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.

    Amateur Laser Construction

    Sub-Table of Contents

  • Back to Sam's Laser FAQ Table of Contents.
  • Back to Amateur Laser Construction Sub-Table of Contents.

    Introduction to Amateur Laser Construction

    So You Think You Really Want to Build a Laser

    While most lasers are extremely high tech devices which require the engineering and manufacturing expertise of corporations, universities, or Government agencies to design and construct, there are a few types that can be built from scratch by the very determined amateur. I'm not talking about wiring a helium-neon laser tube to a power supply or a laser diode to a driver circuit. A truly home-built laser may start out as 4 foot lengths of various sizes of glass tubing, mirrors, special gases and chemicals, scrap metal and hardware; electronic components like transformers, rectifiers, capacitors, and resistors - and laser and high voltage warning signs! Converting this collection of materials into a working laser will require many many hours of effort as well as blood, sweat, and possibly tears. :)

    Considering the enormous gap between your likely capabilities and those of a large corporation, the final result probably won't match a commercial laser in either performance or appearance (though there are a few exceptions). However, that shouldn't be the point of the exercise. Rather, a laser (or multiple lasers after you are hooked) should be built for the education, challenge, and opportunity for experimentation afforded by dealing with all aspects of laser construction. And, the realization that the laser you built is a complex precision device you were able to bring to fruition successfully.

    To the amateur scientist, experimenter, hobbyist, or even weekend tinkerer, there is something about the idea of actually creating a working laser that makes its construction from the ground up a very attractive and rewarding project with "first light" - the instant that your home-built laser first emits a coherent beam - approaching the excitement of a religious or sexual experience!

    Although the laser is a device or instrument based on fundamental quantum mechanics which is very simple in principle - an excited medium between mirrors - building one successfully may require mastering several disciplines not normally found in even the high tech home. These include: glass working, vacuum systems, gas handling, high voltage electronics, and precision mechanical fabrication. Dealing with these can in itself be an excellent educational experience. Access to a university or industrial lab will make things a lot easier but isn't essential - it is possible to build a laser without outside assistance. Academic studies in laser physics or related subjects are also not necessary unless you want to attempt to do serious research as all the lasers that can be reasonably constructed at home are based on well established principles where rules-of-thumb and simple calculations will suffice. However, the cost of such an undertaking can be significant - experience in scrounging is a definite asset! And, the construction of home-built lasers can be quite addictive and may impact other activities like social interaction, eating, sleeping, and the timely performance of other bodily functions. :)

    In this chapter and the one that follows, we provide basic information on the construction of various types of lasers from scratch including: home-built laser safety, setting up a home laser lab, sources of supplies and chemicals, vacuum systems, glass working, structural materials, power supplies, and more.

    Then, a variety of specific types of home-built lasers are described in more detail. Much of this material is derived from the Scientific American collection "Light and its Uses" [5] and from the email, Web sites, articles, and experiences of those who have been successful in building their own lasers from basic components and getting them to work (not taking the easy way out and using commercial tubes or laser diodes!) - or have given it their best shot trying!

    While this will not substitute the hands-on of actually having built one of these lasers or detailed construction plans, it may provide the spark to get you started!

    Reasons NOT TO Build a Laser from Scratch

    First, let us consider some ill-posed justifications for attempting to build a laser from (almost) raw materials:

    If these are your only reasons for wanting to do this, you will rapidly tire of the endeavor and the parts will end up in a box alongside that dusty old partially ground telescope mirror you also never completed. :-(

    If you want a working laser for a particular application, save your pennies and buy one. The cost of a used laser appropriate for what you have in mind may not be as terrible as you may think. The result of building a laser from scratch isn't likely to be something you can use reliably day in and day out without constant maintenance, repairs, and the occasional disaster. (The one exception to this might be the axial flow CO2 laser which if properly constructed, is less finicky than the other types discussed in the following chapters.) A system that starts life on and under a workbench will also probably never be packaged in a nice self-contained cabinet and may have to coexist with the home washer-drier, family car, or kitchen table. :) Anything home-built is also going to have many potentially serious hazards associated with it unless significant effort has been made to provide the necessary beam blocks, electrical and thermal protection devices, and safety interlocks.

    Reasons TO Build a Laser from Scratch

    However, there are many justifications for embarking on an adventure of this type:

    And, there is something to be said for being able to claim to have built the only working laser from scratch on your block - or more likely, your town, city, state, or country!

    Experiences of People Who Have Built Lasers From Scratch

    Here are a few of the (mostly) success stories. The fact that this list is so short may be some testimony to the difficulties involved! The first is from one of the original authors of the SciAm Amateur Scientist column, who recently contacted me via email:

    (From: Syl Heumann (

    I am the author of the original SciAm article on the design of the pulsed Argon laser. I did another article for SciAm Amateur Scientist - a year or two later. It was how to make a hologram.

    I got to know Red Stong, the Amateur Scientist editor quite well - visited him at his home on Staten Island.

    My first laser was a HeNe from the original SciAm article in about 1964 or 1965, It was raw AC powered and had a life of about 15 minutes! Then I went to RF power and that was better, till I went to DC. I had met some of the guys from Spectra-Physics and they were very helpful. Then I did a bunch of HeNe lasers - my best one did 16 mW.

    They gave me the original notes on the first mercury vapor laser and I had one of those going in the next few days.

    I guess my argon (also krypton) ion lasers were about 1966 or so.

    When I did the argon laser, and looked at the light dispersed by a prism, there was an extra line in yellow. I thought I had discovered a new transition - but after some research, it turned out that I had a leak - it was nitrogen! But it was exciting.

    I no longer have a high vacuum system, so don't mess with lasers any more. But please check out My Web Page for some other hobby interests.

    (From: Mark Wilson (

    I was born in a very small town in Idaho. I was fascinated by physics and laser technology. When I saw the Scientific American article on building your own Helium-Neon (HeNe) laser, I decided that I wanted to build it. The Scientific American HeNe laser was extremely difficult to build and I could not have done it without a lot of help. I got Spectra-Physics to donate a set of laser mirrors to me, a glass shop in my home town to help me cut and grind the Brewster window angles on the tube. The tube was made from lead glass from a sign company, and I also used neon sign electrodes. The optical rail was a 3 foot long piece of 2"x6" extruded aluminum that I got from a glass company which used this material to make doors for commercial buildings.

    I followed the directions in the Scientific American article to the letter. I sealed the microscope slides to the glass tube using flexible colodian that I got from a pharmacy. I filled the tube at a sign company which had He, Ne, Ar and other gases on a glass manifold. I assembled the laser and made the power supply using a mercury rectifier tube and a neon sign transformer. I got the tube to lase for a brief time, but since it was not a hard-sealed tube it quickly died probably due to helium diffusion. The tube would light up but not lase for a while then that too stopped.

    I then made a dye laser, again from a Scientific American article. This was much easier, and the materials were much easier to obtain. I did not need any wavelength selective mirrors, vacuum system, or high voltage supplies. I ordered samples of a couple of laser dyes (Sodium Fluorescene, Rodium 6G (possible spelled wrong)), and mixed up dye samples which were flowed through a piece of quartz tubing. A flash lamp was located at one focus of an elliptical reflector and the dye tube was at the other focus. The reflector was a juice can that I polished up as per the article. The flash lamp power supply was very simple which put several hundred volts across a large capacitor, and then tripped the flash with a tickler coil. This laser was easy to build, and actually worked for quite a while, but I couldn't set anything on fire with it.

    I then built a flowing gas CO2 laser again using glass and equipment from a neon sign company. The Brewster windows were 2 near perfect salt crystals that I got for a salt company in near by Salt Lake City. I had to keep the windows inside plastic bags a moisture absorbing material when the laser was not in use to keep the windows clear. I made my mirrors from round glass blanks that I got from a local eye doctor. He ground me a set with a -1/8th diopter (-8 meter focal length). I then drilled a small hole in one of the glass blanks to allow the output beam to escape. I coated the mirrors with gold using a sputtering machine that I built, again from a Scientific American article. I assembled the tube and mirrors onto a extruded aluminum base, and then connected it to a vacuum pump. This pump was two Fridgidare compressors connected in series with each other and a cold trap. Later I replaced the pump with a 2 stage Cenco Hyvac pump that my friends at the sign company donated to me. I made a gas manifold including a vacuum gauge, to mix gases for the laser. I got a cylinder of CO2 (used in pop machines), and also a small cylinder of Helium, bubbled the gases through water to add water vapor and then flowed this mixture through the tube. I used a center tapped neon transformer and a set of solid state rectifiers to make a DC power supply to run the tube, and with an input of 150 watts of power, I calculated that I got about 5 watts of IR power at 10.6 um. I could burn holes in things so I was finally happy.

    (From: Steve1W1 (

    I built my first C02 in 8th grade for our school's science fair with plans from "Roy Davis Laboratories" (if anyone is old enough to remember those days). It took about a year's worth of work and scrounging, but it worked so such a project CAN be done.

    (From: Steve Roberts (

    I tried to build several lasers when I was a kid, dye worked, nitrogen worked, argon failed. But, I found making Brewster angle windows nearly impossible. The guy at the local refurb shop told me I went about it wrong, all I really needed was a belt sander and some abrasive, the brass bladed sawing technique in "Light and its Uses" doesn't work. I have also repumped a commercial laser tube or two with mediocre results.

    (From: Keith (

    I have built a variety of lasers from scratch.

    I have a tube blown by my friend, with microscope cover slips sealed on using Torr-seal epoxy I made in 1988. Its filled with an 80/20 mix of helium and neon, and is about three feet long. It uses several neon sign electrodes. Its large diameter (compared to 1 mm bore) gives it a TEM11 output. With periodic gas fills, it still runs today, though I don't mess with it much.

    I have made very small argon lasers from scratch, but the cathodes are quite large, and do look just like early he-ne lasers. The power supplies are very different, however. I have an original "tube", not solid state, pulsed argon laser power supply from bell labs, 208 VAC, from the era of the "Telephone pole theory of laser action": Hit a telephone pole with enough energy and it will lase. :) Hundreds of amps per pulse, into a large diameter tube, before CW capillary tube type ion lasers.

    I've also built an more unusual laser using the element thallium. I'm sure you know that thallium is named after its strong green spectral line. It just happens that it has a very low melting/vapor point, and like mercury needs no buffer gas to lase, it just needs a vacuum. It is not very critical of vapor pressure as in the HeCd and HeSe. The use of argon or neon as a buffer does make it more efficient but is not absolutely needed. The mirrors from a dead green HeNe laser tube work fine (as long as their radius of curvature is consistent with stability for the resonator).

    (From: Thomas Rapp (

    My interest in lasers started very early when I was a young boy and read about the invention of the ruby laser in a popular German magazine (Stern). But of course, I who tried to build radios from old parts found in the city dump, was not able to get either the materials nor the knowledge to build a laser. However, years later, when I started to work as a technican at the Munich University, I obtained an old HeNe Laser which didn't work anymore because of gas cleanup. When I unsuccessfully tried to reanimate the tube one older scientist told me about the article in Scientific American about an easy to build dye laser. Some weeks later, after a small fire hazard because of blowing up the flashlamp and ignition of the methanol dye solution, my first laser really worked. After that I began to read all about laser mostly in "Review of Scientific Instruments", "Applied Physics Letters", and "Journal of Scientific Instruments". After building a couple of nitrogen lasers, I changed to copper, manganese, and lead vapour, and then I built a CW CO2 laser. After that my interest changed to other fields of physics, building all kinds of vacuum devices, like a quadrapole mass spectrometers and even a small scanning electron microscope, microwave devices like a home-brew radio telescope. And of course, I was working in the rapidly growing computer area. But all sorts of pulse power devices and their application in lasers are still my favourites. Late in 2002 I got the job to construct a nitrogen laser for a guy I got to know via eBay. After that I decided to build lasers again, create my Pulslaser Web Site to share my knowledge and experiences with other laser enthusiasts.

    (From: Sam.)

    Hopefully more to follow. :)

    Flavio's Comments on Amateur Laser Construction

    (From: Flavio Spedalieri (

    The reason for building a laser from scratch, is to learn how lasers work through physical hands on construction. Also, one major feature that you have with home-built lasers, is the unlimited freedom of experimentation, you can control many variables like; power (voltage and currents), gas, optics, materials etc, which otherwise is very difficult or impossible with commercial lasers.

    Another rule that I have set when it comes to building lasers: Build the lasers that are more expensive, exotic, and least obtainable like:

    In the list above, you may notice that I have NOT included typical lasers like Helium Neon, and Argon Lasers, the reason being that HeNe lasers are now too cheap to even consider building, and are all the too common. I have not limited myself to not experimenting at all; I do have a HeNe laser tube that I, one day would like to cut the vacuum nipple off, and connect the tube to a vacuum system, and back-fill with HeNe, but the chance of getting it to lase is quit small, due to the fact of the need of very critical gas mixture and purity, and a very good vacuum system.

    CW argon ion lasers, on the other hand, require huge amounts of current, and the necessary materials to build the tube are not easy to work, toxic, or both (e.g., beryllium oxide, tungsten). Also, glass work is a major component with the need of correctly angled Brewster windows. Today, like the HeNe lasers, argon lasers can be obtained with ease, and relatively cheaply. Argons, also require very good vacuum systems as well as very pure and critical gas pressures.

    So, as a summary, and as a rule of thumb please take the following in mind:

    1. If you are building a laser for experimentation, learning/educational purposes, then all the better, as your learning curve will be very steep.

    2. If you are intending to build a laser purely for a 'workhorse' application like wood or metal cutting, especially if for a commercial venture, it might be more economical and much easer to purchase a second-hand laser system from the surplus market.

    3. DO NOT expect great powerful beams of light outputting from your laser, and you will be shortly setting yourself up for a disappointing downfall. As mentioned, home-brewed lasers will require much experimenting around, you may be lucky enough to have the laser produce an output at first go, but also be prepared that you MAY NOT get an output at all.

    4. Use the 'KISS' Method - "Keep It Simple, Stupid". In doing so, you can reduce the number of areas that can cause problems. For example, the electrodes of a CO2 laser can just be copper pipe fittings (one at each end of the tube). There is no need for neon sign electrodes which require additional assembly and possible glass work - and can possibly fail.

    5. Keep you laser tube within reasonable lengths. At the very least for your first laser. Build your first laser, get this running and lasing, conduct experiments and chart results within a spreadsheet or something. voltage/current versus gas pressure/flow rate/mixture, etc. Once you have succeeded in building your first working laser, then you can move on, and try building a bigger laser, at least you will not be disappointed if the laser does not work as you have already built a nice small working model.
    I hope that this will bring some reality into your projects, but please don't interpret these comments as discouragement from building lasers. I'm actually trying to create more of a challenge for anyone who is or will be embarking on this wonderful area of laser technology, yet to keep in mind that you may not have a working product at first - a little like trying to build a tall building, expecting it to stay up without the foundations.

    Diane's Home-Built Laser Experiences - The Beam and I

    Only rarely do I receive email demonstrating true enthusiasm and determination for *anything*, let alone laser building, from an early age. Here is one I had to include. Diane now has a Web site which in addition to the description below, has some nice photos to go along with it.

    (From/by: Diane Neisius (

    The Beam and I

    Do you still remember the first Star Trek series? Ah, that was something for a child's heart. Ha, there *was* a woman in a starship (so don't tell me a girl can't fly to space silly boy you!!!), and they had these "phasers" raying around during their adventures. Nevertheless, I was a big girl (10) and knew, it was a TV series. A kind of technical fairy tale. The more I was surprised when I found a popular science book at the local city library. There were guys who made beams - really, not on TV. I didn't understand much of the stuff described, but got that there were some sort of mirrors and a ruby in it. They called it L.A.S.E.R., and it was real. For the fact my grandpa could make for himself *everything* (he repaired all the electric and mechanic stuff for our family, even hopeless cases), I believed he (and also me) could make our own laser if we just had one of these expensive rubys. I well knew a ruby was a high-priced gem. When I asked grandpa, he told me something about precision and that it is not quite easy to reach this in the living room of a hobbyist. Ok, I could not buy a ruby from my spare money. But from those days on I was convinced one day I *will* have my own laser. One day.

    Childhood dreams came to an end. No, to be honest, they only slept until I was in the final High School classes. It was the beginning 80s, and being a frequent reader of the German issue of Scientific American, one day I found the famous description of the mercury laser in it. Huh? A guy somewhere out there built his OWN laser? The fever came back again. When he could do it, I also can do it, I decided. However, I knew a lot more about physics than before, also liked to spend time with the school's little HeNe laser (on rare occasions). Got a basic knowledge of what is important for lasers to become working. A few telephone calls made me quite unhappy. I had learned you can get all the stuff you need if you really want - and can pay the prices. HIGH prices. Again there was the "ruby problem": lasers are *expensive*. "Silly, one cannot have one's own laser", a fellow laughed about me, "that's only for laboratories." I better not tell him about my desire. So I thought: "One cannot? Let's see about that."

    Don't ask me today why I started to study mathematics. We had a quite famous Department of Quantum Optics on our university, perhaps I should have gone to the laser business. But those days computers were still more exciting than lasers to me, and that's it. Now, being a student, I had access to real scientific literature, and I learned a lot about the theoretics of lasers (and, reading the business laser magazines, also about technical realisations). I studied various descriptions of laser types and decided to try about a flashlamp pumped dye laser. This one at least needed no expensive ruby. :) To make a sad story short, I learned a lot about how to blast stroboscope tubes using a voltage doubler and really BIG capacitors, and I guess the carpet in my room at the student's community will still have these nice pink rhodamine spots. :( I gave up on dyes. But I still wanted a laser, and some more telephone calls brought a fine small Siemens HeNe tube to me (it was a new 1.5 mW LGR7621). Shall I say I heard some of the well known "brzzzz's" from self-wound transformers until I looked for a used neon transformer? Still a descent of my grandpa I was... So I spent my time to build a casing for the transformer and a rectifier out of a box of 1N4007's (never tell this an engineer student. I did, the poor boy almost got a heart attack). By the way, the LGR7621 is quite robust. One day I caused a short which *detonated* the anode resistor (another lesson about BIG capacitors). After my eyes had recovered from the resulting supernova inside the tube, I anxiously looked if there's still anything alive. A visual inspection showed up a lot of very small cracks along the inside of the bore. Replaced the mortal remains of the blown resistor by a fresh one, it still started and lased! I used it for years after this accident without significant drop. Brave little HeNe. :)

    After some time, HeNe became boring to me. Using an internal mirror tube is one task, to build a device like Scientific's mercury completely on your own, quite another. So I started doing some experiments in that direction for the next time. I got a simple vacuum pump and an unsealed neon tube and began to work with glow discharges. To keep it short, I had to perform lots of experiments to learn about vacuum, outgassing and purity of gases. Over the years, my little laboratory grew: a self-made voltage doubler for the neon transformer, a self-made mercury vacuum gauge, a better pump, noble gases in liter bottles, a hand-held spectroscope, longer discharge tubes. To work, all this took years of learning by doing. Then the next strike came, of course again by Scientific American. It was the copper vapor laser. How exciting... For I knew very well I had no experiences about laser optics until now - but the superradiant copper lines would need none. In words, I could start immediately. But unlike the old days, I decided to study a bit about superradiant laser before blindly begin to "hammer and saw". And by this I found out about the still simpler N2 laser. I decided to have one. And in 91, a self-made 10 cm test tube of acrylic with quartz windows lased! I danced in my room, for after 10 years I finally got it!

    For the fact the 10 cm tube lased quite weak I studied more about N2 laser design. The major problem was, I could reach only 10 kV with my equipment thus having poor energy densities in the discharge. Longer self-made tubes also lased weaker than expected (I tried several), even if attaching a metal mirror to one end. I had to pinch the discharge somewhat, but doing this in a tube made of acrylic would easily overheat and smoke the walls. Those days I already worked toward my Ph.D. thesis, and the research center where I did it had the most precious thing I ever saw in a library: the COMPLETE set of the Review of Scientific Instruments! Complete means complete: from issue #1 of 1929. And after some hours with it, I found the most useful paper about N2 type gas lasers I know. It is about the "strip line" type laser using a segmented discharge bore originally designed for the UV lines of the hydrogen laser [1]. I adopted the design a bit to my power supply (shorter strip-lines, shorter discharge segments) and pinched the discharge even more by adding short pyrex capillaries inside the segments. Believe me, it was a *lot* of work to drill holes for 56 electrodes and fill in the pyrex tubes successively from the end of the outer acrylic tube. But finally I had a 3 mm x 80 cm bore with a nice high energy density. For its strange appearance, I baptised it the "German Flute". The first thing I noticed after the first tests was, this baby would lase with every gas containing a bit nitrogen, even dirty air. :) Mirrors were good for nothing, and I easily got all three UV lines on a fluorescent screen using a "water prism" (triangle pot glued together from thin pyrex pieces and filled with water, which absorbs UV much less than a massive pyrex prism). Of course I also tried other gases, and the green superradiant Ne line at 540.1 nm was strong in this tube, too. Over the time it became my favourite. And on few occasions, after long times of green Ne, it was also possible to get the much weaker orange line at 614.3 nm at a lower pressure. But normally it disappeared after some time from outgassing impurities. The still weaker yellow line I never caught.

    Impurities were what finally drove me tired. It was common for my "German Flute" to be run with gas through flow. Otherwise lasing stayed only for seconds. On one occasion I tried a bore cleaning via He bombardment which took several hours. But after finish, what I call the "dirt spectrum" (N2 band, H alpha line plus Hg lines -- mercury from the vacuum gauge) reappeared in half an hour. Whenever I liked to start my laser, I had to spend days and hours in front just for basic cleaning the vacuum devices. And, the cost for the needed constant flow of Ne burned a hole in my pocket. Pure lab-grade neon isn't that cheap. And then, a few years later, I got a pen-sized red diode laser which made roughly twice the output power of my home-built in the green. It was depressing.

    Sometimes in life one has crises and has to separate from several things. So it was for me, and it was such a crisis which made me giving away a lot of things - including all my lasers. The university didn't take them (security reasons of course), but I found a physicist collecting strange devices, and I hope my baby still has a home there (even if it doesn't lase any more). So, off I were.

    Yes, until now. Some weeks ago I visited some friends, and I wondered to see a yellow HeNe in their rooms. They told me to use it for illumination of large naturally grown quartz crystals, for they like the "golden shattered glow" in them doing so. But they knew less to nothing about lasers, and I talked about lasers for an hour or so. And thought about I also would like illuminated crystals. But not red or yellow - green and blue, perhaps from an argon ion laser it would have to be. In the days of internet it is an easy task to feed "argon laser" to and see what happens. Of course it leads to Sam's Laser FAQ and to lots of surplus advertisements. I have the chance to get an "all-included" ALC 60X (head, fan, cable, power supply, tube refilled) for roughly $1,000. Goddess, I have to scratch off those bucks somehow...

    The beam has me back! :)

    1. Kirkland, Dogett, Kim: Vacuum-UV H2-laser excited by a traveling-wave discharge, Rev. Sci. Inst. 52(1981) p.1338.

    I bet you're sorry you gave all that stuff away, even if you do know how to do much of it better now. A 60X is a nice laser but not quite the same as something built from scratch that one can fondle and tweak! :)

    (From: Diane.)

    What did you do... Talking about fondling and tweaking my home-built laser... Asking if I regret to give it away...?? Ah, sigh... :) Finally I phoned the guy I gave all the equipment and he told me I CAN HAVE IT BACK! Now, I guess I will have to do lots of maintenance work on my baby if it's home again (stored at a garage for the past 3 years).

    So I guess the beam really has me back! :)

    Building a Femtosecond Laser at Home?

    Well, probably not. But after you have constructed all the SciAm and other "common" home-built lasers, it could be something to keep you occupied. :)

    (From: Anonymous (

    Info on how to build a femtosecond laser? I have had a rather strong academic interest in the field, if nothing else, on this subject for the last 10 years or so. never in that time have I seen a 'how to' build such a laser. if you would like to see what others have done the academic journal 'Optics Lasers' is your best bet, any major university library should subscribe to them, and certain articles are available via their on line archive at the Optical Society of America Web site.

    If on the other hand you are simply looking for background information on mode locked lasers, nowadays most systems (or at least the easiest to build systems) use a self starting mechanism and only really need:

    1. Gain medium with sufficiently high bandwidth to support the pulses you are interested in - Ti:Sapphire and dyes are the old standbys, but there are all sorts of different hosts that will produce sub-picosecond pulses.

    2. A system for achieving GVD compensation. For a long time, group velocity distribution has been accomplished by a pair of prisms tailor made with 'x' apex angle and 'y' refractive index. However, in recent years it ahs been possible to use multi stack dielectric mirrors for GVD compensation. If I'm not mistaken this was first done by Newport (or perhaps they had the first commercial mirrors, I don't remember which).

    Some Photos of Home-Built Lasers

    (From: Chris Chagaris ( (From: Laserist ( Also check out the links in the sections: General Resources for Amateur Laser Construction or Amateur Laser Construction Sites.

    Diagrams Showing Major Components of Typical Home-Built Lasers

    These drawings show the structure and power supplies for some of the lasers built by amateurs. The first seven are based on the laser articles from Scientific American (including the book: "Light and its Uses" - see the section: On-Line Access to the Scientific American Laser Articles Their purpose is to give you a flavor of what this type of laser construction entails - but are NOT intended as dimensioned plans and are NOT drawn to scale. Refer to the more detailed descriptions in the chapters on each laser type following the introductory chapter: Home-Built Laser Types, Information, and Links and the relevant Scientific American articles. The final one is from the journal: Review of Scientific Instruments. See the chapter: Home-Built Pulsed Multiple Gas (PMG) Laser for more information. There may be additional diagrams in each of the chapters on specific home-built lasers, particularly those constructed by various contributors to this document.

    Comments on a Universal Experimenter's Gas Laser

    A question that comes up occasionally is: "How can I build a laser that I can use to try out various gases and other parameters?". Here are some suggestions for a gas laser testbed optimized for visible and near-IR operation. Actually, these are more like random thoughts to get you started:

  • Build your plasma tube with Brewster windows at both ends on stems that are at least 2 or 3 inches long to keep them away from the discharge. For a wide wavelength range, putting these on ball and socket or flexible mounts may be desirable to permit their angle to be varied slightly.

  • Provide electrodes suitable for your expected types of lasers. The positive (anode) electrode can usually just be a wire or sleeve (cooling and sputtering aren't significant). However, the negative electrode should be made of a suitable material (e.g., aluminum, heated tungsten, etc.) for the laser operation - pulsed or CW, low or high current, etc. Place the cathode(s) in a side-arm so prevent sputtered material from getting to the Brewster windows.

  • Build universal mirror mounts that can accept a variety of mirrors - you won't always be able to find the same size or thickness. I would suggest something like the mirror cell arrangement described in the section: Mounting Laser Mirrors.

  • Install a Helium-Neon (HeNe) alignment laser as a permanent part of your testbed. Mirror alignment is one of those things you will be doing constantly. It will be desirable to be able to do initial set up and checking without having to assemble the alignment jig every time!

  • Start with broadband HR mirrors having as high a reflectivity as you can get or afford over the wavelength range of interest. The HRs used in HeNe lasers may be much better than 99.9 percent. (If an 'other color' home-built HeNe laser is in your plans, particularly one that does green, even the OC reflectivity will have to approach these values!) The higher the reflectivity, the lower your lasing threshold. The radius of curvature (r) should be at least equal to the distance between the mirrors (L) but probably not more than 2*L since if the focal lengths (f = r/2) are too long, alignment becomes more difficult.

    As noted below, some HRs are not polished on their rear surface. It may be possible to attach an optical flat (e.g., piece of a good quality microscope slide) with optical cement or Epoxy to reduce scattering and reflections from that surface but this won't be ideal.

  • For initial experiments, take the output beam off the Brewster reflections or from the leakage through the (HR) mirrors. Once you know which line(s) you want, a specific OC mirror spectral curve and reflectivity can be selected. Reading through the chapters on each of the types of home-built lasers that follow should provide some of the details. Start with the chapter: Home-Built Pulsed Multiple Gas (PMG) Laser which deals with an approach along these lines.

    (From: Bob.)

    Also, take note that most HR mirrors are a lot better than 'just' 99% reflectivity, at least when you are talking about intracavity mirrors. Also a lot of commercially available HR mirrors are not designed for laser use - they only have their back surface fine ground, not polished. This means you can't get a HeNe laser beam through them for alignment purposes (particularly optics from Thor Labs and CVI). Another reason why generic broadband lasers may not be suitable, is that if you have any plans on making a high power pulsed laser (e.g., ruby or YAG), most off the shelf optics can not cope with the power/energy levels you would be exposing it to, and will quickly fail. Just a few things to watch out for!

    What About a Home-Built Solid State Laser?

    Note that there is currently no mention here (or as far as I know, in the Amateur Scientist articles of Scientific American) of ruby, YAG, vanadate, and other solid state lasers. However, there is a chapter: Home-Built Pulsed Solid State (PSS) Laser which is under construction and another one: Home-Built Diode Pumped Solid State (DPSS) Laser which is already fairly well along.

    Since there is no realistic possibility of actually growing, shaping, grinding, and polishing a raw laser crystal in your basement, there are, never were, and never will likely be any truly built-from-scratch SS lasers. You will have to buy the crystals ready-made. And, since there is less standardization on SS laser components than for many other types of lasers, it isn't even possible to suggest many sources for parts with particular specifications unless you are willing to pay new (and very high) prices. That's the bad news. The good news is that SS lasers are much easier to get working even with a less than optimal match between the lasing medium, pump source, and mirrors, than many other types of lasers. While the gain of a HeNe laser may be 10 percent per meter, the gain of a solid state laser rod with flashlamp pumping may be 10 percent per cm! (And when diode pumped, the gain is much higher still.)

    Pulsed (flashlamp pumped) SS lasers have been popular projects since the invention of the laser and with reasonable care, a successful outcome is likely. They are by far, the easiest lasers to construct capable of blasting holes in things. :) There are many surplus components and partial or complete systems available at reasonable cost. As companies switch over to Diode Pumped Solid State (DPSS) lasers from lamp pumped types, more and more pulsed SS laser components and systems are showing up on the surplus market.

    Building a DPSS laser, especially one with frequency doubling to produce green (532 nm) output is a more complex - likely much more expensive - undertaking, but one that can be accomplished successfully. And, because DPSS lasers are becoming more popular, components for these are coming down in price, at least somewhat. So, there is hope. :)

    Home-Built X-Ray Laser?

    The following is from a recent paper: "Generation of millijoule-level soft-x-ray laser pulses at a 4-Hz repetition rate in a highly saturated tabletop capillary discharge amplifier", C. D. Macchietto, B. R. Renware, and J. J. Rocca, Optics Letters, vol. 24, no. 16, pp, 1115-1117, August, 1999.

    (Portions from: Bob.)

    There are a good number of people who have built or are currently building their own lasers. from simple systems, to the extravagant. I was going through some current journals today, as things were slow here, and I saw something that caught my eye, making me instantly think of this group: A table top soft X-ray laser.

    Basically, this system was quite simple. It had a 0.32 mm ID aluminum oxide capillary evacuated to roughly half a Torr filled with argon gas, pre-ionize by a discharge. It was pulsed with a high current pulse (approximately 26 kiloamps!) with a fast (40 ns) rise time using a water capacitor and series spark gap switch. The water served as both the dielectric of the capacitor and as cooling for the capillary. The capacitor was charged by a 4 stage Marx generator located in a separate box. The laser itself occupied an area of only about .4 x 1 m (16 x 40 inches). Since the laser operates in a highly saturated regime, no cavity optics are required. The output beam profile had a ring shape (due to plasma density gradients in the plasma column) with a half-angle divergence of about 4.6 mR. The output energy averaged about .88 mJ at 4 Hertz.

    I (Bob) kinda like the idea of having an X-ray laser in the corner of my lab. So I'm gonna build one, though maybe others might like to as well. For your information, when the NOVA laser was used to pump a soft X-ray laser, they got out 8 mJ pulsed, with a repetition rate of about 1 pulse per 30 minutes (!!) or so. It would be kinda cool to have a 1 mJ X-ray laser, especially if it could operate at 4 or 5 pps. And, with higher average power than the NOVA laser pumped X-ray laser!!!!! (Although I must admit that one generated shorter wavelength X-rays: 15 nm instead of 47 nm. Well, you can't have everything.)

    General Resources for Amateur Laser Construction

    There are actually a larger number of places than you might think to find information on home-built lasers as well as some ways of interacting with like-minded individuals on-line. Check out the amateur laser construction Web sites for examples of lasers others have built, or are in the processing of building, as well as much related information. Many of these sites have descriptions, diagrams, and photos of their home-built lasers.

    There are also a number of companies that may sell complete plans, parts, and other items related to home-built lasers:

    Scrounger of the Month Award

    Here is a success story on obtaining help in glasswork, AND inexpensive neon sign transformers and vacuum pumps from the same source - you guessed it - a neon sign shop! It seems that many neon sign types are also interested in lasers (or at least fantasize about building one) and are therefore sympathetic to the needs of amateur laser constructors! For more info, see the section: Tips for Dealing with a Neon Sign Shop.

    (From: Tom Miller (

    Ok, today was GOOD! I visited a local sign shop, one of the larger ones, and got to talk to the owner. He was the one who initially nixed selling any transformers to the general public. After a brief discussion explaining what I wanted to do, he took me on a tour of the whole facility. It seems he may be interested in constructing a laser also. He was checking prices for CO2 lasers for use in cutting plastic sign material and for a 50 watt unit, was seeing prices in the $50,000 range. Now granted, this included the mechanical positioning equipment, but still, he thought it to be too high.

    Anyway, I left him a copy of Sam's Laser FAQ and my business card. He wants me to give him a drawing of the glasswork and he will put it all together. Says it will take less than an hour, so I told him I would pay for his time. He suggested that I use a "tubated" electrode on each end and connect the vacuum pump to one end and the gas supply to the other. This way, the flowing gas will cool the electrodes.

    We got around to talking transformers and I ask what a 15 kV, 30 mA unit would cost. He asks if I could use a 15 kV, 60 mA transformer if it was used. I was completely surprised when he said it would only cost $10. This transformer was sitting on his glasswork bench and he was using it to test tubes. The guy showed me a pile of neon sign transformers under a very large workbench. Must have been 20 to 30 there. Also, the owner told me he had MANY old vacuum pumps just sitting around. I saw at least 5 of them.

    Next, he asked what I would mount the laser on. I said I would like to use an aluminum I-beam about 3 to 4 inches wide and about 4 feet long. He took me to a different area of the shop and found a scrap piece of the stuff. I figured that I would stop using all of his time, and told him I would go and make a drawing for the tube and get it to him sometime in the next few weeks.

    So, anyone having problems, just load up with Sam's Laser FAQ, go find a large old neon shop and talk to the guys who actually do the work. You will be surprised how much interest they will have in a good powerful laser.

    So today was a good one. :)


    Information from many sources has been used to compile the chapters on amateur laser construction. Wherever possible, I have attempted to identify the individual contributor. However, if you feel that there is something here you wrote without an acknowledgement, please let me know.

  • Back to Amateur Laser Construction Sub-Table of Contents.

    Setting up a Home Laser Lab

    Safety Issues in a Lab for Home-Built Lasers

    There are a variety of issues that are important for any sort of home lab or workshop but the following, in particular, apply directly to lasers and laser construction:

    There didn't appear to be a critical mass of lawyers present at the time most of the articles in "Light and its Uses" were written. Therefore, they tend not to deal with the safety issues as emphatically as might be desired. Most of these projects have aspects (most often the high voltage power supplies) that are potentially dangerous or lethal. Safety must be at the top of your list of priorities when undertaking such an endeavor!

    Work Area - Setting up a Laser Lab

    Since any of these lasers represents a long term comittment, it is essential that an area be set aside for your laser lab. Therefore, the kitchen or dining room table is NOT an appropriate place to be constructing a laser. It is possible to do without the sort of setup depicted in the section: Possible Laser Lab Layout but there are some basic requirements for a safe, functional, and convenient space:

    Possible Laser Lab Layout

    I wish I had this! Note: Two means of exit and two fire extinguishers!

    Also note the chair - most important - and the bench for your guest (though probably should be s eleep-sofa so they can snooze while you spend the afternoon adjusting your gas mixture or performing mirror alignment. :-)

        |<------------------------------- 12' ------------------------------>|
      ^ |    |                                                         |     |
      | |    |        Storage Cabinets/Shelves (above work area)       |     |
      | |    '---------------------------------------------------------'     |
      | |          Electrical Outlets (two circuits) all along wall          |
      | |                                                                    |
      | |            Work Surface - thick hard-plywood (3' x 12')            |
      | |____________________________________________________________________|
      | |           |                                                        |
      | |           |  Vacuum System on floor (beneath work area)     Gas    | 
      | |           |                                              Cylinders |
      | | Test      |                                              __________|
      | | Equip.,   |                    ________                 |          |
      | | Power     |                   |        |                | Wet area |
        | Supplies, |                   | Office |                | Glass-   |
     10'| Misc.     |                  (| Chair  |)               |  working |
        |           |                   |________|                | Ventila- |
      | |           |                   '--------'                |  tion    |
      | |___________|                                             |__________|
      | |                                                                    |
      | |S Power Switch            _ _        __________                     |
      | |(on Wall)          .-======'======-.|          |                    |
      | |        /          |               ||  Bench   |           \        |
      | |      /       Fire |    Storage    ||          | Fire        \      |
      | |    /         Ext. |               ||==========| Ext.          \    |
     _v_|__/         _______|_______________|____________________         \__|

    Sources of Special Parts and Supplies

    This section deals mostly with the items to equip your lab, small parts, chemicals, and so forth. Also see the sections on vacuum equipment, optics, and power supply components, in this and the chapter that follows, as well as the chapter Laser and Parts Sources. And note that many of these items appear regularly on eBay and other auction sites as well as high-tech flea markets. Apparently, at least one person is even selling small quantities of chemicals to the public on eBay!

    Develop a relationship with a teacher/instructor/professor/researcher at a high school/technical school/college/university/industrial lab. Some people will be more than eager to help and mentor you - even to the extent of loaning equipment or donating small quantities of chemicals, electronic components, hard to find optics, etc., to your cause. Use of their lab may even be possible. And, universities sometimes toss out the most amazing things - like complete vacuum systems - when a grant runs out and they need the space! There are various programs as well to encourage students to go into science and technology fields. Who knows, they may even pay you to do this!

    Call up laser and optics manufacturers. Sure, many won't give you the time of day unless they think you will be ordering $1,000,000 worth of equipment. But, all you need is one to say yes! There are always such things as cosmetic rejects or seconds - that are useless to them because they cannot sell the parts - but fine for your needs. The trick is to hold their attention long enough - or be such a (polite) pain in the neck that the easy way out is for the company to provide what you want! I have heard of people obtaining all sorts of material, parts, equipment - some of it worth quote a lot of money - in this manner.

    In summary - possible places to find useful stuff:

    (From: Chris Chagaris (

    Here are some resources that I have not seen mentioned anywhere on the Net:

    For chemicals used in various aspects of laser construction and laboratory glassware at unbeatable prices, a fine source is:

    For quartz tubing and quartz windows of all sizes, at very good prices: I would be glad in assisting other individuals in locating some of the more difficult to procure items needed in some aspects of constructing these various lasers.

    (Portions from: Steve Roberts (

    While Sargent-Welch, Edmund Scientific and Dunniway might be what come to mind when thinking scientific suppliers, they are most expensive, expensive, and not cheap, in that order. :-)

    The ideal thing to have is the Laser Focus World (LFW) Buyers Guide, a phone book sized list of suppliers put out by Laser Focus World for their subscribers. Subscriptions are free to qualified individuals, so you need a company name to subscribe. If somebody is in the laser business, they are in LFW. Photonics Spectra is also a good freebie if you can qualify.

    If you are into building your own HeNe (or other) laser from the ground up, these suppliers may come in handy:

    Glass and glass working equipment suppliers: Flanges, glass-to-metal seals, electrode material:

    For cathode material, just ask for what the machinists call "gummy" aluminum, the really hard to work soft stuff that gums up tooling, and you've got it. Not 905 or 2025, maybe 6061, but 6061 has a lot of weird stuff in it like silicon monoxide. One guy I watched once sputtered the cathode in a oxygen discharge, then cleaned it with hydrogen in a soft glow. Nickel is the metal of choice for anode pins and wiring inside the tube, but its going to take some work to spot weld it to the kovar or dumet lead throughs, you need to find or make a "thermocouple" welder designed for cap discharge welding of small wires.

    You can get pure aluminum, nickel wire, tungsten wire, titanium sheet, etc. from Small Parts, Inc. (Miami Lakes, Florida, 1-800-220-4242). They sell it chopped up in small quantities at decent prices, with no minimum order. They specialize in what they call "Engineering Findings", in other words, small quantities of all the goofy parts you need to make industrial prototypes, and they tolerate us hobbyists. Reading the Small Parts catalog will keep you spellbound for a few hours, it's worth the call.

    The other source for cathode type aluminum and nickel tubing/wire is MDC, Inc, also in Florida. They also sell high vacuum parts such as flanges and seals and have the tubing and the low temp. brazing materials you need to make tubes. They don't stock the aluminum as cathode material, but it's the right type. Surprisingly the stuff the hobby shops sell in the K&S displays is pure aluminum, otherwise they couldn't extrude it that cheaply. Both Florida companies have Web sites but I (Steve) don't recall what they are.

    Miscellaneous parts:

    (From: Joe or JoEllen (

    A good resource for components found in "Light and its Uses" is:

    (From: John De Armond (

    Duniway Stockroom for items like Dow corning Silicone High Vacuum Grease. Less than $10 for a lifetime supply. Also a nice stock of Viton O-rings (what you want to use instead of Buna-N for vacuum), fittings and other vacuum goodies. Might also take a look at their "vacuum epoxy". Really a common industrial epoxy made by Hysol that has a sufficiently low vapor pressure to be used at the low end of high vacuum. Same stuff Varian sells as "TorrSeal" for many bux.

    More on Obtaining Gases

    British Oxygen Company (BOC) Gases has a variety of technical and safety information on-line as well as handy units conversion tables on-line. They have request forms for an extensive catalog and other technical info.

    Spectra Gases supplies all sorts of gases and support equipment for lasers and related applications including a variety of mixes for CO2 lasers, pure gases for helium-neon, argon and krypton ion, and excimer lasers. See their Laser Gases and Equipment Page. They have been recommended and will sell in small quantities to the private individual (more below). They also are planning some technical and reference pages for their Web site but they are not presently complete.

    The following comments deal with a variety of gases required for laser construction.

    (From: Cass (

    The CO2 laser-mix gas is typically sold in "H" bottles for $85.00 per fill. The company that I checked with will allow one to specify their own blend of CO2, He, N2. Your local welding/medical supply company may vary. Try to find the largest gas company in your area as many of the smaller ones simply use them to fill your bottle(s) and tack on an additional charge.

    (From: Steve Roberts (

    I went to purchase a tank of nitrogen today. I own the tank, so a fill was $10.80 + tax for a size "P" tank (It seems no two gas places use the same letters) because I own the tank. I needed N2 to test a nitrogen laser for a customer. As I was walking out the door it dawned on me to ask what the price for a CO2 mix would be. The fellow couldn't give me the mix percentages, but said that for my tank, the fill of UltraMix by AGA would be $16.85. My tank is about 14" tall by 5.5" in diameter cost $89.00 when I bought it last summer. When I need a fill, they just swap tanks. So I guess I'm getting a tank of mix next week when I'm done with the N2.

    I actually think this is good news because when I talked to the rep on the phone and a leased tank of gas in the same size was quoted at about $200. The tanks all have the same CGA 580 fitting on them so I've been able to get oxygen, N2 and Ar in the same size with the same regulator.

    If your operation is set up like our local gas shop, you have an industry side and a retail side of the business. I'm not a industrial customer, but I make it a habit of going to the industrial dock in person for my gases, its much cheaper then calling in and getting the sales person who seems to jack up the price when you pick it up at the retail end of things. It's easier and cheaper to go buy it from the dock clerk. BTW, if you're hunting for something pure like krypton, Try Spectra Gases. Or one of the others that cater to the laser industry. Even after the hazmat shipping fee, buying a bottle from Spectra was much cheaper then getting it locally. The large company around here varied their krypton price on a day to day basis like it was on the stock market. Spectra's quote was about 1/3rd their price even after shipping.

    (From: Tom Miller (

    The place I checked gave a price of $160 to purchase a 40 cu. ft. tank with a fill. Refills are $60. I have access to the "R" tanks used for medical O2 and these are 20 cu ft (tank is ~2 ft tall and 4 in diameter) and have a tank valve like scuba type gear. I guess the valve could be changed to some standard but it would be nice to use an O2 regulator as they are designed for low pressure and low flow. I wonder if the O2 regulator would work with the high He mix? When I talked to the gas supplier, they said to use a He regulator.

    (From: John De Armond (

    You can change the valve on the "R" tank, as almost all tanks use 3/4" NPT threads. This is a job, though, because the valves are put in REAL tight. when I owned a welding gas supply company, we had a clamp rig that would hold the tank while we pulled on a closed-box wrench attached to a 6 ft cheater. still usually required some heating. One thing to watch for is the thread sealant. If the gas co used pipe dope, you'll never get it clean enough (short of vacuum baking) for high purity gas.

    You really do need to use a helium regulator for He. The reason is the He atom is so small that it'll through the rubber diaphragm of most regulators like a dose of salts! He regulators typically have a 316 SS diaphragm and a copper crush gasket. Helium's kind of tough to hold onto for any length of time. At the welding gas co, we kept close track of the code dates on our tanks because if we let one sit around for a long time, the He would diffuse through the walls and lower the pressure. Got complaints from customers. :-) Also be aware that ordinary He sold for balloons has a goodly chunk of air in it. You can get an inexpensive catalytic gas scrubber from Matheson Gas that will clean up the stream but probably not worth it. Just get clean gas.

    I buy my high purity gas from Spectra Gases. They sell gas in disposable cylinders. A 20l cylinder that looks like a propane torch cylinder but with a needle valve filled with Neon is about $50. A 100l cylinder is about $140. This is spectroscopy grade gas. And when the tank is empty, it is useful for a number of things plus the 100l tank has a standard CGA valve so you can put it on your "E" tank :-) I use a 20l tank as a "day tank" on my bench. I refill it from the 100l tank. That way if I leave a valve open I don't trash the whole 100l tank. These are really good guys to work with and they have all kinds of gas. Need 20l of Xenon? Got a BIG pocketbook? They got it. :)

    Tips for Dealing with a Neon Sign Shop

    Neon sign shops can be a fabulous potential source for inexpensive neon sign transformers, glass tubing, electrodes, vacuum equipment, and glass working services. Gas lasers and neon signs share a lot in common so the people who work on neon may be very cooperative if you convince them you are serious about building a laser. I have even heard of someone not only getting his HeNe glass work done by a friendly glass bender, but having the tube filled with helium and neon at the proper pressure (along with the required bake/back-fills) as well. However, you do need to use at least some common sense when you walk in (and maybe a bit more) so as not to just be shown the exit:

    (From: John De Armond (

    When someone contacts me with a request for a transformer, for example, I'm pretty cooperative once I determine that they probably won't hurt themselves and aren't planning on using it for practical jokes. I don't need the liabilities. I buy used transformers in bulk from a large regional shop, test and refurb them and either use them for my neon or resell them to experimenters. I get a buck a kilovolt for as-is units. If you're lucky, a shop may give you transformers but it generally greases the skids for future requests to offer to pay a bit.

    A few other things to keep in mind when visiting a neon shop:

  • Back to Amateur Laser Construction Sub-Table of Contents.

    Vacuum Technology for Home-Built Gas Lasers

    Do to its size, this material has been given a chapter of its own. Please go to Vacuum Technology for Home-Built Gas Lasers.

  • Back to Amateur Laser Construction Sub-Table of Contents

    Introduction to Glass Working

    Laser Tube Fabrication

    The following mainly applies to the traditional gas lasers like the HeNe, Ar/Kr ion, HeHg, and possibly CO2 where the entire laser discharge tube may be a single glass structure - it is made in one piece from various individual pieces that are fused together. The N2 laser does not require glass working of this type.

    As laboratory apparatus goes, what you need for any of these lasers is pretty mundane: A few tubes joined together with butt or tee joints, a few dimples or bumps, some angled cuts, and pieces attached with glue.

    Note that at least in principle, it is possible to construct these lasers without actually fusing glass pieces together as Epoxy or other adhesive and/or vacuum rated flexible tubes and clamps can be used. However, such a structure is not nearly as stable and are not recommended. In addition, the added nooks and crannies of clamped pieces and places with glue that can outgas mean that achieving the required level of vacuum and maintaining it is much more difficult.

    There are basically two ways to go about obtaining the needed assembly of tubes, electrodes, Brewster windows, and so forth.

    1. Have someone else do it! Assuming you can find a cooperative individual or pay a neon sign shop or laboratory equipment fabricator, this is by far the easiest especially if you have to start from ground-zero. For someone at all experienced in this sort of stuff, the assembly of the main portion of a typical laser tube (not including the Brewster windows) is a 20 minute job if all the bits and pieces are available. If you can arrange that your design uses the same sizes and types of glass tubing that they commonly deal with, so much the better.

    2. Learn enough of the skill of glass working to do it yourself. This is by far more fun and who knows, maybe you have a talent for the sort of glass art exhibited in museums. This really isn't as difficult as it might seem at first. Glass working is a skill and you will no doubt create some pretty interesting failures at first. But with a little practice (OK, maybe a lot of practice!), butt and tee joints, and dimples and bumps will become second nature. We aren't talking about fancy decorative glass blowing, mostly just basic cutting and joining.
    Glass working as it relates to laboratory apparatus fabrication has been covered in the Amateur Scientist columns as well. See: Glass blowing, technique explained, Scientific American, May, 1964, pg. 129.

    A recommended book on this topic is:

    In sections below provide the briefest of introductions to the glass working skills that are needed for laser tube frabrication. But, here's a Web site with extensive material on scientific glassblowing or glass working (same thing): It includes information on safety, equipment, materials, and terminology; an extensive illustrated tutorial on basic techniques, tips, glass recipes, useful data and tables, and more.

    Types of Glass

    What we call "glass" is made from silicon dioxide (SiO2) and other additives to produce the wide range of properties of various glasses that we are familiar: from window glass to Pyrex cooking and labware; colored glass bottles and stained glass windows, light bulbs of all shapes and sizes; and optical glass for lenses, mirrors, and prisms. SiO2 is the same stuff that constitutes beach sand and the insulating layers of integrated circuit chips.

    Glass is an amorphous material - it has no crystalline structure and is really a liquid at room temperature. A liquid, you say???? Well, a slow moving liquid at least. As its temperature is increased, glass becomes softer but has no distinct melting point (compared to water, salt, or any other material that forms a crystalline structure where there is a distinct phase transition from a solid to liquid state).

    The two types glass which will be of most interest for laser construction are:

    Fused silica (Vycor) and pure quartz are two highly heat resistant materials that you hopefully won't have to shape since they have even higher softening (or in the case of quartz, a crystal, melting) points as well!

    If you order common laboratory glass tubing, it will likely be made of S-L glass though other types are also available - make sure you specify what you want since for some of the laser parts, heat resistance is an issue. Most beakers, flasks, and anything else that may be heated are made of Pyrex or the equivalent B-S formulation of another manufacturer. However much other labware is of the S-L variety. Since the coefficient of thermal expansion also differs for the two types of glass, there may be problems in attempting to mix them in a given structure.

    Cutting Glass Tubing

    For small tubes - say less than 1/2" in diameter, cutting is, well, a snap!

    All you need is a small triangular file (new or in excellent condition, not rusty and clogged with something disgusting) and perhaps some spit. :-)

    Sometimes, wetting the filed location with a bit of spit or tap water will aid in the process.

    Practice on some scrap pieces of tubing. In on time you will be turning all the neon tubing in your neighborhood to small bits suitable for making beaded necklaces!

    This also works for larger diameter tubing (like CRT necks) but a longer crevice may be needed - try to keep it straight. In some cases, one pass all the way around will be needed.

    There are also hot wire cutters (the heated wire produces local stress which fractures the glass). For large or irregularly shaped objects, the best tool is a power driven diamond grit glass cutting wet wheel - a water cooled miter saw for glass and ceramics! For small pieces, a Dremel tool (compact high speed multipurpose hand-held grinder/saw/sander/drill) can be use though you may go through (non-diamond coated) cutting wheels rather quickly. :( Dentists have nice high speed drills with diamond impregnated bits, cutters, sanding disks, buffers, and other widgets - and they are water cooled! Perhaps, you know a friendly dentist? "Please be patient Ms. Jones, we'll get to your root canal as soon as we finish cutting these glass pieces for Jerry's laser". :)

    Any sharp edges left by the cutting operation should be smoothed with fine sandpaper or in the flame of your glass working torch.

    Glass Working 101

    All glass working consists of four steps:
    1. Heating. This is going to be done with a flame of some kind:

      • A common propane torch or natural gas burner using air is just hot enough to soften S-L glass. A bunsen burner works - barely. Other types of lab burners are better.

      • With the use of pure oxygen, the flames from these all run much hotter and that is what you is really needed to be able to do any sort of glass work easily and consistently (or borosillicate glass at all).

      • An oxy-acetylene or oxy-hydrogen torch will be needed to easily deal with some types of heat resistant glass and fused quartz. (CAUTION: Hydrogen flames tend to be invisible!)

      At the proper temperature, glass has the consistency of soft taffy - easy to bend and shape but not so soft that it runs or drips. Part of the skill (and fun) is keeping the glass at just the right temperature as it is worked. As the glass approaches the proper temperature, the flame will take on a yellow tinge from the sodium ions in the glass (the soda part) and the glass itself may appear red or orange-hot itself.

    2. Working: Bending, joining, pulling, dimpling, blowing, etc. is done while the glass is maintained at a relatively constant temperature in the flame. Glass cools quickly so repeated or constant heating is needed. Some positive pressure in the glass parts may be needed to prevent them from collapsing - or to blow bubbles! The surface tension of the soft glass is going to be both our friend (since it will help smooth out much of the damage you will inflict) and foe (since it will tend to want to cause tube ends to collapse or other holes to expand). Usually two hands and a mouth (safely at the end of a length of rubber tubing!) are enough but at times you might wish to be an octopus!

    3. Once the particular joint or whatever is formed to your satisfaction, the piece must be cooled so that it solidifies. However, you cannot just dunk the whole affair in a bucket of water as the sudden temperature change will cause your hard work to shatter into a million pieces (sometimes it will do this even without such help!) It must go through a process of annealing where a lower temperature flame is run back and forth over a large area of the glass - beyond that which was dealt with originally. The cooler flame can be obtained by reducing the air or oxygen supply to the torch. Fortunately, this takes only a couple of minutes for anything we are interested in constructing (unlike the 17 foot diameter Palomar telescope mirror which required over a year of annealing). Note that there is no real way of knowing how much annealing is enough - it is just something that one does based on recommendation or experience.

      Surrounding larger pieces with warm vermiculite after flame annealing will further slow the cooling process. Vermiculite is ground up mica and is sold in garden shops. :) This is probably not necessary for the sorts of things needed for home-built lasers but one never can tell where your activities will lead!

    4. Cooling. The worked and annealed area will still be very hot. Set it down on a non-flammable material or better yet, in such a way that the hot parts do not touch anything until it is cool enough to touch. This allows it to cool slowly and uniformaly, further minimizing the chances of stress cracks.

    Gas Flames

    A gas flame (natural gas, propane, etc.) adjusted for hottest temperature (optimum fuel:air ratio) is divided into several parts:
             Tip---> /\ (Dark Blue)
                    /  \
          Cone --> / /\ \ (Light Blue)
                  | |  | |
         Burner |          |
    Great diagram, huh?

    Note that it is mostly shades of blue - there should be minimal yellow or orange (indicating that there is adequate air/oxygen) but the flame should not begin to separate from the burner (indicating too much). There should be no smoke or soot from such a flame.

    The hottest location is just above the inner cone.

    With soda-lime glass, once the glass is hot enough to work, the flame will take on a yellow color due to the sodium ions in the glass.

    With the air/oxygen supply cut off, the flame will be long and yellow and may produce black smoke and soot. This will be the proper temperature for the annealing step.

    Note: Where you have control of the air/oxygen supply as with a professional glass working torch (or Oxy-Acetyline welder, for that matter), light it up by first opening just the gas supply a small amount and then adding air/oxygen and adjusting gas flow after the flame is lit. Shut down in the reverse sequence. This avoids unsightly pops, bangs, and other explosive behavior. In other words, always make sure the gas is turned on first and shut down last!

    Glass Working Examples

    The only way to really be come proficient at this is to practice. You will create many many interesting disasters at first but glass is cheap. After a while, these sorts of 'simple' procedures will become automatic and second nature. Who knows, even your failures may find a place in the Museum of Modern Art!

    Home-Built Glass Working Lathe

    Even with a lot of practice, it is difficult to make consistently high quality joints between various size glass tubing and other parts free-hand. A glass working lathe is a fixture that permits the parts to be mounted in such a way that they can be rotated together in the gas flame(s) while maintaining precise alignment and leaving at least one hand free (the other would do the rotation - this needn't be motorized) to do other things. It probably doesn't make sense to acquire or build a glass working lathe for the fabrication of a single laser tube but if several more complex lasers are in your future, it might be worth considering.

    A commercial glass-working lathe is a fairly expensive device that can't be justified for sporadic and limited hobby work. However, A Home-Built Glass Working Lathe provides a short description of a piece of equipment which can be used in the fabrication of the plasma tube, water jacket, and other glass components for a variety of lasers as well as for other scientific glassware applications.

    Learning Glass Working Skills

    There are many Web sites with glass working information. Teralab - Glass Blowing for Vacuum Devices is just one example. But more hobbyist-types do this than you might thinkg.

    (From: John De Armond (

    One way to start is to subscribe to a couple of email discussion groups related to glass working. There are 2 lists I'm involved with:

    See the section: Laser (Email) Listservers for other possibilities.

    The list discussion tends to be esoteric. There's no substitute in learning to blow than to do it! I taught myself to blow glass so it can be done. First, get the books mentioned in my article on the web page. Then find a neon shop and arrange to spend some time in the shop. Most shops will let you watch for a bit but it's only proper to compensate the benders for their time. most benders work on piecework so time = money. When I first got started, I slipped a bender in Atlanta a C-note to let me tag along behind for a day. I had his undivided attention. This is VERY important. The early motor skills you learn will be the foundation that you build on with practice and if you learn incorrect basics, it will be hard as hell to correct it. It is also important to understand that there are some people whose motor skills will not let them learn to blow glass. I've seen people who went through $6k, 9 week neon bending classes and not be able to do a 90 without it flattening out. You can usually tell within a few hours' work. If you end up being one of those people, just live with it. You'll waste a lot of money and make no progress. Avail yourself of someone who can.

    After you progress to one level, don't be afraid to spend a little money for a day of tutoring from a master. I've been bending for several years and I still try to go a couple of times to someone better than me and study under them. Now you can usually get someone to do it for free but things are a lot more clean if you simply pay the expert for his expertise. You'll profit in the long run.

    Here's Someone Who Has Offered Glass Working Services

    Here is a possible way of getting your custom glass work done. I do not know what he charges but it sounds like if you supply the proper specs and drawings, there will be high quality results:

    (From: Georges Koff (

    Although the bulk of my work is for industry and research, I do work for individuals.

    I am a professional scientific glassblower, a member of "The American Scientific Glassblower Society". For the last 27 years, I have been in charge of the glassblowing shop of McGill University in Montreal, Canada. I have been working in research for the last 33 years in Europe and Canada and have a lot of experience in the design of scientific glassware, high vacuum technology, glass-to-metal seals, special glasses, etc.

    I design and work from blue prints so any drawings could be submitted including those in electronic form (most formats are supported).

    I do not consider the type of laser but the kind of glassware to be blown. So as long as the design of the laser tube is fully specified, it doesn't matter whether it is for a CO2, HeNe, argon ion, or other type of laser. I most likely should be able to make them.

  • Back to Amateur Laser Construction Sub-Table of Contents

    Electrodes and Getters

    Types of Electrodes Used With Home-Built Lasers

    The electrodes used with home-built gas lasers are of two types, both of the 'cold' variety:

    A third type may be found in some lasers and is, of course, is commonly used in vacuum types including CRTs and microwave magnetrons:

    The cathode (negative electrode) is usually where the most heat dissipation occurs due to ion bombardment. Thus, it is generally made large and of a material and shape to minimize heating. The anode can often just be a little wire through a glass-to-metal seal.

    Getters in Home-Built Lasers

    A getter provides a means of ridding a permanently sealed system of the last traces of unwanted gas molecules by chemically combining with them to create a stable, non-gaseous compound. Common getter materials are designed to react with as many gases as possible. This includes O2, N2, CO2, H2, and many others but NOT noble gases since they won't react with much of anything. So these getters can be used with noble gas lasers like the HeNe and Ar/Kr ion type, but not with most others. The getter is that silvery or black metallic spot visible in a vacuum receiving tube (if you remember those) or CRT (though it may be hidden by the external or internal coating). Some HeNe laser tubes have getters but this doesn't seem that common today and those that do often have not had them activated (the getter electrode is present but the getter spot is missing. I suppose that modern vacuum systems and processing methods are so good and hard-seal tubes don't really leak, so there is no need for a getter). Argon ion laser tubes may also have getters but they are more likely to be hidden behind metal and beryllia.

    The getter is a one-time use device and is ruined by exposure to any significant amount of gases in air like oxygen and nitrogen (more than a dozen molecules or so - OK, just a slight exaggeration!). Thus, their use doesn't make sense with a flowing gas or continuously pumped system even if all it uses are noble gases. And, the residue from the getter reaction may be a powdery substance that will contaminate the laser tube.

    Therefore, none of the home-built lasers requires or should include a getter unless the intention is to permanently seal off the tube - which isn't recommended unless your gases are ultra-pure, your vacuum system is superb, and your attention to minimizing contamination is equally superb. If the tube isn't hermetically sealed, air is almost certain to enter the tube between uses at the very least.

    So, you ask, "How does one install a getter if it is ruined by air?". Good question! The answer is that it is manufactured in an inactive form as a compound that is inert and contained in a small U-cross-section metal structure (the getter electrode). This must be activated by heating (once the tube has been pumped down and sealed off) to decompose the inert compound driving off a reactive metal that forms the getter spot. Should that metallic spot turn milky white or red (depending on the actual compound used), the getter has outlived its usefulness and the tube is probably leaky and no longer functional.

    Apparently, depleted uranium (which is not radioactive - or at least not very radioactive depending on purity) makes an excellent getter after being processed to convert it to the so-called 'active' form. For obvious reasons, I'm not going to go into any more detail here but a search of the relevant scientific literature of the 1960s will turn up the complete recipe if you really want to build a nuclear laser and can get a few grams of the stuff. :)

    Methods to Activate Neon Sign Electrodes and Getters

    All neon sign (and most other types of) electrodes need to be heated via an external power source to prepare their surface for use ('activation'). Getters in these lasers as well as some commercial helium-neon and other gas laser tubes may need to be activated or reactivated to clean up contamination due to poor manufacturing or air leakage.

    During the manufacture of commercial tubes, they may be made red hot. My solar heater (see below) probably doesn't achieve this (for HeNe laser tube getters) but it also takes awhile. The minimum temperature for activation probably depends on the type of material used but I expect that it is at least several hundred °C. There are a variety of ways of providing the source to heat the electrode or getter:

    CAUTION: Make sure the glass-to-metal seals do not overheat while attempting any of these procedures! Also, any procedure you use should restrict heating to only the electrode itself or its contents. For getters in particular, heating the 'white cloud of death' (which might be inevitable with the solar heater) will likely result in an increase in unwanted gases as it releases the contamination that had been collected over the eons. :(

    (From: John De Armond (

    All regular electrodes must be activated. Running an un-activated electrode will crap up a tube in short order. There is one company that will supply uncoated electrodes on special order (maybe "Tecnolux" but I'm not sure). In any event the coating can be removed with a nitric acid wash. The electrode will still need to be heated to outgas it though. I'm not sure how one would attach an electrode with epoxy and still get it hot enough to properly outgas.

    Also note that barium carbonate (reduces to barium metal when activated) is one of the common activation ingredients. The metallic barium that results functions well as a getter. I've experimented to see just how bad a vacuum I could get away with. Using my usual brand electrodes, EGL, I can leave up to about 50 microns of air in the neon tube and still have the electrodes clean it up fairly rapidly.

    Basic Induction Heater Circuit

    An induction heater is just a source of high frequency driving a coupling coil. This acts as the primary of a transformer where the secondary is the object to be heated. The low voltage but very high current induced thus heats the object resistively. Induction cooktops may require multiple kW. For our purposes, power is much lower, perhaps 50 to 200 watts (though the schematics aren't all that different).

    The typical circuit is a line powered high frequency driver providing about 300 V p-p to the coupling coil of the induction heating head. It consists of a voltage doubler and filter, a half-bridge controller chip like the IR2151 or IR3M02, a pair of N channel MOSFETS, snubber, and return capacitor network. A half-bridge consists of a pair of switches (MOSFETs in this case, though bipolar transistors or IGBTs are often used) in series, connected between the positive and negative DC voltages. The output is taken from their common point. Clearly, only one had better ever on at a given instant!

    The key part of the typical circuit is shown below:

         +150VDC o--------+-----------------+               Coupling
                          |                _|_ C1             Coil
                      .|--+ Q1             --- 1uF      +------+
               P1 o---||<-. IRF850          |  250V     |       )
               _      '|--+                 +-----------+       )   
             _| |____     |                 |                   ) O Item to be
                  _       +----+------------|-----------+       )     heated
             ____| |_     |    |   Rs   Cs  |           |       )
                      .|--+    +--/\/\--||--+           +------+
               P2 o---||<<-. Q2             _|_ C2
                      '|--+ IRF850         --- 1uF
                          |                 |  250V
         -150VDC o--------+-----------------+
    The schematic of a commercial unit that I've seen is based very closely on the IR2151 Half-Bridge Driver Datasheet.

    A typical coupling coil consists of about 200 turns of #18 to #20 magnet wire. At this time I do not know if it has a ferrite core - I intend to find out.

    Amateur science and hobby outfits like Information Unlimited have plans, kits, and assembled induction heater drivers and heads suitable for this application. Commercial versions may be available from scientific and neon sign parts and equipment suppliers. Or, perhaps, you could borrow that old diethermy machine sitting in your doctor's back room. :)

    For an alternative to a special induction heater driver, see the section: Sam's Recycled PC Power Supply Induction Heater.

    Sam's Recycled PC Power Supply Induction Heater

    The induction heater driver circuits for firing getters and activating neon sign electrodes that I have seen produce 300 V p-p at around 50 kHz.

    Well, you know how lazy I am and always wanting to build stuff with recycled parts. :) So, rather than going to the hassle of constructing a line powered half-bridge induction heater driver, guess what's inside a typical PC switching power supply? Figure it out yet? A dual MOSFET 300 V p-p driver capable of 100 to 200 W! Cool, huh? :)

    So, I had this really bedraggled slightly water-logged no-name PC power supply sitting sort of in pieces (minus its case) in a box minding its own business and decided to see what it could do. I located the common point of the MOSFET pair and took a look at that with a scope (grounded to the source of the lower MOSFET, everything on an isolation transformer for safety) with the power supply driving my head light load. Almost perfect! 300 V p-p but the frequency is a bit lower than I would like - 20 kHz rather than the 50 kHz used in the commercial induction heater driver. The lower frequency would require a larger number of turns on the coupling coil and a lower single-turn voltage on its secondary (the getter or neon sign electrode), neither of which is desirable. Ah! No problem, just change a capacitor or resistor! The controller is an IR3M02. Its timing capacitor looked like a high quality job so I left that alone. The resistor was 30K. I substituted 12K and presto! It now runs at 50 kHz, apparently none the worse for the experience. The MOSFETS, transformer, and rectifiers still seem to be cool enough. The waveform still looks clean. I suppose the switching losses are now about 2.5 times what they were before, but I won't be running this continuously at full power. I don't know whether the range of load regulation will be the same as before but I really don't care as long as nothing blows up. The supply even had the same configuration of split caps and snubber as the sample induction heater circuit so I didn't even need to add those components! I then attached a cable and connector for the drive and return connection to the coupling coil in the induction heating head.

    Note that due to the low duty cycle at the light (no pun...) load, the waveform looks more like a modified sine-wave than an ideal squarewave but should be acceptable. With a larger load, it should more closely approach a squarewave shape.

    WARNING: Consider doing this at your own risk. PC power supplies are directly connected to the power line - use an isolation transformer for safety. The output drive is 300 V p-p with a few hundred WATTS available, at least momentarily. This is an extremely dangerous setup in any case - make sure everything is well insulated. Never change connections with the power on. The main filter capacitors in the power supply can store a lethal charge for quite some time - always confirm that they have discharged and/or discharge them if necessary before touching anything - a couple of 47K, 2 W resistors should be added as bleeders if there aren't any originally (though it will still take a minute or two for the caps to fully discharge). It is also quite possible that any attempt at changing the power supply operating frequency will result in instant smoke. Although it worked once, this doesn't necessarily generalize to other PC power supplies and not all designs are based on the dual MOSFET drive circuit.

    To calculate the inductance needed for the coupling coil, I assume that the maximum available power is about 100 W (to err on the safe side, hopefully!) which for the 100 VRMS drive (which is about what this is) would have an impedance of 100 ohms. The inductance then follows from: L = Z/(2 * pi * f). For f = 50 kHz, the result is about 300 microHenries.

    According to an ancient Allied Electronics Data Handbook, the following formula will give the inductance of a multi-layer coil "to within approximately 1% for nearly all small air-core coils".

                                         0.8 * r * N2
                            L = -------------------------------
                                 (6 * r) + ( 9 * l) + (10 * b)
    Where: Don't you just love the mixed units - this was before the forced conversion to everything Metric! :)

    Anyhow, from this equation, it looks like about 100 turns on a 1 inch form should do it. Stay tuned. Next exciting installment: Winding the coupling coil.

    Sam Schwartz's RF Power Oscillator for Heating Getters

    So, perhaps my whimpy PC power supply based system isn't going to be adequate. Here's one that should be!:

    (From: Sam Schwartz (

    I have built a power oscillator for firing getters that should work well for HeNe and other types of laser tubes. The complete schematic is shown in Sam Schwartz's High Power RF Oscillator.

    The oscillator uses an 813 vacuum tube. I burned up a handful of power MOSFETs before deciding a tube was a better way to go, at least easier for the amount of time I wanted to put into the project. The work coil should be made of aluminum rod and not copper, unless it is water cooled. The heating in the secondary of the transformer is all skin effect loss, and requires water cooling. Alternatively, the heating problem of the transformer secondary might have been solved by using a rope size piece of Litz Wire. Notice that the primary is Litz. That is no inconvenience in my setup as water is used to cool the diffusion pump and the baffle on the pump. All are connected in series with plastic tubing. I forgot to mention two items in the drawings: There is a twisted pair between the RF head and the Main Box that connects the heat control pot in the RF head to the solid state control circuitry in the Main Box. Also, there is a wrapping of fiber glass tape, available at any plastic store, around the work coil. This prevents accidental breaking of glass by inadvertent touching of the coil to the glass. Notice that a current rating is given for the 0.003 uF. transmitting mica capacitor that tunes the tank coil (transformer primary and reflected secondary) in the RF head. That is a necessity as the circulating current in the primary is about 5 amperes as measured with a current probe. Multiply that by the transformer ratio, and you will find a work coil current of over 100 amperes. That helps to explain the skin effect heating.

    The frequency of the oscillator is 250 Khz. It will vary somewhat depending on the size and shape of the heat coil on the transformer secondary.

    I have used old 1/2" getters out of a vacuum, in open air, as test subjects. They will get cherry red in the center of the work coil, even when moved around so that they are not centered in the coil. So that should work for your uncentered getters. Similarly, one can tie a large nut on a wire and suspend it in the coil and it will glow cherry red. It is interesting to put small chips of broken glass in the getter ring and watch the glass melt and conform to the ring while heated with the work coil ! The heat control is very effective in controlling the amount of glow (heat) of any ring shaped object in the work coil. The diameter and number of turns in the work coil is not critical, within reason, and may be varied to suit the application. Also, it may be oval shaped or other than round.

    When flashing a getter, the heat should be brought up slowly, over a few seconds, until the getter starts to glow and then lowered quickly as the getter itself goes into a self heating mode. Too much external heat causes the getter ring (stainless steel or nickel) to vaporize and deposit material over the getter material. The result is a black deposit as seen on the glass. Am now looking for a spare 813 tube in case my 'old but goody' gives up.

    If you cannot slip the work coil over your tube in the vicinity of the getter, you can direct the field with a piece of ferrite like that from a flyback transformer. Some of the color picture tubes use this technique when the neck is too short to accommodate the getter.. They have what's known as an antenna getter, which is a getter on the end of a long flexible strap attached to the end of the electron gun. When slipped into the CRT, it follows the curvature of the deflection area around into the funnel and a part of it finally rests against the inside of the cone (funnel). It is activated from the outside of the tube by a work coil having a round ferrite core that is placed in the proximity of the getter. The back of the getter has a ceramic seat that contacts the glass and prevents a "hot spot" that otherwise could crack the glass CRT.

    Here are some notes:

    1. The Power Oscillator provides a variable range of power up to about 120 watts for induction heating. The output power is controlled by duty cycle modulation of the main oscillator and gives a continuous heat control over a wide range.

    2. The RF coupling transformer was built on a ferrite E-E core which was a TDK PQ 40/40. The 40/40 refers to the outside dimensions of 40 mm wide and 40 mm high. These are 'E' core halves with a rounded interior profile to accommodate a round coil. The round center leg has a diameter of 15.2 mm. The core material is 7C4, which has a high saturation flux density rating commensurate with low core loss, and is made for switching power transformers. It, or its equivalent, is probably available today from other ferrite core suppliers such as Siemens, etc. NO air gap is used as the transformer carries no DC current, since it is capacitively coupled from the 813 plate feed choke.

    3. It is essential that the primary winding in the output transformer be made of Litz wire. In my version, I used a bundle of small wires, typically #30, twisted together with an electrical drill.

    4. It is also necessary that the secondary of the output transformer be supplied with cooling water, otherwise the solder will melt out of the connections.

    5. The Work Coil should be bent from solid aluminum rod (not copper) as it will not oxidize appreciably in a hot oven or from its own heat. One can melt a copper penny located in the work coil at full power!

    6. It is necessary to deliver the output power from the source to the object to be heated, which is usually some distance away. For convenience and to prevent loss of the output power in the transmission, the output transformer and tuning capacitor are mounted in a separate small box with the work coil mounted on it. A connection between this output box and the main oscillator is by means of large coax. Therefore, the power is delivered at a high voltage and low current (about 2 A) over a relatively long distance without appreciable loss and then the current is increased to 125 amperes in the work coil. The buildup of current comes about through the circulating current in the tuned circuit and through the step down ratio of the output transformer.

    Additional information on the system and details of the coupling transformer can be found in Sam Schwartz's Power Oscillator Notes.

    Simple Solar Heater

    I built this contraption in order to activate the getter inside a contaminated helium-neon laser tube (see the section: HeNe Tube Lases but Color of Discharge Changes Along Length of Bore. Solar Heater for Activating Tube Getters shows a HeNe tube in position ready for a treatment. (I haven't actually tried it on anything else so far.)

    The solar concentrator is just a $1 7" x 10" plastic Fresnel lens which is supposed to be used as a reading magnifier. It is taped to a wood frame with two degrees of freedom so that both height and angle can be adjusted depending on the location of the Sun and height and orientation of the structure of interest inside the tube. Despite being rejected as really terrible for its intended application, this cheap lens is capable of producing a nicely focused spot less than 1/4" in diameter. Based on an estimated 600 W per square meter of noonday Sun where I live, this works out to about 25 W of light energy in that area. Sounds like a nice size laser, huh? A larger lens can be used but don't go to extremes - even this one can be nasty (especially for any unsuspecting bug that might wonder too close!).

    It might be a good idea to cover the base with some sort of non-flammable or fire resistant material to prevent unfortunate 'accidents. Whatever the focus of the beam hits will char instantly and 6 foot flames won't be too far behind!

    I expect it could be used to make some pretty decent hot dogs as well. :)

    So, you ask: "Why isn't this considered as dangerous as a 25 W laser?". The main reason is that the Fresnel lens does not produce a collimated beam and it is virtually impossible to convert it to one. Yes, putting anything at its focus will cause the item to be charred, cooked, vaporized, incinerated, or otherwise damaged. However, 2 or 3 inches away, it is just a very bright source of light. Nonetheless, DON'T ever stare into the light from the Fresnel lens!

  • Back to Amateur Laser Construction Sub-Table of Contents

    The Central High School Cyclotron

    These sections summarize my experiences in high school many years ago with an atom smasher, sort of. :)

    Why is This Here?

    OK, so why is this included in a Laser FAQ? Well, the simple answer is that it would seem pretty silly in the IR Remote or VCR repair guides, wouldn't it?! :) Seriously, there is a great deal of commonality between atom smashers and lasers - at least in terms of the technologies involved.

    First, the major players (complete names are not included since I am unable to contact them for permission to make this public):

    Now, Central High School (CHS) wasn't a technical high school but was and still is to some extent, one of the principle academic high schools in Philadelphia, PA, USA. At that time, it was also an all boys school, with Girls' High, a long block away. Now, it is co-ed.

    There were the usual debating, rock climbing, sports, cycling, and all the other associations, clubs, societies, whatever, typical of any high school.

    There were also the technical ones. We had an amateur radio society and photo society, of course. There were also probably organization for chemistry and biology but I don't really recall as I wasn't enthusiastic about chemistry (despite having acquired a cabinet full of beakers, flasks, and such) and biology was even lower on my list of fun activities. But I do recall a field trip to the Pine Barrens in New Jersey to check out the trees. :)

    I also managed to get involved in the Advanced Physics Lab (APL) where I did inherit a ruby laser based on a mid-60s Popular Science design (that never worked or maybe I was too chicken to turn up the capacitor charging circuit to reach threshold). (See Modern Mechanix - Build Your Own Laser for the original article if you're curious.) Mostly, what was done in the APL is that Fred and I abused burned out light bulbs by among other things, driving a 110 VAC to 2.5 kV transformer on 220 VAC and I would swear that for several seconds, achieved a plasma jet through the bulb in open air. But it also looked good on my academic record!

    There was also the... Cyclotron Society. Myself, Doug, and Fred were the principle members. There was also a guy named Gary but he used the place more for political reasons and always seemed to find excuses NOT to do anything technical. Fred and I eventually got him thrown out of the Cyclotron Society on the basis of some infraction which to this day I don't know was real or not. Douglas was in some ways similar and the room (to be described below) DID make a good book repository and hangout! He, at least, didn't interfere with progress.

    Now, lest you think this is the sort of accelerator you have seen on documentaries or even bad Sci-Fi movies, you would be disappointed. It was small - the maximum energy was theoretically if everything was optimal, maybe 1 million electron volts (1 MEV). The diameter of the magnet pole pieces were 7 inches! However, everything was there - magnets, RF source, vacuum system, gauges, the works.

    What is a Cyclotron?

    The cyclotron is the earliest type of atom smasher which uses the combination of a magnetic field to confine the (usually) protons and an RF field to accelerate them. It was invented by Ernest O. Lawrence in 1930. In a cyclotron, both the magnetic field and RF frequency are constant. A charged particle traveling perpendicular to the magnetic field lines will follow a circular trajectory with the diameter determined by its energy (speed). For velocities much less than the speed of light, the period of each orbit and thus the frequency of rotation is a constant. So, if a little boost if given to the particles as they circle in the magnetic field, their energy will increase along with the orbit diameter.

    Cyclotrons can be of a wide range of sizes. Lawrence's original invention was only about 5 inches so ours at least was larger than that! The upper limit on size is imposed by relativity. Once the energy of the particles approaches their rest mass (from E=mc2 or equivalently, their speed approaches the speed of light, the constant magnetic field/RF frequency behavior doesn't work anymore and they would lose synchronism with the RF. I've seen a cyclotron with magnet pole pieces about 2 meters in diameter but even this is probably pushing things. There were atom smashers called "syncro-cyclotrons" which were physically similar but varied the RF frequency as the particles spiraled outward to maintain synchronism. Nowadays, the modern versions are mostly "synchrotrons" which do away with all but the outer ring and may be miles in diameter producing energies of 100s of Giga electron-volts (GEV). The "Super Conducting Super Collider" was supposed to be over 20 miles in diameter (perhaps more, I forget) but that died when BIG science fell out of favor in Congress. The largest one was at Betavia, IL for awhile but that has probably been surpassed by something at CERN in Europe by now.

    The CHS Cyclotron Facility

    I wish I could have been involved in the original construction of this thing, but my tenure at Central had more to do with what might be termed a major rework and upgrade. The details of how a public high school ended up with an atom smasher may forever remain a mystery but what is known is that the original version was cobbled together by students in the years prior to my high school days. The main mover and shaker was probably Donald, who if I recall correctly wrote a paper of sorts on "The effects of High Energy Protons on Semiconductors" - he stuck transistors with their covers removed inside the cyclotron and measured some combination of parameters and how they changed with exposure to high energy protons. Now, this was most likely totally bogus for reasons that will be come clear below, but I think it did win him a National Merit Scholarship.

    The Cyclotron Room was on the main corridor in the school basement just around the corner from the Bookstore and opposite one of the storage areas for those Nuclear fallout rations the Government was so fond of stockpiling around the country (I think the mice benefitted mostly). This was sort of fitting and I suppose there were some teachers who wouldn't rule out a nuclear accident from our activities! We shared the approximately 15 x 20 foot room with one of two huge water chillers for the school. (I never did quite figure out if these were for the fountains or something else.) However, the 5 gallon container of refrigeration oil left by HVAC techs next to the chiller came in handy from time-to-time.

    The power supplies, RF source, and instrumentation were mounted in a 6 foot and 3 foot rack. The magnet with the vacuum chamber was behind it with the vacuum system underneath. The "instrumentation" was a microamp meter for measuring beam current later upgraded to one with a tube based preamp. We had a collection of probably non-functional radiation meters most likely "liberated" from that storage room across the hall. There may have also been an actual working Geiger counter as well. Our test equipment consisted of a very abused Triplet 1K ohm/volt VOM with a bent needle, having been unwound from the right-hand stop more than once.

    The magnet:

    This was a HUGE 7 inch resistive (hey, no one knew about superconducting magnets in those days) magnet wound with a lot of #20 wire. (Once, we found the coil with an open connection - possibly cut by a saboteur! Maybe that Gary fellow I mentioned - really never found out. Or, it may have just been a natural failure. Needless to say, there was a minor panic until the break was found!). The magnet was on a 99 year lease for $1 from some company, maybe GE, like they'd ever want it back! I don't know how they got it into the place, in pieces I guess.

    The magnet was horizontal. To install or remove the vacuum chamber required unscrewing some huge bolts and then prying apart the yoke. Now, keep in mind that this thing weighed in at about 6 tons even though it is a rather small magnet. So, replacing anything inside the vacuum chamber was always an interesting exercise. Underneath was a very fragile glass diffusion pump that somehow survived.

    Originally, power for the magnet came from a 200 VDC or so power supply. We eventually concluded that this was grossly underpowered (another reason to suspect the cyclotron never really worked until the prior management) so we changed it to the 2,400 VDC electric utility pole transformer based power supply that was originally used for the RF source (mercury vapor rectifiers and all that). It was controlled by a BIG Variac. With this improvement, it probably was running near the 20,000 (2 Tesla) limit of an iron core magnet. It was quite impossible to extract any ferrous/magnetic objects from between the pole pieces with the magnet fully energized.

    The vacuum system:

    We had a Sargent-Welch rotary vane mechanical pump, maybe a 1405 but could have been the "economy version", and an all-glass (except for the tower and heater) oil diffusion pump. The Welch usually needed to be started by hand as do many of these older pumps. Belt guard, what's a belt guard? :) There must have been a box fan or something to cool the diffusion pump since it did not use tap water. The diffusion pump was about 3 inches in diameter with a glass O-ring flange-flange reducer for the 1 inch coupling to the vacuum chamber squashed between the magnet pole-piece right above it. There were no baffles, cold traps, or dryers.

    An ion gauge provided our only real indication of vacuum level other than that wonderful clacking sound the mechanical pump made when it has drawn down below 10 microns or so. We had the obligatory hand-held Tesla ('Oudin') coil for leak testing though it was used much more often to chase the unwanted visitors from the cyclotron room. :) Anyhow, once the whole affair was drenched in Red Glyptal, leaks really weren't an issue!

    I seem to recall a scavenged thermocouple gauge as well but that may have been for the "Linear Time-of-Flight Resonance Mass Spectrometer" I started to build. That at least got me a free cruise on a Navy destroyer escort and my first plane flight to the Newport Navy Base as a consolation price at the local science fair since it was not completed and never really worked - but the charts and front panel were impressive! I still have the Science Fair posters. Oops, that's another story.....

    The vacuum system would do at least 10-6 Torr when it was cooperating. Usually, it didn't take long to get there but since we really didn't trust the ion gauge all that much, we usually left the thing running overnight. (The actual pump-down time was probably measured in a very small number of minutes.) But, the ion gauge tube was never damaged from loss of vacuum. :) The controller might have been a Varian RG-31X, a semi-antique even at that time.

    The vacuum chamber:

    For our huge 7 inch magnet, we needed a vacuum chamber large enough for the Ds (the electrodes that actually accelerate the protons - shaped like the letter 'D') and a little clearance. In all, it was about 12 inches in diameter and two inches thick. The 'D's (actually only one, see below) was mounted on a metal stud fixed in Epoxy passing through a glass insulator.

    Most cyclotrons have a pair of 'D's and drive them with a balanced RF source. Someone decided that this wasn't necessary, so ours had a single brass 'D' made of sheet metal brazed at the edges with the RF between it and the ground of the vacuum chamber. In principle, this would work though I suspect there would be less of a focusing effect. What did we know? I seem to recall later replacing the single 'D' with a pair made from machined aluminum.

    The RF source:

    Originally, a 200 W amplifier driven by a 50 W exciter generated the RF (about 20 MHz if I recall correctly) for the 'Ds'. This was based on a pair of 811 tubes (remember those?). I later built a 1 kW linear amp (from the AARL Handbook) using an EIMAC 3-400Z or something like that. The original power supply (866 mercury rectifier tube based) was moved to the magnet and a new solid state HV power supply was built for the amp.

    True R and D

    I consider the years I spent working on that machine to be more closely akin to true scientific research than anything since - designing giga instruction per second high performance 3-D visualization/graphics accelerators or microchip lasers just doesn't have the same feel!

    I had to learn - on my own - about high vacuum technology, high power (well, relatively speaking) RF, instrumentation, at least a little E/M and high energy physics, and much more. Keep in mind that no one else - including any of the teachers at Central High - had a clue about ANY of this! I'm sure they thought we were building a bomb...

    I also learned a lot about locksmithing (including all about making master keys using solder-fills and lock picking) - one has to be resourceful to succeed in these endeavors. The teachers were aware of this and kind of accepted it (as well as benefitting at times), realizing the limitations of an environment where advanced science was more along the lines of dissecting a worm. :)

    To be continued as I think of more tid-bits...

    Other Home-Built Cyclotrons

    I don't know how many of these were successfully constructed prior to the CHS Cyclotron in the early 1960s. However, there have been several, if not many, built since then both by individuals and as projects in various high schools and colleges. Check out the following links:

    Some brief references to our cyclotron may even be found by a determined Web search (at least one in the bio of a person who recalls having been involved). I welcome any info or links to the CHS Cyclotron or any other home-built cycltrons that have existed.

    The First Generation Central High School Cyclotron

    I recently heard from one of the original builders of the Central High School Cyclotron, Lawrence H. Zuckerman, Pleasanton CA, (Email: (, who was in the 218th class. Here is a paraphrase of our email thread:

    (From: Larry.)

    I graduated with the 218th class of Central High School. I was a member of the Cyclotron Society and helped to build this accelerator when I was a freshman in the '58-'59 school year. When we completed construction and performed a brief test, our work was publicized by the Heart Association, and a newspaper article with pictures of us appeared all over the country.

    I remember guys who were much older than me -- had just graduated in the 210th class and started college, such as Nathaniel Ostroff and Joshua Horowitz. I may think of other names later.

    Our faculty advisor was an MIT-educated physicist and later School District of Philadelphia science curriculum executive, was Fred Hofkin.

    (From: Sam.)


    Hey, wow! You're the first person from the original construction team I've ever heard from. I have seen a few Web references to other "personal" cyclotron projects, but nothing about Central until what I wrote.

    We were involved from '66 to '69 or '70.

    Anything you can add or correct would be most appreciated, with whatever attribution you'd like!

    (From: Larry.)

    It was Hofkin who told us that the cyclotron really only needed one Dee. I did not know any physics then and never thought about it all these years. Now that I am thinking about it, the single Dee architecture worries me.

    You mentioned that the magnet assembly is there on a 99 year lease. I vaguely remember that fact, and also that the "lessor" is the University of Pennsylvania, where I went to study Physics starting in 9/62.

    One of my tasks, assigned to me by Hofkin, was to determine the number of turns on the magnet (so that, using the current value, the field density and deflection could be calculated). I recollect (maybe incorrectly) that there were four coils, two slightly dis-similar pairs. He said I could use any method I wished, except for unwinding it. ;-) I measured the wire gauge to determine ohms/thousand feet. I measured the resistance of each coil. The total of 72,800 turns rings a bell, (but that was 50 years ago!).

    My other task was to build the control panel, which was nothing more than a relay rack panel with switches, chassis, and 120 Volt AC connectors. I did that OK, except giving the older boys a laugh by using bell wire; so Nat re-wired it with #18 zip/lamp cord, using both conductors in parallel.

    The cyclotron calculations (which I certainly did not do at the time) said that the frequency needed for the Dee was about 7 MHz. You mentioned a small ham rig for that, which could have been a DX-20. I don't remember the power amplifier.

    We fired up the equipment, opened it, and found a mark on the side in a plausible location. Was it from a proton beam? I think it was too far from the Dee to be an RF arc.

    The main (second stage oil diffusion) vacuum pump was donated to us by EIMAC (Eitel-McCullough), a fact mentioned in the AP/UPI released newspaper article that had the picture of the six of us standing around the cyclotron. I think this happened late in the school year.

    I shall check with the CHS Associated Alumni (of which I am a member) to get more information. If the cyclotron is still there, I would like to see it the next time I visit Philadelphia, which is once or twice a year. I have also been looking for information about Fred Hofkin.

    As a bit of irony, for the past fifteen years, I have been living about 10 miles from the (E. O.) Lawrence Livermore Labs. I am an RF design engineer at NATIONAL SEMICONDUCTOR.

    If I think of anything else or obtain more documentation, I'll send it to you. Nice work on your web site!

    (From: Sam.)

    Thanks for the info.

    Using the same basic calculations as you did, we concluded that the magnet was way underpowered by the original power supply and moved it over to the rectified pig-pole transformer supply.

    Hmmmm, let's see. Assuming a mean diameter of 2 feet and #20 AWG wire, your 72,800 turns is about 450,000 feet with 10 ohms/thousand feet for #20 AWG wire, that would be roughly 4,500 ohms. I seem to recall we measured 6,000 ohms, so not too far off. :)

    I built a 1 kW linear amp from the AARL handbook that was then driven by the DX-20 exciter. It replaced what was a home-built RF amp, supposedly about 200 W.

    That calculation of 7 MHz may have been correct for the weaker magnet, though I think it was even weaker than that. I seem to recall we eventually ran it on 1 or 2 kV, rather than the 200 V it was on originally. I'm not sure what the RF frequency we were running at, perhaps 20 MHz for a (20 kgauss) saturated iron magnet. But our only means of measuring field strength was by the student and hammer method - how many average CHS students it took to pull our standard hammer from the core when the magnet was energized! :)

    The single Dee thing always bothered me also, but I seem to recall other references to it, so perhaps it's not totally bogus, though probably less efficient.

    As far as whether it ever worked, I rather doubt any real proton beam from this machine would have had enough power to make a mark but who knows. We finally got it to expose a piece of dental X-ray film, and that was marginal.

    The most dodgy thing was probably the ion source due to its simplicity, had to be operated on the border between high vacuum and a pressure too high to sustain a beam.

    And I'm still amazed that the glass diffusion pump survived! :)

    As far as to whether the thing is still there, I have no idea. But who would want to move a 6 ton magnet!

    Anything else you come up with will be welcome.

    (From: Larry.)

    It is great that you guys resumed work on the cyclotron after my days at CHS and did so much; I knew nothing about this. Following my freshman year, I built much amateur radio transmitting and receiving equipment while in high school, but I have no recollection that anyone ever went back into the cyclotron room.

    I imagine you are correct that any proton beam from this machine could not make a visible mark.

    If the cyclotron was ever cleared out, it was probably to make room for a girls' lavatory!

    (From: Sam.)

    Or to add another water chiller. :)

    I think we were actually the third generation. Between your team and ours, there was Donald and several others.

  • Back to Sam's Laser FAQ Table of Contents.
  • Back to Amateur Laser Construction Sub-Table of Contents.
  • Forward to Vacuum Technology for Home-Built Gas Lasers.

    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.