain      Comments
      Day 1    0.75 %     Started  5.25 mA,  slightly pink
      Day 2    0.80 %     Small improvement, still somewhat pink
      Day 4    1.00 %        "             "             "
      Day 9    1.00 %     Unchanged

    Due to the large uncertainty in the measurement of gain, "unchanged" doesn't really mean much, just that the change, if any, wasn't dramatic.

    Interestingly, for a few seconds after being powered on after being un-powered for awhile, a weak red beam would appear and then die out quickly. If the "off" duration was several hours, the beam might start out at around 1 mW and take 25 seconds to disappear completely. With a shorter rest, there would be a less intense beam of shorter duration. This is similar behavior to what it was doing several years ago, but then the output power was higher (with some orange even possible when installed in the tunable laser case) and the duration of the lasing was longer. At that time, extended running had at best no effect, and possibly was making it worse. But, it has obviously deteriorated further since then.

    Conclusions: After several days with absolutely no change, it seemed obvious that the tube was too far gone to recover. Will probably try again in the future though. At least, it doesn't now appear to be deteriorating while sitting on the shelf. If only, those bad gas atoms or molecules could be dispatched to a place where they wouldn't interfere!

    Aerotech LS4P HeNe Laser Tube - No Lasing

    This is an interesting early (probably late 1970s) polarized HeNe laser tube. It consists of a soft-sealed two-Brewster plasma tube with full diameter glass extensions on which the mirrors are mounted. The OC mirror is Epoxied in place but the HR mirror is on a 4-screw (yes, 4) adjustable mount. The HR mirror itself is rectangular which almost certainly means it's planar and cut from a larger piece. (This is the only rectangular cavity mirror I've ever see on a HeNe laser!) But it is also about the most finicky mirror as well. Breathing on the mount changes alignment, and power variations due to thermally alignment changes can be 2:1 or more.

    The discharge color was excessively pink and there was no beam, even after fiddling with the (adjustable) rear mirror. This color discharge sometimes means that recovery is possible with extended run time. So, the patient was placed on continuous run current therapy and sure enough, after a few hours, a beam appeared.

    The original treatment was performed a year or so ago and unfortunately, the records were lost. However, now (2007), retreatment was needed since these soft-seal tubes deteriorate with non-use. An I must admit to neglecting the required frequent petting to keep this one happy.

    Here are the stats for the retreatment.

      Time on     Output
      Therapy     Power     Comments
       Start      0.0 mW    Initial powerup after long rest.
       12 hours   0.6 mW
       24 hours   1.1 mW
       36 hours   1.1 mW
       48 hours   1.1 mW
       60 hours   1.2 mW
       72 hours   1.3 mW
       84 hours   1.4 mW    1.8+ mW peak with magnets during warmup.
       96 hours   1.7 mW
      108 hours   1.8+ mW   Will not stay lit below 7 mA.

    The tube was allowed to rest at night, so each treatment period was approximately 12 hours. The power output reading is the peak that could be obtained by pressing on the HR mirror mount. Running continously (no rest at night) would probably improve it slightly, but the decay rate is so high that it could only be maintained with nearly continuous running. What's interesting that the tube does seem to be able to recover to having an output power similar to what it was when I first got it.

    This tube was always very particular about current and required a 10K cathode ballast resistor to run stably at 6.5 mA, but now as the gas cleanup has progressed, needed 7 mA to be happy. In addition, if the discharge dropped out and attempted to restart, there would tend to be a very weak glow inside and outside of the bore for several seconds or longer until it actually started. This is very unusual for any tube.

    Conclusions: It's possible that running for more time would get it closer to the spec of 4 mW, though this is doubtful. However, the high current means that a normal life will never be possible. I don't know for sure what the optimum current was supposed to be, but it certainly was less than 7 mA! The patient has been sent home with instructions to return every few days for treatment. But it almost certainly won't, so we'll be doing this full regiment again in the future. :)

  • Back to Sam's Laser FAQ Table of Contents.
  • Back to HeNe Laser Testing, Adjustment, Repair Sub-Table of Contents.
  • Forward to HeNe Laser Power Supplies.

    Sam's Laser FAQ, Copyright © 1994-2009, Samuel M. Goldwasser, All Rights Reserved.
    I may be contacted via the
    Sci.Electronics.Repair FAQ Email Links Page.