Most of the major differences are with respect to the power supply. Rather than a simple rectified neon sign transformer and Oudin/Tesla coil trigger, it uses a 15 kV thyratron controlled pulsar in the basic power supply and also includes a more complex design capable of better control of pulse width and energy. This 'cascade' system includes a high voltage trigger and lower voltage pulse forming network - similar in some ways to what is used to drive a xenon strobe or solid state laser.
When used with argon or krypton, some of the same wavelengths are produced as with normal Ar/Kr ion lasers. I assume this to be the case with xenon as well but I don't have any data for a xenon ion laser. The Xe lines lase with both the simple exciter as well as the cascade system though most of the Ar and Kr lines require the cascade system.
I also don't know if the O2 lasing process is similar - it could very well be a totally different mechanism being a non-noble gas diatonic molecule. No data is available for O2 with the cascade system. This indicates that the O2 lines may not lase with a wide, lower voltage pulse. In this case, it could be lasing in the afterglow (as has been suggested by some people) which would only operate with the short HV pulse provided by the simple exciter's thyratron directly.
With pure neon (Ne), the PMG laser should produce coherent light at the orange (611.9 nm) and yellow (594.1 nm) wavelengths of the corresponding non-red HeNe lasers. This is discussed in more detail later in this chapter. And, you are of course free to try whatever other gases you desire!
The PMG laser is also very similar in many respects to the Ar/Kr ion laser in terms of the difficulty of construction. A flowing gas system like that of an axial flow CO2 laser is suggested - a mechanical pump at one end of the laser tube and a controlled gas leak valve at the other end. A medium vacuum system and some glass work will be needed. However, since easily obtained (or generated) O2 can be used, the gas handling can be much simpler than for the Ar/Kr ion or HeNe lasers.
Avoid eye contact with the direct or reflected beam. This includes the 4 pairs of beams reflecting off the Brewster windows which may be quite strong.
WARNING: Do not use a high rpm direct drive type vacuum pump with pure oxygen. It is better and safer to use a slower rpm pump like the older Welch 1402 or 1397 belt driven pumps. The concern is that due to mixing of O2 with standard vacuum pump oil, the higher rpm pumps could, as they say, EXPLODE! So, a different (and very expensive) type of oil called 'Fomblin' needs to be used.
For more information, see the chapter: Laser Safety. Sample safety labels which can be edited for this laser can be found in the section: Laser Safety Labels and Signs.
The following site may have some useful information, if not explicitly for a PMG laser:
The following may be useful in conjunction with PMG power supply design:
Much of the information in this chapter comes from the paper: "A Cold Cathode Pulsed Gas Laser" by R. K. Lomnes and J. C. W. Taylor in: Review of Scientific Instruments, vol 42, no. 6, June, 1971. Thanks to Daniel Ames for to extraordinary effort in digging out this material.
I have simplified the laser tube somewhat (more along the lines of the SciAm designs). The authors likely had access to a well equipped machine shop to fabricate the flanges and what-nots for the their laser head. So, while it has some nice features including ease of disassembly and parts replacement - everything is screwed together with O-ring seals and the like - more than basic tools and metalworking skills will be needed to duplicate their design. However, if you have access to a good machine lathe and know how to use it, know a friendly machinist who owes you some favors, or have unlimited funds, go for it!
I have also updated the cascade system to use a solid state rectifier rather than the mercury vapor tube that was state-of-the-art in 1971 as well as suggesting alternatives that do not require a thyratron.
While I cannot put this paper on-line due to copyright restrictions, I will be happy to send a most excellent fabulous photocopy of a somewhat mediocre photocopy (best I have, sorry) to anyone who provides a stamped self addressed business letter size envelope. Contact me via the Sci.Electronics.Repair FAQ Email Links Page for instructions.
However, unless you are intent on duplicating their design exactly (or are offering to draw the tube assembly in AutoCad to be included here), the amount of additional solid information in the paper is quite limited. In fact, the amount of solid information in the paper is quite limited - period! The narrative is in many ways much more like that in the "Amateur Scientist" articles of Scientific American than a true research paper. I hope this was due to space limitations in publishing and not to a lack of actual results. One possible conclusion is that assuming the authors actually got the laser to work as advertized, they knew very little about what they were doing!
In any case, the section: Home-Built PMG Laser Description includes all the important dimensions, materials, and power supply schematics so duplicating the laser in terms of function, if not style, should be straightforward.
Another interesting and possibly relevant paper (also unearthed by Daniel) is: "Super-Radiant Yellow and Orange Laser Transitions in Pure Neon" by H. G. Heard and J. Peterson, Proceedings of the IEEE, Oct. 1964, vol. #52, page #1258. This discusses a PMG-like laser using neon which will lase superradiantly in a narrow tube (e.g., 40 cm long x 1 mm ID) in the orange (611.9 nm) and yellow (594.1 nm) with orange being the strongest. Superradiant means that no mirrors are used although the addition of a Fabry-Perot cavity does improve the lateral coherence and output power. The authors used a pulsed high voltage power supply for excitation (they didn't attempt to operate the system in CW mode but speculate that it should be possible). So, if you have a source of neon, try that as well and report back with your results! :)
A wealth of information can also be found on the oxygen ion laser in U.S. Patent #3,928,820 - High Gain Pulsed Ion Laser. See the section: Daniel's Oxygen Laser (near the end) for the abstract and some of the major topics covered in the patent.
Another one worth looking at is U.S. Patent #3,646,476: Pulsed Gas Ion Laser. It has a good description of a somewhat practical low rep rate PMG built with Pyrex and an indium cathode.
The only photo I know of is in the paper listed above. However, my mediocre photocopy doesn't show much (just a decent approximation of a black smudge) and it really isn't worth the cost of postage if that's all you what!
Refer to Typical Home-Built Pulsed Multiple Gas (PMG) Laser Assembly for a simplified (very rough) diagram of the overall glasswork and power supply electronics (the complete cascade system).
<------ Wavelengths (nm) -------> Gas Simple Exciter Cascade System ------------------------------------------------ Oxygen 464.9 559.2(1) Xenon 496.5 496.5 499.1 499.1 515.9 515.9 526.0(1) 526.0(1) 535.3 535.3 541.9(1) 541.9(1) 597.1(1) 597.1(1) 627.1 627.1 Krypton 520.8(1) 530.9(1) 568.2(1) Argon 472.7(1) 472.7 488.0 496.5 501.7 514.5 Neon(2) 543.5 594.1 611.9Notes:
According to Steve Roberts there is also a line at around 545 nm (or should this be the green HeNe line at 543.5 nm?) which will lase brightly in a transverse discharge similar to that of the nitrogen laser and also lase in a long tube but not as strong.
More information on these power supplies as well as alternatives are provided in the sections starting with: Power Supplies for the PMG Laser
Below are some estimated values for the PMG laser based on the RSI paper and a (70 cm) cavity length. I find it very hard to believe that with O2, the electrodes and the Pyrex cavity walls showed no significant signs of deterioration after 100 hours of use, but then again, the article does not give any specs as to how long each of the four gases including krypton were tested proportionally during those 100 hours with the "simple exciter" power supply.
Pulse repetition rate versus run time at 1 us/pulse:
Well, for this one, there aren't any real guidelines except for those dealing with safety - this is the true experimenters' home-built laser! :)
For argon, krypton, and xenon, there isn't much choice - you have to buy a small (lecture bottle?). There simply aren't any pure sources of these gases available in common household appliances.
Oxygen can easily be obtained in the same way or you can generate it using a variety of chemical methods. One simple one I remember fondly was to mix hydrogen peroxide (from your drugstore) with manganese dioxide (from dead carbon-zinc batteries or a chemical supply house).
Ordinary air can be obtained, well, you know where to get that! However, keep away from your local politicians due to the contamination and excessive temperature (which affects lasing). :)
Home-Built PMG Laser Using HeNe Laser Tube shows one possible configuration driven by the 'Simple Exciter' described below. This would make a particularly nice setup especially if the tube's mirrors are already known to be properly aligned.
Depending on the gas fill, a diffusion pump or ion pump (in addition to the mechanical pump shown) may be desirable but probably not essential, especially if the flow-through gas system is used. However, where the vacuum connection is only made at one end of the tube, a better vacuum system would make life easier by reducing the number of pump-down, back-fill cycles needed to achieve necessary gas fill purity.
As an example of a conceptually similar approach, see: H. W. Morgan, P. A. Staats, P. M. Griffin, G. K. Werner, K. L. Vander Sluis, "Conversion of a Simplified He-Ne Gas Laser to Pulsed Operation with Ar, Kr, and Xe, American Journal of Physics, Volume 37, Issue 9, pp. 938-939 (1969). This paper assumes a tube with Brewster windows so the mirrors are external, but where the relevant wavelength(s) are close those of a suitable HeNe laser tube, internal mirrors are a lot more convenient!
(From: Daniel Ames (firstname.lastname@example.org).)
Well, as you know, making a laser from scratch or even modified parts is a big undertaking, especially for one that uses more than 15 kVDC and has very little existing info and specs. A Challenge. :)
Here is the present status:
(About a month passes.....)
Well, I finally got some pics taken of my present oxygen laser (or PMG laser, take your pick). It's not finished as yet. As with most home-brew laser projects, it seem as though the closer I get to completion the, more considerations or obstacles I encounter. :)
A simplified diagram of the electrical configuration is shown in Danial Ames' PMG (N2/O2/Air) Laser 1.
(Several months later, semi-success but partial false alarm.....)
Well, I finally finished up the small version of my O2 PMG Laser. I had been trying to get it to lase for three (3) days now and just when I was ready to shelve it again and start building a much bigger tube... Drum roll please! I am very pleased to announce that - on April 10th at 3:00 pm PSDT, the Oxygen Laser is LASING = YES!!!!!!!
It actually LASED. :) A beautiful violet/blue, probably the 441.4 nm or 441.6 nm lines - love those wavelengths, or maybe it's the 464.9 nm line, but it appears to be more the color of a 442 nm HeCd laser.
I still have some modifications to work out with the power supply, but hey, it lases! YES, i'm ecstatic!!! After a little more than 1 year of research, design, and manufacturing, I can hardly believe it. I will be adding photos and more details later. Right now, the output power is quite low and unmeasured, but it is very likely that by modifying the power supply and the OC optic that the output power will greatly be increased. Crowd goes WILD!!!
(Two days later.)
As you know I reported that my small O2 laser was in fact lasing in the Violet/Blue reign. Upon further testing and evaluation, I am now somewhat doubtful that it was the oxygen that was lasing. :(
At first, the only way this 7.5" cavity would lase was to close off the O2 supply and open both valves from the tube to the vacuum pump. At this time, the tube had an air leak of about 20 Torr in 5 minutes. Upon fixing the leak, I could not get any sign of lasing at all with just O2. Then I found that by drawing in plain old air into the tube, it would lase again, still only in a pulsed mode.
Here are the present operating parameters:
I found that with one of the optics removed, all lasing stopped. I tried substituting a multiline argon HR mirror for the original 442 nm OC - again no lasing. In general, the optical alignment had to be correct or all lasing would stop. I also tried substituting an uncoated blank optic having a 44 inch RoC for the original HR and had similar lasing to the original 442 nm pair.
I also tried N2 and air with very negligible difference using all optic and spark gap combinations above. The gas pressures were similar within the lasing high and low limits.
My conclusions at this point: I suspect that it was the N2 in air that was actually lasing, NOT oxygen as I found that I was fooled by the target, (plain old white printer paper). This paper was actually fluorescing the violet/blue color that I originally thought were the oxygen lines 441.4 and 441.6 nm. Upon further investigation, I substituted a non fluorescing target and found no visible output. Oxygen does have at least 4 UV lasing lines, but since this laser's output was very similar using air or N2, I have for now, ruled out that it was lasing the O2 UV lines, but instead, it was producing a coherent UV beam from the pulsed N2.
I know from testing this laser with varying parameters, that the UV lasing requires a very fast pulse, Sorry but I am not set-up yet to measure the pulse width. However, the electrical connectors to the tube are made from 1/2" wide brass for a shorter pulse delivery. All the other connectors to the caps and spark gap are also 1/2" wide and as short as possible. This setup works well, but when I add an inductor into the anode circuit (20 feet of 18 gauge copper wire), the lasing ceased. Since it does lase (pulsed only) with either N2 or just plain old vanilla-flavored air, the obvious assumption would be that it is the N2 lasing as in an N2 Blumlein type laser, unless I am getting different UV lines for air than with just N2. If so, than maybe the N2 in air is transferring some energy to the O2 atoms? or helping the O2 atoms to return to their ground state? I have never heard of an N2 Ion laser. :)
For now, my only plans for this very short laser will be to try to increase the voltage and try again with O2.
One thing is that I wouldn't expect the mirrors to matter that much with N2 as I understand it so maybe you have discovered something new. :) Of course, if it is indeed UV, those mirrors really shouldn't do much anyhow - about the same as plain glass which I think is what you found out. But the fact that alignment is critical is at the very least, interesting....
Don't rip it apart - there is something strange going on and it deserves a proper investigation.
(From: Jon Singer (email@example.com).)
There is an N2 ion laser which runs at 428 nm, and could be what Daniel was seeing. In fact, Diane Neisius mentions this laser further down on the same page, though she refers to it as He-N2+.
I also think Mark Csele also mentions it on one of his pages -- if his students don't put quite enough N2 into the excimer laser, they sometimes see the visible line.
(From: Steve Roberts.)
Wow, what a problem! N2 is usually superradiant, if your on the traditional N2 line, the only thing I can think of is the mirrors are just kicking it past the lasing threshold. Is the pure N2 power (measured on a scope, not eyeball) greater then the air power? Air impedes the pulsed N2 line. Also N2 doesn't like low pressure.
Most of the pulsed gas lasers reside in the low pressure area which explains why you would get lasing when the valves are locked off and the pump fluid is outgassing. Pump fluid causes the pump to hang at 100 microns or so for a while, till the air and water vapors are expelled from the pump oil by mechanical heating. That 100 microns to 250 microns is a good area for pulsed lasing.
Tough question. I don't know what you have. You need to find a old UV-VIS spectrophotometer from a hospital or similar place and rip the calibrated monochromator out of it. I think we can rule out the pulsed argon UV lines, it's only a very small partial pressure of air. Unless it's the N2 molecular line, you have a ion line going there or God knows what else.
I need to think about this for a while......
(Several weeks later.)
If it is lasing the N2 in the air, then according to Steve Roberts, it is lasing in the afterglow and not superradiantly. But O2 is said to hinder N2 from lasing. Go figure again. :)
I do not get any UV lasing using the same pressure with just O2, so if it is lasing the O2 in air, then I would propose that the N2 is required for similar reasons that HeNe lasers need He and/or CO2 lasers need N2 + He. In the case of O2, possibly it provides a molecular vibrational energy transfer from the N2 to the O2, or the N2 helps the O2 return to it's ground state before the next pulse.
The strange thing is that it does not lase on every pulse especially at higher rep rates. At 1 to 2 pps, it lases on most pulses, but at, say, 10 pps, or higher (up to 60 pps tested) it's a hit and miss situation.
The funny thing is, I was not and am not attempting to achieve a UV output, although O2 does have (4) UV lasing lines at 298.43 nm, 374.949 nm, 375.470 nm, and 375.985 nm . (Reference: Journal of Applied Optics, vol. 4, no. 5, May 1965, pg. 576.)
PS. My next O2 (P.M.G) laser will have a 115 cm length with an ID of 7mm.
Hmmmm. The only way to really narrow this down would be to determine at which wavelength(s) it is actually lasing. The three near-UV wavelengths listed above would certainly pass through a glass or plastic prism or diffraction grating; the one around 298 nm might. They would all work with a reflection grating. So, there may be hope of determining at least which grouping the wavelength is in or if it is actually somewhere else.
Right you are. I fired it up today with an Edmund Scientific transmission grating. At a distance of 68-1/8" from the grating for the 0th order spot, the 1st order spot is 6" (+/- 1/16") away. The spot from a HeNe laser using the same setup is 11-3/8 (+/- 1/16").
Using the approximation that sin(x)=x for small x, the ratio of distances is equal to the ratio of wavelengths. This results in a wavelength of about 342 nm. Hmmmm, pretty close to the 337.1 nm N2 line. :( (Without using the approximation, the result is about 345 nm.) If your measurements are off by 1/16", it could be exactly the N2 line.
Not bad at all. I figured it was probably the N2 line in air, but I did find some interesting ways of stimulating it, such as lasing it in the afterglow and adding O2 at a certain N2 pressure to amplify the output. So, I guess this laser does fit the description of a PMG laser after all. If only I could find that darn air leak, I have isolated it to the laser or it's vacuum line connections, I guess I will just have to develop the poor man's helium leak detector.
I made a more precise measurement and now come up with a wavelength of 334.3 nm. I also drew a cool diagram: Determining Relative Wavelengths using Diffraction Grating (currently not available).
Did you know that N II lases at around 350 nm? I found a reference to it back in 1964.
Interestingly, with this lumped capacitor longitudinal Blumlein type discharge, the length of the inductor (as used with traverse Blumlein N2 laser circuits), is unusual. I am using two caps with a common negative and the spark gap is across one of the caps. The 2nd cap is discharged through the cavity just like the Blumlein N2 lasers, except the discharge is longitudinal and the caps are not the flat plate type, but rather industry standard pulse caps.
I found that by using the larger (.05 uF, C2) cap for the spark gap and the smaller (.03 uF, C1) cap discharging through the cavity, that the output was significantly increased. The inductor lengths tried across the two + sides of the caps at 18 to 22 kVDC with a pulse rate of 1 to 2 pulses per second with pressures ranging between .5 to 5 Torr:
The pulsed fluorescent blue/violet beam spot of this laser - as viewed on plain white printer paper at a distance of 3' and using just plain ole (air) as the lasing medium is now so bright that if it were continuous, you would not want to look at it without laser safety goggles at 400 to 450 nm and UV blocking. Within a certain pressure range after the air starts lasing, adding a small amount of O2 increases output even further.
I was looking through my assorted optics tonight and found some dielectric (UV) flat bar type mirrors. I tested them with this ?UV beam for reflection and transmission. Reflection was great, transmission was zilch meaning 0 detected... hmmmmmm... I might try one of these for the HR since now I have two very bright UV beams exiting the laser, one at each optic. I would love to try lasing a dye in a quartz cell, without a resonator. It just might work as the comparable output of my home made PMG laser on the same white fluorescent target appears much brighter than a commercial TEA N2 laser I had which would lase superradiantly a dye without a resonator.
Assuming the inductor is in the same place as for the N2 laser, I bet it isn't behaving like a simple inductor when you get to the really long configuration - it has capacitance as well and is shorting your cavity.
What happens if you replace the inductor with a string of resistors? For the Blumlein, the inductor should really behave as an open circuit when the spark gap fires. A resistor of a few K ohms should do the same.
I'm also not really sure why you need C2 (the capacitor across the spark gap) at all for this configuration.
I haven't tried a resistor as a substitute for the inductor yet, but I do know from my experimentations that a 50' 18awg inductor that is wound (layer upon layer) on a small spool similar to the ones used in pulse forming networks for flashlamps and for pulsed argons but no output. 50' of 18 AWG wire is approximately. .31925 ohms, inductance was not measured. It would seem as though that with these larger cap values (larger uFs than most TE N2 lasers) that possibly the inductor plays a roll in shutting down the pulse through the cavity? Since O2 is said to lase in the afterglow, a sharp drop off of the ionization pulse, on the trailing edge is needed, that is why the Canadians used a (donated) thyratron.
As for why I used C2: It's because at these low pressures, below about 1 Torr, the gas in the cavity would continuously (barely) ionize even when the spark gap was in the (hold/off) cycle. It could be that a different spark gap and triggering source might solve the problem.
I still suspect that your larger inductors are acting more like capacitors. :(
You really need to measure the pulse duration of your discharge. The usual way to do this is to put a low value resistor (.1 ohms, .01 ohms, etc.) in the discharge return path and measure voltage across it. But you can't do that with your setup because there is not ground point in the discharge path.
I'd suggest constructing pulse transformer - put one of the conductors going to the tube through a ferrite core. Wrap a few turns of insulated wire around the core and terminate those with a low value resistor (a few ohms). Put your scope across the resistor.
I now have different optics in this laser (same one as my Pulsed Multiple Gas Laser Project). The UV output is now stronger than ever using just AIR!!!!!! No exaggerating - the output of my air laser is tremendously brighter when viewed on plain white printer paper than my commercially built TEA N2 laser (using the same paper) with it's very expensive pressurized and triggered spark gap using pure N2. Note: The TEA N2 laser has lased some dyes to superradiance without a resonator.
My air laser's output at a distance of 5" from the cavity no longer has the bright (blue) fluorescent color on white paper. Now it is so bright that the bright blue is starting to take on a white/blue fluorescent color. The TEA N2 laser using pure N2 never did that. The real puzzling part is this is that the air laser actually seems to like the air combination of 80% N2 and 20% O2 with an additional 1 to 3 Torr of O2 with an approximate ratio of air + O2 up to about 1:0.5.
I now have a fabulous animated version of my PMG laser. Check out Daniel Ames' PMG Laser Operation (currently not available). Since this was designed and drawn for pulsed gas lasers that lase in the afterglow (i.e., the trailing edge of the ionization pulse) the stimulated emission is shown after the first glow of ionization.
I have succeeded in lasing oxygen with a little bit of green 559.1 nm from this little converted argon tube. Verified that it is actually visible output and not just week UV fluorescence on white paper appearing to be green. That was easy, just insert a piece of (most any clear plastic/acrylic) into the beam's path. If the lasing vanishes, it is a visible wavelength, easy, huh?
Back to the 559.1 nm lasing, the output so far is very very weak, probably no more than .05 mW and it is very very critical to maintain the exact mirror alignment.
Congratulations! Even .05 mW of visible is fabulous! But, yes, absolute cleanliness is likely to be critical for the lower gain visible wavelengths.
It is possible that different optics, or spacing out these optics much further, or changing the ionization pulse width, or changing the phase of the moon :), etc. might also make a positive difference in the output. Their RoC is 2 m. So they are way to long for this short (19") resonator for normal alignment. I am using two HR mirrors, so both have the 2 m RoC. The output beam waist diameter is very small, perhaps 0.2 to 0.3 mm. The bore diameter is approximately. 1.2 mm. I did try the HR and OC together, but no lasing found yet with this combination.
I suspect that it would lase more strongly if it was not for some brass deposits inside the cathode's Brewster stem and possibly on the inside of the Brewster window as well (although the latter is not noticeable). So, Material Recommendation coming: Don't use brass for hole plugs, electrodes, or anything else that the discharge and it's afterglow overglow can come in contact with. The original brass (non-OEM) gas inlet fitting on the anode did not cause this problem, but I used brass 1/8"NPT hole plugs with TorrSeal by Varian to plug the two holes in the cathode bell (for the original filament and getter from the ion tube). The cathode does of course get hotter than the anode. Whether it is the additional heat, (I doubt it) or the fact that at very low pressures below 1 Torr, the discharge/ionization overshoots the cathode and travels down the Brewster stem towards the window... There's a hint in here somewhere. :)
Yep, it's called sputtering and is more of a problem with materials like brass especially when running at low pressure which increase the mean-free-path so that positive ions strike the cathode with higher velocities knocking off metal atoms which go anywhere they want. :)
Well, enough testing with oxygen with this little tube, so my plans for it are to revert it back to an N2 compatible (O2/N2/Air UV laser) which totally amazed me concerning the strength of the UV output. I am currently making up a small dye resonator to test with this Air laser's output.
(A few weeks pass.)
I have now succeeded in getting Coumarin XX to lase superradiantly (a little) but the real problem is that the pump beam waist diameter and divergence from the modified argon ion tube PMG laser is so narrow (1.5 to 2 mm diameter), that even with a quartz cylindrical lens, most of the pumping energy to the dye cell is still centered in the middle of the line. I used a borrowed dye cell, but hopefully I will be able to keep it long enough to add a resonator, at least an HR mirror to it plus a vertical (primary) cylindrical quartz lens at 90 degrees to the second one to widen the beam before it enters the secondary horizontal cylindrical quartz lens and dye cell. Hopefully the losses of adding a second lens will not be counter-productive.
And, now for my future plans:
Today I started (finally) acquiring the parts to built a much longer PMG laser. The head housing is that of a Liconix model 3220 HeCd, (the big one) and measures 37" long which features a nonmetallic 4 rod resonator frame that is just begging to be used for a high voltage laser application. After considering countless laser tube an electrode designs for getting the longest possible active cavity length plus the feasibility of actually producing what was on paper or (the monitor), I finally decided to go with plan "L"...... Plan "A" will be the grand daddy of them all with a 56" in active cavity length and water-cooled cavity and electrodes, but that's for a later date. This current design, (plan L) will feature an active cavity length of 31" plus a large gas reservoir for temporarily sealed operation plus the means for a flowing gas mode, which also helps during the evacuation and outgassing process. Here is a summary of the features of my universal laser design:
The resonator optics can be easily converted to or from a Fabry-Perot to a DFB (Distributed FeedBack) design using a diffraction grating HR.
This sounds like a plan. Of course, you realize that your dram vacuum system is probably way overkill for testing of gas lasers. Perhaps, you are going to go into the tube refurb business??? :)
A few weeks later:
It is oxygen lasing, just not on the transitions mentioned in the Review of Scientific Instruments article. See: U.S. Patent #3,928,820: High Gain Pulsed Ion Laser.
Here is the abstract of the patent:
"A high gain pulsed ion laser is described in which a short discharge pulse of less than about 2.0 microseconds is applied to a discharge tube in which oxygen is present in an amount sufficient to establish an oxygen pressure in the range between about 10 and 100 millitorr. The pulsed output from the ion laser occurs during the "afterglow" of the discharge pulse and is attributed to doubly ionized oxygen. The wavelength of the output may be selected from certain lines in the visible and ultraviolet portions of the electromagnetic spectrum."
Having obtained the patent, here are (some) of the highlights:
This is a good one for reference!
(Several months later)
The big oxygen laser is still shaping up nicely, but not completed yet. Ran into some design considerations concerning whether to place the gas and vacuum valve in the laser head (preferred) or external (not attached to the head) and whether to use standard glass stopcocks or Ace Glass brand P.T.F.E. vacuum valves that have 3 O-rings for seals. I am leaning towards the Ace Glass valves mainly because of the (oxygen lasing gas) and the glass stopcocks expose the cavity/brewster windows and oxygen gas to the necessary vacuum grease. The main reason for desiring the valves to be inside the laser head is so that when it is disconnected from the vacuum system, it is all self contained. The main drawback to mounting the valves on the resonator's base frame is that opening and closing the valves will add strain or twist to it and possibly effect the optical alignment.
Here is a summary:
The "Querflöte" laser I built basically is of the same design as most
other N2 lasers except two major details:
The second feature is more easily to explain. To get an idea, imagine an ordinary N2 laser like a butterfly - both transmission lines are the wings and the butterfly's body is the laser tube. Now the butterfly folds its wings together (sitting on a flower or so): both wings touch each other, and the body is at the edge of both wings.
Back to the transmission lines, they are folded over in a way the common line touches itself. So both transmission lines are the outer sides of a capacitor stack, with the dielectric and the common line (as a single plate) in between them.
I adopted the segmented bore design from the following paper which describes a hydrogen laser:
Hydrogen lasers have very narrow discharge bores even if they are transverse excited. The trick the authors used was to build no true transverse discharge tube but a long series of short longitudinal discharges with alternating pin electrodes in a long thin bore. I liked that idea, for a small volume means high energy density - and by that a better chance to get lasing out of other stuff as nitrogen or air. (Note: I built a number of different ordinary N2 lasers before starting that Querflöte project in 1991. The tip seen occasionally: "Just put neon in your N2 laser and you'll get green lasing" JUST DIDN'T WORK!
The spark-gap sided electrodes of the bore are connected to a common aluminum plate which serves as transmission line and base plate in one. The other transmission line is slotted, so every of the electrodes connected to it has its own transmission line strip. In a way, it's a long series of short Blümleins which have the spark-gap-sided transmission line in common. This design allows a variable length of the laser - just short some strips to ground to change.
A surprising fact was that no charging resistors are needed, as I found out by experiment (The Kirkland et. al. paper have them in their design, as I had initially). I have no idea why it works this way - perhaps there is some residual ionization from the preceding pulse acting as resistor during the next charging phase of a single strip?
The bore itself I modified somewhat - I felt not happy about a pure plastic surface exposed to the violent thunderstorm of a gas discharge. So I decided to make the Plexiglas bore wider, inserting short pieces of Pyrex tubes between the pin electrodes to protect the plastic walls. As few hours of operation showed up, it works: near the electrodes, where the Plexiglas outer tube is uncovered by the Pyrex, white and brown decomposition products became visible.
The spark gap itself is in one corner of the transmission line - however, I observed never asymmetric laser output when running without a rear mirror. Design of the gap is the common type with a screw acting as one electrode - pulse repetition rate is adjusted by turning the screw (to do this without getting zapped, I attached a plastic rod to it by the help of a short piece of vacuum hose).
This is an ordinary surplus 4 kVAC neon sign transformer delivering 35 mA. Nice thing, got it at the back door of a neon sign shop nearly for free. :) It feeds into a voltage doubler/rectifier producing 10 kVDC.
The vacuum and gas supply unit
This is a single stage rotary vacuum pump with exhaust filter (aquarium filter charcoal) connected to one side, low pressure gas can (12 bar) with needle valve and mercury vacuum gauge (0 to 80 Torr) connected to the other side of the tube. The needle valve is used to adjust pressure, the laser is operated in flow through mode.
The tube lases without any mirrors. However, an aluminum mirror very close to the rear tube window increases the brightness of the output beam visibly.
Separation of the mirror to the discharge bore is a simple method of measuring pulse length: Recall that directly after the pulse, nearly all of the molecules/atoms are in the lower energy level of the laser transition. Thus the just stopped discharge is the perfect absorber for the laser light left in the bore. So, the reflected laser pulse from the mirror *must* have traveled down the entire bore *before* the discharge stops to have a visible effect in output. The trick is to measure the distance d from the tube where your mirror just no longer has any visible effect: light:
t = ( L + 2 * d )/ c
L is the length of the tube and c is the speed of light. Then t is an estimate of the pulse length.
Gases used in the Querflöte laser
Summary of the Querflöte laser
R1 HV (10 to 15kV) o--------------/\/\------------+-----------+------o TH-P 100K |P _|_ C1 C2 R2 --- o --- .1uF,20kV Trigger in o---||---/\/\---+------------------- - - - G _|_ (+300 V .1uF 1M | ^ - Pulse) 500V / | | TH1 5563 R3 \ +----+--------+ | Thyratron 500K / | | | | \ R4 / | F F | 1M \ | (5VAC,10A) o 20kV / | -300 VDC \ | Tube+ +-------------+ Tube- _|_ +-------|-| |-|-----+ - +-------------+ _|_ Laser Tube -A thyratron is the vacuum tube equivalent of an SCR - you can turn it on (with a positive pulse to its grid) but plate current must cease for it to switch off.
The 10 to 15 kV high voltage supply charges C1. When a positive pulse is applied to the grid of the thyratron, it turns into a short circuit (sort of) discharging C1 into the laser tube. (Note: I am suspect of the grid network part values. As shown, the input pulse will be attenuated to 1/3 of its peak value (down to 100 V). I do not know if this would succeed in triggering the thyratron.)
The source of the HV is not specified. My choice would be a high frequency inverter or the HV guts of a deceased TV or monitor. A suitable design can be found in the document: Various Schematics and Diagrams. The voltage for the thyratron filament can be easily provided by a modified microwave oven transformer. See the section: Rewinding a Microwave Oven Transformer for use as a Low Voltage Filament Supply. CAUTION: Filament winding must use insulation good for more than 15 kV! The thyratron and laser tube position could be swapped to eliminate the need for the HV insulated filament leads, but then both ends of the laser tube will be at a high potential.
There is one serious problem with the simple exciter: With no control of the shape of the pulse into the laser tube, the current will be more or less a decaying exponential. This means that if the laser's output pulse intensity depends on tube current, it will also not be anywhere close to constant. The cascade system remedies this deficiency. (This appear to apply only to Ar, Kr, Xe - O2 could need the short pulse and high peak current of the simple exciter and thus may not benefit - or work at all - with the cascade system.)
The cascade system consists of two parts. The trigger supply is nearly identical to the simple exciter except for replacing the .1 uF capacitor (C1) with a .001 uF capacitor since very little energy is needed for the starting pulse.
The PFN itself consists of four 2 uF, 5 kV capacitors and three homemade inductors. It is isolated from the trigger supply by a 5 kV diode capable of high peak currents. (The paper specifies a mercury vapor rectifier which was state-of-the-art in 1971). A microwave oven rectifier may be suitable today.
R5 L1 L2 L3 15KV 1 to 3 kV source o---/\/\----+--^^^^--+--^^^^--+--^^^^--+----|>|---o TH-P _|_ C3 _|_ C4 _|_ C5 _|_ C6 --- 2uF --- 2uF --- 2uF --- 2uF _|_ _|_ _|_ _|_ - - - -It may be possible to connect the pulse forming network directly to the laser tube anode (Tube+) rather than to the plate of the thyratron as shown here and in the paper. I believe this should work just as well without requiring the high current pulse to flow through the thyratron (or an alternate trigger device - see below). WARNING: This DOES mean that the laser tube anode will have a lethal voltage on it as long as the PFN supply capacitors are charged rather than being blocked by the trigger device!
The only information on the inductors is that they are 400, 500, and 600 turns of #18 AWG wire on a Plexiglas form (for L1, L2, and L3 respectively). There is no mention of the diameter, or the winding arrangement or packing. :(
For the 1 to 3 kV charging power supply, a simple Variac and small neon sign transformer (or for higher power, a microwave oven transformer - DANGER!).
D1 H o--------+ T1 T2 +--------+--|>|--+---+-----o HV+ )|| ||( | | | Variac )<--------------+ ||( | D2 | | 0-110V )|| )||( +--|--|>|--+ / 5A )|| High Voltage )||( | | \ R1 )|| Transformer )|| +--+ | | / 5M +--+ 3kV,100mA )||( | | | D3 \ 5W | )||( | | +--|<|--+ | 5kV N o-----+-------------------+ ||( | | | | ||( | | D4 | | | +--|--+-----|<|--+---+-----o HV- | | G o---------------------------+-----+ _|_ ////
(Thanks to: Chris Chagaris (firstname.lastname@example.org) for digging up these specs.)
HY-6301 Ceramic-H2 Thyratron:
Multiple SCRS (like 15 or 20!) could theoretically be used but the SCR is also slower than a thyratron and might not be suitable for this reason alone. There are several potential alternatives to a thyratron (and I believe that this level of performance is not needed in any case - see below):
The important thing is to give all those electrons in the PFN a place to go - it doesn't matter how you do it!
However, by using a scheme similar to that of the pulse starters used for some helium-neon and most argon/krypton ion lasers, it should be possible to do away with the high voltage switch entirely. See the sections starting with: Starting Circuits Using Pulse or Flyback Transformers and Igniters, Pulse, Bypass, High Current High Voltage Diodes for HeNe and Ar/Kr lasers, respectively.
Some areas of concern that remain unresolved include high voltage insulation with respect to gas supply lines and water cooling since the electrode assemblies as described in the paper also come in contact with either the gas or vacuum line, as well as the (optional) water jacket.
I keep an old (1987) Laser Focus Laser Buyers Guide around.
It shows that the Florod Corporation actually did make Xenon lasers, I don't know about present day manufacturers. (These are NOT excimer lasers - those are in a different listing for Xenon Chloride and Xenon Fluoride lasers.)
The models listed all operated at wavelengths from 488 to 540 nm with a pulse length of 4.5 us, beam diameter of 3 mm, and divergence of 3 mR. Models LMT and MFA could do 1 to 2 pulses per second; model MEL-30 up to 30 pps.
Krypton lasers were made by Coherent, Inc. (at least through 1987). This one is apparently NOT the same as a conventional krypton ion CW laser but lases on one of the major krypton spectral lines: Their Model I100K3/468-ASE operated at a wavelength of 647.1 nm at a repetition rate of 7.6*107 pps, pulse energy of 8*10-9 J, and pulse length of less than 120 picoseconds (!!). However, note that this is NOT a pulsed tube but is mode-locked and cavity dumped resulting in a pulsed output. The beam diameter was 2 mm (TEM00) with a divergence of .5 mR.
P.S. The laser I really like is the nitrous oxide laser which was made by Lumonics (now GSI Group), model TEA-200-2. "The laser that makes you HAPPY", The more you use it (vented to the room ambient) the happier you get. :)
Just want to mention some characteristics of the Oxygen Laser.
Oxygen will only lase in a pulsed mode. Back in 1971, some Canadian researchers theorized that you could use ordinary air for the O2, but other researchers back then found that other gases, including argon, only degraded the output power. Plus, air has a lot of moisture in it and other contaminates like smoke, smog, dust, and whatever else is floating around. You would need one heck of a special filter and drying system in order to keep from polluting the laser cavity and fogging up the inside of the Brewster windows or optics. And to make things worse, oxygen - especially hot ionized O2 - is very reactive with metal electrodes and even with quartz glass tubing. Welders' grade oxygen would be the next purity grade up from just using plain vanilla favor air. I am not sure how much moister and other contaminants would be in medical-grade O2. Lab-grade O2 would of course be the ultimate choice but it's much more expensive than just air. :)
Also, you would need to be very careful about what type of vacuum pump oil you use especially if you are using a high RPM pump since: Oil + O2 + the compression stroke in the vacuum pump = possible BOOM. Oxygen itself is not explosive, but it sure makes anything that is flammable burn extremely well. I am not trying to discourage anyone from attempting to build an oxygen laser, just listing some of the basic concerns and cautions. It's definitely ~ a challenge. :)
(From: James Sweet (email@example.com).)
That's definitely a very good point, I've seen heavy machined brass oxy welding regulators blown apart from small amounts of oil contamination in the high pressure side.
(From: Steve Roberts.)
Well I don't know how it is in your state but around here, Medical Grade O2 is four nines, i.e., 99.9999% purity and must be checked for CO2, CO, etc. The goal is to build a low cost unit, and if you have to replace windows or cold cathodes made from aluminum stock, that's cheap for the power you can get from the unit.
The Canadian design for the oxygen laser is designed to have throw away, easily replaceable electrodes and Brewsters, because its a pumped, not a sealed laser, you can take some liberties on the tube design such as brewsters held on with O-rings. It's not like a argon ion laser where everything has to be ultra-high vacuum and high temperature compatible. Electrodes, optics, etc., are demountable, so you can clean or replace them. The warning about O2 and pump oil is a good point, but you're bleeding so little O2 through a leak valve that I wouldn't be too worried. You also have a leak valve on the pump side, otherwise the O2 regulator will just slam open to about 25 psi and bleed all your O2 right into the pump. This is true with any flowing gas system, both vacuum and source have to be throttled, otherwise you loose all your gas in a few minutes. Usually even a dying pump is good enough to suck the tube out fast.
You brought up some good points about the method for gas regulation. :) I found that a very small diameter capillary tube about 1" long inserted into the 02 supply line and a needle valve downstream also works very well to prevent the low pressure regulator from slamming open. I am using a 5 psi secondary regulator too. With this set up, I eliminate the possibility of over pressuring the laser tube and gas supply lines, oooops. :( Plus, this way I can pull full vacuum during the initial pump down with my Welch 1402 pump (slow RPM). I admit that there are advantages to both gas flow regulating methods. I would go with Steve's suggestion of using a metering valve on the roughing line (pump side) if one is using a high RPM oil filled vacuum pump or drying pump - (LN2 and molecular sieve).
The concerns about the potentially explosive nature of mixtures containing high pressure oxygen are certainly valid. However, given that these lasers run at a Torr or less, as noted, the only danger would be due to carelessness or a major leak or regulator/metering valve failure resulting in atmospheric pressure O2 entering the vacuum system.
175 cm length, 10 mm discharge tube with a near hemispherical resonator of 195 cm length. Optics: HR broad band mirror at 99.9% from 400 to 670 nm, various OCs of 3 to 10% transmission. There are 21 available lines. They documented 6 in detail. This is with a 5 us, 10 kV pulse and a current density of 6*103 A/cm2. Pulse repetition rate was 10 Hz. Optimum pressures varied from 5 to 20 mTorr.
Lases at: 455.8 nm, 515.7 nm, 534.5 nm, 549.9 nm, 559.2 nm, 669.9 nm.
Peak Power Line (when optimized) ---------------------------- 526.0 nm 110 W 535.2 nm 60 W 539.4 nm 120 W 595.5 nm 100 WThe 455.9 nm line fights with 539.4 nm.
Abstract: A pulsed xenon ion laser with a output power of 5 kW at 364.5 nm designed as a pump source for several blue dyes with conversion efficiencies in excess of 20% with dye laser output pulses if 120 ns.
Lases at 231.5 and 364.5 nm
Construction: 190 cm discharge volume of 8 mm diameter Pyrex tubing with quartz Brewster windows. Excited by a .3 uF capacitor charged to 12 kV with spark gap triggering. Pulse repetition rate 10 Hz, rate limited only by power supply. Used dielectric coated mirrors with a 99.99% HR and a 40% transmissive OC.
Output power at 364.5 nm was 5 kW with a pulse width of 200 ns FWHM. Xe pressure of several microns. They also describes a matching dye laser. Refs: Vol QE-11, pp. 935-937 for the green version of the same unit.
Test conditions: 122 cm discharge tube with 6 mm inner diameter. Mirror separation 183 cm. HR with 10 meter radius and reflectivity of approximately 97.5% at 664.0 and 99% at 559.2 nm. OC is flat with reflectivity of 98.5% at 664.0 and 97% at 559.2 nm. 65 us pulses at 1 pulse per second. Laser operation was observed over a range of 16 to 35 mTorr in oxygen
At 2,300 V across the tube, saturated power of 2.3 W at 30 mTorr on a 350 A peak pulse. Threshold of the O II 664.0 nm line was around 150 amps. Threshold for the O III 559.2 line was 320 A with an output power of 4 Watts. They also had a nitrogen line at 648.2 nm.
It used a 3.5 mm diameter tube, 123 cm in length, with a hot cathode made of tungsten. It had a anode end side arm filled with molecular sieve material to hold iodine, and a cathode end sieve without iodine. The tube was in an oven. Controlling the temperature of the molecular sieves controlled the iodine pressure, optimal sieve temperature at the anode end was 120 to 130' °C. Cataphoresis causes the iodine to be pumped to the cathode, using a hot cathode prevents the iodine from reacting with the cathode materials.
The cavity used was a confocal cavity with two high reflectors coated for 100% reflection from 450 to 700 nm. A Brewster plate inserted into the cavity extracted about 10 mW of output. Optimal conditions for a current level of 400 mA were 4.7 Torr of helium and a sidearm temperature of 130 °C for the source sieve.
The material used for the sieve is not stated, although one can assume most molecular sieves made for sorption pumps will work.
Lases at: 540.7 nm, 567.8 nm, 576.1 nm, 612.7 nm (highest gain), and 658.5 nm.
The authors state that with proper end mirrors, this laser could be scaled to 30 to 40 mW of power.
Reference: R. A. McFarlane, "Laser Oscillation on Visible and Ultraviolet Transitions of Singly and Muliply Ionized Oxygen, Carbon, and Nitrogen", Applied Physics Letters 1, vo. 5 no. 5, Sept 1964.
Wavelength (Angstroms) Element/State ---------------------------------------- 3749.47 O II 3754.68 O III 3759.892 O III 4347.38 O II 4351.23 O II 4414.93 O II 4416.97 O II 4605.52 O Unknown 5592.48 O III 6721.38 O II 4647.25 C III 4650.11 C III 3478.76 N IV 4097.29 N III 4103.26 N III 4510.45 N III 4514.41 N III 4630.31 N II
(O III lines are in the afterglow, O II lines coincident with the pulse.)
The nitrogen lines are from NH3 (ammonia) and N2 in Air. The carbon Lines are from CO2 (carbon dioxide). The O2 lines in oxygen. The discharge is in a 7 mm bore, 1 meter long, pressures 20 to 50 milli-Torr with pulse currents in excess of 500 A from the 2 uF capacitor with a pulse width of the order of 20 us.
Both mirrors were high reflectors, radius unknown. The main set of mirrors peaked for 430.0 nm. Others were visible broadband.
Just so you know where I'm coming from, I have built several N2 lasers at home for fun, and one of them would lase open air at 1 atm giving at best, about half the pulse energy, as perceived by eye, of the output when LN boil-off was used.
In the section: Home-Built PMG Laser Description, Steve Roberts mentions that:
"there is also a line at around 545 nm (or should this be the green HeNe line at 543.5 nm?) which will lase brightly in a transverse discharge similar to that of the nitrogen laser and also lase in a long tube but not as strong."
I don't know about that wavelength, but I do understand, and you even admit it in the chapter on the nitrogen laser, that neon lases at 540.1 nm in a fast transverse discharge. This might possibly be the line Steve is talking about. According to my refs, this line lases with about 1/10 the power of the 337 nm nitrogen line in the same laser.
Additionally, I have built a PMG laser for a research project when in college, years ago. It was supposed to lase the Xe 3+ lines that you talk about - but when there was an air leak, which was quite common before the plumbing was fixed, you could see the discharge change color, and a strong line at 409 nm - beautiful purple, would come up. It was about as strong as the Xe lines. It was due to nitrogen, atomic neutrals I think.
This laser was Pyrex, had a 3 to 4 mm bore, approximately 60 cm long, was pumped to several tens of microns pressure, and the discharge was a several 100s of pF, charged to several kV, with a repetition rate of between 30 and 100 Hz. It had an indium cold cathode and a large foil-ring anode, and blue-green mirrors, 99.9% and about 98% R. (Sorry, this laser was nearly 20 years in the past. I don't recall the exact details.)
It would also lase the Ar+ blue-green lines, at lower voltages and higher tube pressures.
Back to the 409 nm line, of course if it were an air leak, it was lasing 79% N2, 20% O2, etc. The output might have been more with pure nitrogen, but I never tried it because I was supposed to be working on the xenon laser. However, there was a bottle of argon and I did try that. BTW this was in Bill Bridges's lab at Caltech, where I went. (He invented the argon ion laser and he probably suggested that I try Ar just for fun.)
This laser had silica Brewster windows and also lased several UV lines 200 to 400 nm with Xe filling at 50-70 microns pressure.
I have been an electronics and laser hobbyist for years, a ham (N6URH) and so forth. I have been planning for years to build a monster TE laser, primarily meant for 337 nm nitrogen, and it turns out there there are dozens if not hundreds of lines that such a beast can lase. Not only Ne at 540.1 nm, but also hydrogen at 160 nm in the VUV, nitrogen molecular ion at 428 nm, some other vis. lines in Ne, and lots and lots of others. Xe is supposed to give a very energetic ~3.3 (3.9?) micron output in such a device, and you can lase nitrous oxide at 11 microns, and so forth. There are also the excimer combinations, XeCl, etc, and some unusual ones (XeO) that will also go. I saved a bunch of papers on all these lines, and a summary of them may be of interest to you. But of all of them, 337 nm nitrogen is nice because the gas is benign and the wavelength satisfactorily useful. I have also pumped home-brew dye lasers with my home-brew nitrogen lasers, including one that had a diffraction grating mounted on a protractor as a cavity mirror. You could literally rotate the grating around, and change the color, and then read off the angle of the grating from the protractor, and calculate the implied operating wavelength.
Fun fun fun. :)
(From: Mike Poulton (firstname.lastname@example.org).)
I have no idea what the proper xenon pressure is, but I can tell you that it probably can't be mixed with any other gases. Since helium has a much higher ionization energy than xenon, it might be okay, but don't count on it. It's probably better to go with straight xenon. A neon sign shop can probably do the filling work for you. You can buy 1 liter glass vessels of xenon at 1 atm from some neon sign suppliers -- check around on the net. I think they are about a hundred bucks.
Look for a local sign shop with only one location. Shops with multiple locations do all their pumping at one, and are generally too busy to be bothered, even for money. Small shops will often do this sort of thing, although it WILL cost you. They will have to attach your xenon flask to the vacuum manifold. If they have an extra spot on the manifold, which many places do, then it's no big deal. If not, then it requires removing one of the flasks already attached (neon or argon), which wastes the rest of it ($14), and keeps them from using that gas as long as your xenon flask is attached. Your xenon flask cannot be removed without wasting the rest of it. They will probably have no use for xenon, and will want it off the manifold as soon as possible. That means you get one try -- you can't go back the next week and do it again without buying another liter of xenon.
While the tube is under high vacuum (1 micron or less), it should be heated to 250 C or more to bake out impurities. The neon guy will want to bombard it -- that would probably destroy the electrodes. Instead, use a hand torch very carefully and slowly. The tube probably ought to be partially refilled (to a few mm Hg) with xenon and then re-evacuated while hot to remove all impurities possible.
Now here's the cool part: Slowly fill the tube with xenon with the power supply attached an operating, and the mirrors fairly well aligned (using a laser). It sounds like a high-gain medium (pulsed high power emission) so getting some beam shouldn't be too hard. You will probably want the capacitor to be at 4 kV or more, and have the trigger firing a couple times a second. Mess with the pressure until you get the best discharge or (if you're real lucky) the best beam.
If you want to get tricky, try refilling the gas ballasts with xenon at a higher pressure than the tube. Open the ballast valves, fill to fairly high pressure, close the valves, and pump down to optimum pressure. Note that you may have one gas ballast and one vacuum ballast to allow for pressure adjustments either way, in which case the vacuum ballast needs to be filled (emptied?) and sealed before filling the tube. Play around. You have no incentive to save xenon gas, since that remaining in the flask will be wasted when it is removed from the manifold.
Have I ever done this before? Of course not. I've never even seen a xenon laser. However, I do know a good bit about gas discharge tubes and processing, and the neon industry. This is an ad-hoc process omitting many important steps and details necessary for optimum tube life. It's experimental, untried, and your mileage may vary. Good luck.