*                  VARIOUS NOTES ON THE SULFUR LAMP                    *
  *                                                                      *
  *                       **** Version 1.02 ****                         *
  *                                                                      *
  *                      Copyright (C) 1996-1999                         *
  *                        Samuel M. Goldwasser                          *
  *                           Don Klipstein                              *
  *                                                                      *
  *                   Corrections or suggestions to:                     *
  *         sam@stdavids.marconimed.com or don@donklipstein.com          *
  *                                                                      *
  *                     --- All Rights Reserved ---                      *
  *                                                                      *
  *    Reproduction of this document in whole or in part is permitted    *
  *    if both of the following conditions are satisfied:                *
  *                                                                      *
  *     1. This notice is included in its entirety at the beginning.     *
  *     2. There is no charge except to cover the costs of copying.      *
  *                                                                      *

         ****************  INTRODUCTION  ****************

Sulfur lamp technology:

There has been quite a bit of publicity lately about a new technology
in lighting - the sulfur lamp.  While lighting is not normally thought
of as high tech, you may change your mind after reading these articles.

This document contains a collection of articles and discussions on various
aspects of sulfur lamp technology.

Microwave energy similar in power and wavelength to what your microwave oven
uses (microwave oven parts may actually be used in some implementations)
excite sulfur in an argon filled bulb (other gasses may also be used and
affect the spectral distribution).  The small bulb must spin as well as being
forced air cooled to prevent an instant melt-down.  The spectra is not quite
like daylight but is broad-spectrum - more polychromatic than most other
non-incandescent technologies.  In the current implementation, the bulbs are
very small (golf ball size or less) but are used to illuminate the inside of a
long light pipe which is actually used to distribute and diffuse the light.

The Smithsonian Air and Space Museum apparently has installed 3 of these
to replace over 100 high pressure discharge lamps with a resulting brighter
more natural illumination and reduced energy.  They kind of look like overgrown
fluorescent bulbs - a substantial fraction of the length of the exhibit hall.
The sulfur lamps and microwave exciters are at each end.

Unfortunately, it is not clear how well this technology will scale down
to residential use.  The excitation requires a microwave generator - magnetron
like in your microwave oven.  At the present time, the bulbs need to be
rotated continuously to distribute the sulfur/Ar mixture so there is also
a motor involved.  Hopefully, these problems can be overcome economically.

An interesting technology.  Stay tuned.

 ****************  ROB'S DETAILED NOTES ON THE SULFUR LAMP  ****************

(From: robpen@wseo.wa.gov (Rob Penney))


These notes are somewhat out of date at this time (1996), but cover the
fundamentals and include contacts for updated information.  I hope this
helps.  Anyone west of the Mississippi would like us to research this
further (the latest articles, papers, proceedings, info from manufacturer
and researchers), contact me at the address below and I can do this as part
of a contract with BPA and WAPA to support energy conservation for utilities
and their customers.


Exciting sulfur and quartz with microwaves creates great amounts of light
with similar properties to sunlight but without the ultraviolet component.

The light is distributed though light pipe for hundreds of feet, replacing
hundreds of conventional fixtures.  A smaller version may be installed in a
torch-type indirect lighting system.  The lamp itself may last indefinitely,
and the microwave generator may need occasional replacement parts.  Lumen
depreciation is negligible, and CRI will remain fairly constant.


Sulfur lamps consist of a golf-ball-sized sphere filled with sulfur, quartz,
and argon.  It is energized by a 5900-watt magnetron similar to that on a
kitchen microwave oven.  The spherical lamp is constantly rotated at about
600 rpm on a glass spindle surrounded by a jet of compressed air.  If the
lamp were ever to stop rotating, it would melt within two seconds.  The
technology is quite similar to a UV light source that Fusion Systems has
been selling to chip manufacturers and printers for 15 years.  Fusion is
planning to release more efficient, smaller models by early 1996, roughly
1000 watts and 140,000 lumens.  Lawrence Berkeley Labs is working on a
75-watt version of this for interior lighting.  They are also working on
making the magnetron smaller by using more solid state electronics.  The
smaller models will not use cooling air and would spin about 1000 rpm.  The
technology has the environmental advantage of using no mercury.

The light emitted is reflected by a parabolic reflector into a 10" light
pipe made of acrylic, prismatic film.  This pipe is almost opaque on top.

The bottom is made of many parallel, curved, reflective grates which catch
some of the light and reflect in down and out to the sides.  The ration of
how much light goes down and how much out to the sides can be varied to meet
design needs.  How much light goes out altogether varies along the length,
with more allowed to pass through farther from the light source and less
near the light source, to create more uniform luminance along the length.

The light pipe would therefore need to be purchased in sections, each with
specific characteristics.  A mirror at the far end of the pipe reflects back
any light traveling that far.  Smaller models may not use light pipe, either
using a more standard fixture or possibly fiber optics.  One such
application being considered is to install the light on a 7' tall pedestal
in an office cubicle area creating a powerful indirect lighting system.

Light output:

It emits 450,000 lumens, 310,000 of which are reflected into the light pipe.

The spectrum is closer to visible light than most conventional lighting
sources.  The chemistry of the lamp can be varied somewhat to adjust the
exact light spectrum.  Light output of lower wattage versions would be less.

Health effects:

There is a greatly reduced component of damaging ultraviolet light.


The efficacy of the lamp itself is 450,000/5900 = 76 l/w.  If you consider
the lamp reflector as part of the lamp, the efficacy drops to 310,000/ 5900
watts, so 53 lumens per watt.  The light pipe is roughly 60 percent
efficient, so the efficacy of the whole fixture is 31 lumens per watt.  That
would be reduced further if the system gets dirty or is not properly
maintained.  This does not compare well with other light sources which have
efficacies up to 180 lumens per watt, although the CRI of the sulfur lamp is
greatly superior to such other lamps.  Looking at fixture efficiency, this
would be 0.7 (reflector) times 0.6 (light pipe), producing a fixture
efficiency of 0.42.  This matches very closely that measured by LBL.

Fusion hopes in increase lamp efficiency considerably.

Life expectancy:

The sulfur lighting system is currently rated to last 10-20,000 hours, but
this is a rough estimate.  Because the components in the lamp do not
chemically react and it has no electrodes, the life of the lamp itself
should be quite long.  What would probably fail is an electrical component
of the magnetron.

Electrical/mechanical maintenance:

Because one sulfur lighting system can replace several hundred conventional
light fixtures, maintenance can be greatly reduced.  In an area with an
inaccessibly ceiling, this can be an attractive feature.  The light pipe
itself needs to be cleaned periodically, probably with something on the end
of a long stick.  The electronic components in the magnetron will eventually
need replacing, but that can all be located in a easily accessible spot.

Lumen maintenance:

Again, because the components in the lamp do not chemically react, light
output and quality should remain unchanged.  However, if the light pipe is
not kept clean, the effective light output will suffer.

Other performance issues:

Many of those who witnessed the first installation of a light pipe system
were distracted and surprised by the noise of it.  This was primarily due to
the cooling system, probably an air compressor which are notoriously noisy.
Using a single source, a large area could lose lighting if the light source
failed.  Systems should therefore be designed with redundant light sources
with automatic backup.


Products were expected to be available at the end of 1995.  Cost estimates
are unknown, but the system installed at DOE headquarters was reported to
cost one-third that of the mercury vapor system it replaced.

Expert resources:

The folks most on top of this new technology are with the manufacturer (Kirk
Winkler at Fusion Lighting 301/251-0300) and with Lawrence Berkeley Labs
(start with Francis Rubinstein 510/486-4096, FMRubinstein@lbl.gov).  LBL is
doing a lot of research for DOE on applications for this new technology.


Fusion Lighting, Inc., of Rockville, MD, a privately-help spin-off of
Fusion Systems Corp., makes the fusion lamp.  301/251-0300.  Kirk Winkler,
x5553.  A.L. Whitehead of Vancouver, BC, makes the light pipe.

          ****************  ITEMS OF INTEREST  ****************

This is the only slightly edited transcript of an email discussion between
Sam (>) and Don.  (From: Don Klipstein (don@donklipstein.com)).

> When will we see household sulfur lamps?:

My answer is, not any time soon.  Consider the electricity cost of 
operating compact fluorescent lamps a few hours a day, and maybe the cost 
of the bulbs.  How much would you invest up front to cut the electricity 
costs by 50 to 60 percent?  The return should exceed that of competing 
investment opportunities.

There are quite a few minor technical hurdles.  The sulfur lamps in use 
now are 5.6 KW (or is that 5.9 KW?) units of golf ball size.  The Fusion
Lighting Co. (unsure of exact name) is working on 1 KW units.  I am guessing
that using a xenon-sulfur mix instead of an argon-sulfur mix might reduce
heat conduction enough to reduce the bulb's diameter by (optimistically) a
half to two-thirds.  This would would reduce the power to around 100-200
watts.  If you blow a jet of air at the bulb to cool it further, they might
be scalable down to the point that power input is only a few times the heat
conduction loss.  I am guessing 30 to 50 watts, as a number out of a hat.

Sulfur bulbs also have a quirk having to do with convection.  The 5.6 
KW bulbs must be kept rotating.  Otherwise, a major hot spot will develop 
at the top of the bulb, destroying it in something like 1 or 2 seconds.  
Use of xenon instead of argon does not help this much.  On a smaller 
scale, convection MIGHT not be as bad, but I suspect the lamp will still 
need a motor.

Another hurdle is getting 50 watts of microwaves into a target the size 
of a pea.  I doubt this can easily be done at the 2.4 GHz or so frequency 
of microwave ovens.  One would need a much higher frequency probably well 
over 10 GHz.  And the microwave source must still be economical, 
efficient, and reliable.  And all of this must be done in a manner 
satisfactory to the FCC.  I don't know if there are any bands in the 
10-30 GHz range where such microwave use is permitted.  Of course, the 
regulations can be changed if the need is great enough.

Since xenon does not ionize as easily as argon, an auxiliary means of 
"igniting" the bulb might be necessary.  This might be some sort of Tesla 
coil, flyback transformer, or trigger coil type of device.  Not too 
expensive once someone gets in the swing of making the cheapest thing 
that works, but it is a minor extra expense and possible aggravation.

Meanwhile, what would be "ignited"?  The gas in the bulb, or the air 
outside it?  Might be a problem if the gas in the bulb has to be at a 
really high pressure, and I have little idea what that might be.

Another consideration is the color of sulfur light.  Generally, the 
color temperature is high.  I saw a wide range of 4000-10,000 Kelvin 
somewhere (see below for where), but they said it works best with color 
temperatures in the middle and upper portion of this range.  A color 
temp. of 5500 K is an icy pure to slightly bluish white.  6500 K is 
definitely a bit bluish; this is the color of "Daylight" fluorescent 
lamps.  Maybe good for outdoor use away from astronomers, but not a 
popular color for illuminating a living room.  Furthermore, sulfur lamps 
are a bit greenish compared to a blackbody source.

As for filtering this light, maybe things aren't too bad:  The #85 
Wratten filter is about two-thirds transparent to 6500 Kelvin light, and 
converts it to around 3750 Kelvin.  A filter gel to convert 5500 to 3750 
would be even better, if a mini sulfur bulb can efficiently produce 5500 K 
light.  If fluorescent materials could be employed to convert some of the 
shorter wavelength stuff to red light, things get even better.

If something can be made for under 100-200 dollars and be satisfactory, 
we might have something.  Otherwise a mini sulfur lamp would be just a 
curiosity, conversation piece, or suitable for a few special purposes.

For some bits of info about sulfur lamps, check:

http://www.webcom.com/~lightsrc, (found not working 3/1/2023)
and find the part with the "archive" of older articles.  The one about
sulfur lamps is available for a month every several months.  The "archive"
rotates in and out some of the more popular articles every month.

Again, I don't expect to find any sulfur lamps in the nearest home 
building supply store any time soon.

Discussion on the feasibility of a homemade sulfur lamp:

>So when can we build one?

I thought a bit more on sulfur lamps this morning.  Don't see problems 
at frequencies higher than already used, in terms of microwave 
penetration (should be fairly constant as freq. increases past what works 
well) or reflection by the plasma (should be even less as freq. increases).
Possible problem with small bulbs is getting microwaves absorbed fairly 
completely by a tiny plasma, but adjusting the fill gas pressure will 
probably fix this.

As for convection, it not only heats the top of the bulb but also 
transports heat from the plasma to the bulb.  This may be a significant 
energy loss at lower power levels.  Efficiency of smaller bulbs may be 
significantly improved by rotating them to prevent convection currents.

How to build one?  Hardly looks like a DIY to me.  Takes quite a bit of 
doing to blow a good strong bubble out of quartz.  Its trickier than 
glass, and also needs higher temperatures.  I will check into this in the 
library when summer approaches, if there is demand for info on how to do 
this in your basement.

MY ADVICE:  Don't try this at home.  Required materials and equipment 
will probably cost thousands of dollars.  You need lots of patience and 
AT BEST some tricky glassblowing.  Prepare for bulbs to explode if flawed.
Stick to Tesla coils, they're easier.

> Would a low power sulfur lamp need to be smaller or could the same
> 1 inch or so bulbs be used?

Underpowering a 1 inch bulb would cause 2 problems:

1. The thermal conduction loss from a plasma at a normal operating 
   temperature is roughly proportional to the diameter of the plasma.  I 
   believe this would be a surprisingly constant fraction of the bulb's size.

2. Underpower the bulb enough, and the plasma temperature drops, probably 
   shifting the spectrum to less visible wavelengths and possibly also 
   causing an undesirable color shift (maybe from greenish blue-white to 
   whitish green-yellow).  Then again, the color might be like that of a gas 
   mantle, which isn't too bad.

   However, I suspect that efficiency and color may be only mildly 
   impaired by operating a xenon-filled (instead of argon) 1-inch bulb
   (1-KW size??) at something like 100-200 watts.  Nice idea.

> So, maybe they buy the premade bulbs.  Just add a microwave oven
> and stir!  Sounds like a recipe.

Sounds nice.  May only be able to be sold as part of a kit with strong 
warning statements.  Consider the similarities to HTI, HMI, and
short-arc/compact source mercury and mercury-xenon bulbs.  Of course, the 
pressure may not be very high and possibly not much UV gets through 
sulfur vapor, and maybe lack of electrical connections makes the 
construction a bit simpler and sturdier, and there is nothing toxic or 
corrosive (at room temperature) inside.  This makes them a bit safer, but 
the Consumer Product Safety Commission might not let anyone sell them where
Joe Sixpack would buy them.

However, I like the idea of somebody selling them in kits or through 
mail-order.  I would probably buy one.  I would probably put it in my 
microwave oven and see what happens (Goggles on face, fire extinguisher 
in hand?).  If nothing breaks, I might trash-pick a microwave oven or buy 
the cheapest junky one, take it apart, and build a working sulfur lamp.

Yes, I like your idea.

Probably has to wait until sulfur bulbs are produced in great enough 
quantities that some could be diverted to hobbyists, or spare bulbs 
become available from whoever sells replacement bulbs (I doubt they last 
absolutely forever).

> So, take the envelope from a burned out HMI bulb (hey, talk about warnings!),
> back fill with sulfur/argon or whatever.  I have a couple of vacuum pumps
> that would probably be good enough.  The tough part would be the fire
> extinguisher. :-)

Can't do this with anything that has or had electrodes.  Hot sulfur and 
sulfur vapor are corrosive to most metals.

Got me thinking however...

If you take a quartz tube and heat one end, you should be able to 
squeeze it shut.  This will need oxy-something.  No torch using any 
combination of air and propane or MAPP gas seems to be hot enough.  
Quartz takes at least 1600 Celsius or more to be worked.  I tried this 
with some tubing from a toaster oven.  After closing one end, go over
the end with the flame and melt it somewhat to be sure it is closed.
After that, do the same with the other end.

If the proper fill gas pressure (or one that we can make work) is 
atmospheric pressure (as measured when the quartz is being worked), then 
WE ARE IN LUCK.  Just blow gas through the tube before closing it off.
Get a bit of sulfur in there first.  Try to work the quartz such that its 
overall temperature distribution makes the gas pressure the same as when 
you anneal the darn thing afterwards.  Annealing requires baking the bulb 
for something like a day at 1140 Celsius or a bit more, with the gas 
inside at atmospheric pressure.  (Or match bulb and oven pressures.)

Or, push your luck and operate the bulb without annealing.  Quartz has 
very nearly zero thermal expansion, so an unannealed bulb just might not 

If the fill gas must be at some odd pressure, then one must seal both 
ends of the bulb, poke a hole in it, and attach a hollow stem to it.  One 
of us will have to look up how HMI or similar bulbs are made.  Then comes 
the time to anneal it, then dump in some sulfur, then vacuum, gas, and 
seal and pinch off the stem (easier below atmospheric pressure than above).

Since there is no metal, we don't have to worry about corrosion by 
contaminants such as oxygen or water vapor.  Traces of either of these
would be a big problem in bulbs with metal parts inside.  However, these
DO impair starting, and should be minimized.

> Maybe someday.

Seems interesting.  I have always wanted to build my own high intensity 
discharge bulb, although most of my life I thought in terms of mercury 
vapor to do this.

          ***********  LINKS TO OTHER SITES  ***********

  Another sulfur lamp FAQ at http://www.sulfurlamp.com/index.htm 
(at the Wayback Machine)

          ****************  REFERENCES  ****************

(In no particular order.)

1. "Sulfur Lighting:  Emerging Technology Could Challenge HID Light
   Sources,"  E Source Tech Update 94-7, September 1994.  Call 303-440-8500.

2. "Electrodeless Lamps:  The Next Generation,"  Lighting Futures vol. 1,
   number 1, May/June 1995, Lighting Research Center. Call 518-276-8716.

3. "A Light to Replace Hundreds of Bulbs", by John Holusha, New York Times,

4. "A New Kind of Illumination That Burns Brightly, but Not Out", by Curt
   Suplee, Washington Post, 10/24/94.

5. "A Quick Look at the Sulfur Lamp", by the Lighting Design Lab.

6. "Energy Department Brings Dazzling Bulb to Light", by Curt Suplee,
   Washington Post, 10/21/94.

7. "DOE Unveils Revolutionary 21rst Century Lighting Technology", a press
   release by Hope Williams and Keith Holloway, U.S. DOE.

8. "DOE Unveils New Lighting Technology", from Femp Focus, Dec. 1994.

9. Journal of Illuminating Engineering Society vol. 26 number 1 Winter 1997
   Two papers written by authors affiliated with Fusion Lighting, Inc.

-- end V1.02 --

Back up to Don's Lighting Page.

Back up to Don's Home Page.