The key here is that neither output of a bridge rectifier (which is most likely at the front end of the power supply) is at Earth Ground potential. They both have a large AC component with respect to Earth Ground. Consider:
D1
H o-----+----|>|-------+---------+-----o DC+
~| D2 |+ |
In from +----|<|----+ | +_|_
AC line D3 | | C ___
+----|>|----|--+ - |
| D4 | |
---+-o N o-----+----|<|----+------------+-----o DC-
| ~ Bridge -
+-o G o------------------------o Earth Ground (also connected to scope)
_|_
///
Hot (H) and Neutral (N) are tied together at the electrical service panel. Now think about what would happen if the scope test probe ground lead was connected to DC- without an isolation transformer. This would basically short out D4 and put D2 directly across the line (among other things). Not good. With an isolation transformer for the power supply, there will be no fireworks. However, an isolation transformer for the scope will not help unless it has an isolated ground, or the scope ground is disconnected.
Take extreme care when making measurements. The laser itself should be powered via an isolation transformer if possible. This is much preferred to isolating the test equipment (or - gasp - cutting their grounding pins or using 3 to 2 pin grounding adapters without attaching the third wire). The reason is that isolating the power supply prevents the situation where accidental contact between the internal circuitry of the power supply or laser head, and earth ground, will be a shocking experience. If only the test equipment is isolated, the power supply and laser are still live with respect to earth ground. This will permit measurements to be made, but is still extremely dangerous.
For tests of the control circuits or even the igniter, a modest size isolation transformer will be adequate. These are readily available new or surplus, or can be constructed inexpensively. It's only when the tube starts that the high current capability will be required. See the document: Troubleshooting and Repair of Consumer Electronic Equipment for more info on isolation transformers.
Note that most isolation transformers DO NOT isolate the earth ground (third prong). Thus, if using one on an oscilloscope where signal ground and earth ground are tied together, there will still be fireworks, and the scope cabinet will still be a potential shock hazard (for the return from the live laser).
Ultimately, safe troubleshooting of line-powered and/or high voltage systems comes down to good work habits - one hand in a pocket, elimination of Earth Ground points to touch, not working in hip-deep salt water, etc.
Also see the document: Safety Guidelines for High Voltage and/or Line Powered Equipment.
For example, the following table shows the expected MTBF hours for an ALC-60X/Omni-532 compatible tube (specific model unknown) as a function of beam current:
Plasma -------- Laser Output Power (mW) --------
Tube Multi- ------ Gaussian TEM00 Mode ------
Current mode ------- Pure Line ------
(Amps) -- All Lines -- 457 nm 488 nm 514 nm Lifetime (MTBF) Hours
---------------------------------------------------------------------------
4 20 10 1.0 7.0 0.0 15,000 - 25,000
6 50 30 2.0 17.6 7.5 8,000 - 15,000
8 110 70 5.0 27.0 23.0 4,000 - 6,000
10 220 130 10.0 44.0 42.0 1,500 - 2,000
12 325 200 15.0 60.0 68.0 1,000 - 1,500
14 430 280 22.0 81.0 98.0 500 - 1,000
Thus, it makes a lot of sense to run at the lowest power that will be adequate
for the application whenever possible to maximize tube life (not to mention
keep your utility bills within reason!). For a multiline tube, reducing
current may still get you the particular color(s) you need with adequate
intensity and/or acceptable color balance. However, each model/age tube will
have a different balance of lines at a given current (i.e., of the 10 for
argon and 11 for krypton). Also see the section:
Laser System Life.
Note: The power output numbers, above, look like they apply to a very 'hot' or new tube - at least compared to what you'll typically find in a surplus 60X laser. The actual power (at a given current) for one of these will likely be somewhat lower. Keep this in mind if you see an advertizement for a 400 mW 60X! Also see the section: Expected Output Power from Surplus or Previously Owned Ion Lasers.
(From: Steve Roberts (osteven@akrobiz.com).)
10 A is absolute max on a average 60X tube. I once talked to a engineer who was on the 60X project, and he prefers 9 A as the upper limit. Keep in mind that's a MTBF for a new tube, not a used one. Also note that the factory sells lasers from 5 mW and up, therefore you don't know what you have in the cavity.
The American PSU can in theory source 14 amps for short periods of time. However this is only on a modified unit. 10 Amps is about what they can do. 11 amps runs the pass-bank way hotter then its 100% duty cycle temp without readjustment, in my opinion.
However I've never seen a manual for it and am going from what I've learned from trial and error. A few newer compact FET based gold boxes made by Marlin can only do 9 amps. Above that they go into meltdown, so one has to be careful.
As for maximizing life of a small to mid-size tubes (up to 2 or 3 W) where the laser is to be used intermittently (between shows, for example - say 10 minutes on, 20 minutes off), it is better to drop it back to idle (the lowest current which will maintain the discharge) when the beam is not needed rather than shutting down completely. (However, large frame ion lasers should be turned off entirely between shows - they like to run at one particular current.)
(From: Steve Roberts (osteven@akrobiz.com).)
For larger water-cooled ion lasers reducing current increases tube life only up to a point. However, running consistently at low idle is not good for the tube. On short frame lasers there is a certain reference current. You can go about 6 to 7 amps below it and an amp or two above it and still have decent lifetime. This point is determined from the factory data sheet for a given tube and by users' experience.
Running above this window leads to quickly shortened tube life. Running much below it can result in plasma instability and PSU damage, as well as eventual tube damage. The rebuilder I used to work with stressed running consistently for long periods at the reference current as the best way to avoid warranty repairs, and I'd tend to believe him.
Also, depending on your PSU design, it is often recommended that you start at the high end of the current range. Otherwise, damage to the PSU and tube cathode may result from repetitive start pulses when the tube doesn't catch but just flashes. Some power supplies will start at a high current regardless of the control settings, and then drop back to the selected current or power.
On white light lasers, or pure krypton, there usually is a 'knee' curve in the plot of light output versus current where the krypton lines really take off. I'd find the knee point for my tube and see where it is versus the factory recommended current and run right around that region or a little above it. What I mean is if you look at the Kr gain curve, you'll see its a very gentle slope from laser start to usually a point around the mid current range, where the line in the plot starts to shoot nearly straight up. If the tube is set up right, that breakpoint should be way below the recommended operating current. The knee point will shift higher as the laser ages, and Kr adsorption rates are much higher then Ar rates. 647/568 nm balance is a good indicator of tube pressure, lower pressures favor yellow, however red has much more gain then yellow, so you'll always see some red.
Owning a good power meter and logging line power on a weekly basis with a prism is a good idea. One of the problems of owning a high power argon or mixed gas is that laser ownership at a point stops becoming science and turns into somewhat of a art.
OK, here are the gory details:
(From: Steve Roberts (osteven@akrobiz.com).)
Lifetime decreases exponentially with current, and you can get to the high speed end easily enough. (See the chart in the section: Argon/Krypton Ion Laser Tube Life.) At currents not much over 10 amps, assuming you have a power supply with enough spare output compliance to not burn up, first thing that happens is it outgases from the cathode and bore, so that's the LAST time it lights. Then she starts sputtering and flashing, at this point some will just wink out from overpressure. If the pressure was low enough to begin with, now the anode starts glowing red very bright, and if it doesn't crack off, and if the cathode feed-throughs hold, and the bore doesn't crack from shock, you'd see maybe a watt or so for a few seconds until the cathode goes dead from meltdown or the return bores could no longer handle the flow.
If you did reach this region, getting the cathode material off the stems and replacing the etched insides of the Brewster windows is hard. You can see the particulates accelerate inside the tube from the intracavity light and they go right to the windows at HIGH speeds, glowing like little fireflys and acting like fast little cruise missiles and stick when they hit the window. If they bounce, they go into the returns where the charge is more neutral.
By this time the gas returns will be filled with crud, and when it cools off and you try to relight it will probably find the returns mucho more interesting then the bore.
Interestly enough, new tubes are the hottest until they season in. It's the Brewster windows, cathode, and pressure that determine tube life.
Hint: Got Brewster windows that are clean both inside and outside with good tube voltage, a positive delta-T, and good magnet?, and yet no matter what you do you get no power, put in two high reflectors for a while, ramp up the current and get rid of your color centers in the windows. This is a last resort sort of thing for bigger lasers.
The former Soviet Union and a few U.S. researchers developed some neat tricks to avoid these things and get to 500 to 1,000 watts, but I m not talking. :-) The tricks are not all that easy to do outside of a lab anyways and they fry the coatings off the optics fast.
HGM Medical, the surgical arm of American Laser, made some glycol cooled things roughly the size of a 60X. A water-to-air heat exchanger is built in. Bore sizes are similar, but the changes are in the gas returns, anodes, stems, tube pressure, PSU etc. They do 3.2 watts new at the OC, for a few seconds at a time at a VERY VERY low duty cycle. The limits are the anode and pressure, the anode is air cooled. If you hold the treat pedal down for more then 20 seconds, the power drops off like a falling stone as the pressure rises until the overtemp switch cuts out to protect the tube. Then you need to set for about 5 minutes and it comes back. These were for eye surgery, and the eye surgeon rarely needs the full 3 watts, 200 to 500 mW delivered power to the eye is more typical. Also the OC transmission and lasing threshold is a lot different on a power on demand tube. You may not get enough transmission on a normal X optic. They usually have PC in the part number and a diode aiming beam, since the tube doesn't lase until needed, call it fire on demand, once the tube gets stable, then the shutter opens.
Factory manuals and practical experience are wonderful. ;-)
Here are some more of the gory details, if you can stand them!
The following take place in NO particular order: They are heat related and plasma/pressure related fast failures!!!
The maximum limits are 700 A/cm2 for graphite and 850 A/cm2 for BeO. It's somewhat higher if there are no gas returns being a tradeoff between mechanical strength and radial thermal conductivity to cooling water. However, the tube will suffer from cataphoresis (gas migration) if it there are no returns. Gas returns internal to the BeO weaken it and raise thermal resistance. (This information from: Kesik and Siejca, SPIE vol. 859, Laser Techonolgy II.)
There's another 10 or so nasties, but I'm outta time, chao!
Buy a bigger tube then you need, turn it down to like less then 1/3 to 1/2 of its design power, cool it well, it will live nearly forever.
Filament Discharge
Manufacturer/Model VRMS VDC at IDC Typical Output Power
------------------------------------------------------------------------------
American Laser 60X 3.0 107-108 at 10 A 20-40 mW, higher multiline
Cyonics/Uniphase 2301 3.0 105 at 8 A 5-20 mW 488 nm
NEC GLG-3030 2.6 85 at 10 A 20 mW 488 nm
Spectra-Physics 091-93 2.9 80 at 10 A 25 mW 488 nm
Spectra-Physics 093-91 2.9 110 at 10 A Greater than 25 mW, 488 nm
Aside from obvious physical damage, there are a variety of conditions that can prevent a tube from lasing or result in a weak beam or one with rings or other artifacts:
(From: Anonymous.)
"Where a tube outputs lower than normal power regardless of optics, alignment, and tube position - assuming the PSU is the correct one and functioning properly - could be low on gas. To find out, take a pair of argon (red) safety glasses (EXTREMELY IMPORTANT!) and power it up. Then look for a red spot on the inside of the anode-end Brewster window. The red fluorescent spot will be the indication of low gas pressure."
To check for cathode sag, with ONLY the filament powered (make sure the high voltage is totally disconnected from the tube - you wouldn't want it to discharge while staring down the bore - or to get your nose zapped!), look through the mirror at the cathode end of the tube to view the interior:
However, Ar/Kr ion tubes will often have some sort of additional optics at one end, either a line selecting prism or ghost suppression wedge. This will make it impossible to easily adjust alignment at that end at least.
WARNING: Both ends (and possible other parts) of the tube are probably line-connected so these procedures are significantly more dangerous than with a typical HeNe laser. Take care. It may be possible to use the approach outlined in the section: Pulsed Operation of an Ar/Kr Ion Tube for checking alignment at least. However, the mirrors will have been adjusted at the factory when the tube was running at normal operating temperature (hot!) so it won't behave quite the same (this is probably particularly true for longer tubes).
One cause of the buildup of a damaging coating on the inside surfaces of mirrors or Brewster windows (usually at the cathode-end) is sputtering. If run with an adequately heated filament and moderate (well within specs) tube current, there will be minimal sputtering. However, with excessive tube current and/or an underpowered filament, sputtering will increase. With some power supplies having a wide compliance range, it is even conceivable that the tube might appear to work (for a while at least) with a filament that has lost all power. (I don't know how likely this is with these high current ion lasers as most power supplies will just drop out if the tube voltage were to climb substantially, as it would under these conditions. However, lower current helium-cadmium lasers have been known to do this with an unpowered or open filament.) The arc could either continue to the filament but with much greater 'cathode fall' (the voltage drop between main part of the discharge and the filament) or even strike to some other part of the tube. Assuming the structure doesn't fail outright, one result will be a large amount of sputtering and rapid overcoating of the cathode-end mirror or Brewster window, rendering the tube useless very quickly.
(From: Anonymous.)
"If an air cooled argon produces rings of actual coherent laser light around the beam that cannot be eliminated by alignment and tube position and optics cleaning then it probably has a lot of hours on it. The source of these artifacts will be a metallic beryllium ring that is formed inside the bore right at its cathode opening."
Hydrogen and helium slip through glass and ceramic, and can stay hidden in certain metals used in tubes. They are also used in processing the porous cathode for various reasons. The heavy argon and krypton ions drive the lightweight hydrogen deep into the bore walls and metal portions of the tube, when a tube sets on the shelf, hydrogen can in some cases seep in from the walls and welds. This happens even in quartz or hard Pyrex, and more slowly into ceramics. Most of the hydrogen originates from welding of the seals and fusing of the ceramic to metal joints in a hydrogen furnace during processing. In glass tubes it can come from the cathode and metal structures.
Hydrogen doesn't ionize easily. Thus it raises the tube pressure while increasing the resistance of the discharge. This can lead to a condition where the tube will not lase. Argon has a wide lasing pressure range but as the pressure climbs too high, the blue and violet lines trail off, the 514 nm and 528 nm (green) lines get a little brighter, then the green falls off and lasing ceases, even though in some cases the discharge will still start. Any gas in the tube other then argon or krypton diverts energy from the lasing process - some more then others - and can cause cathode poisoning.
Tubes are carefully filled and designed to center the positive column and the negative dark space and negative glow in specific areas of the tube. That's one reason why cathodes are powered by center tapped AC - to move the arc around. When this alignment deteriorates, you get bore wall erosion, localized heating, and other undesirable effects leading to tube damage.
Thus, a tube with large amounts of hydrogen that the getters can't handle needs reprocessing soon as the hydrogen will change the plasma physics in the tube by moving the cathode dark space and other regions of the discharge which promote erosion to places in the tube not designed for erosion. The main damage will be to the cathode, causing a hot spot to form ultimately leading to cathode sag or total failure (filament breakage). In fact, it is this small hot spot, not gravity, which is the most likely cause of cathode sag. The hot spot may also cause the electron emission agents mixed into the cathode matrix to boil off and dump more impurities into the tube. Low tube pressure will also move the discharge back toward the cathode, heating it excessively.
Other impurities have similar damaging effects, including some that are much more reactive with the tube materials. Carbon dust, for example, will be ionized and accelerated toward the windows, damaging them.
You normally leave the end-bells floating. The only time you would connect to a end-bell is when you want to heat the getter. I don't know where the getter is on Cyonics tubes. Often they have a 3rd pin for it, if not, it's in the cathode end-bell. The getter is a corrigated or fan fold ribbon with little tabs sticking out of it located toward the front of the end-bell so it's not directly bombarded by plasma. It is an active device using the floating ions and electrons for activation. Current is passed through it at the factory to convert carbonates to pure metals and to outgas it.
It is not safe to fire the getter on a 60X, I (Steve) have tried it a few imes and a hard start tube becomes a no start tube real fast! For some of their older tubes, Coherent has a getter procedure for a worse case situation where hydrogen has outgassed when the tube sets on a shelf and is thus at high pressure. Otherwise, no manufacturer would ever suggest lighting off the getter. It takes 7 to 10 A to heat the X getter, a ribbon secured to the cathode at one end. So, it really gets heated a bit during normal use. Further heating would just evolve more gas or burn off more material causing it to bury argon and/or liberate whatever it has gettered. The 60X getter is unique in that its coating is on both the end bell wall and the ribbon after its heated.
On other tubes, getters are designed differently. On Lexels, for example, there is a getter in the ballast tank and the whole inside of the tank is coated. A electrode sticks out of the tank to activate the getter at the factory.
The getter ts there mainly to collect hydrogen and water vapor. Having loose H2 (which diffuses out from the welds and brazes which are done in a hydrogen furnace) raises tube voltage and lowers power. That H2 has a tendency to make water vapor which is really hard on the cathode.
(From: Steve Roberts (osteven@akrobiz.com).)
If your tube is getting old and the gas pressure is starting to drop, refill the tube before you start damaging the cathode. Consult a professional to find out what your tube voltage drop should be for a given current. A lower voltage across the tube equals lower gas pressure. Don't wait too long or it won't be possible to rebuild it.
The following comments were stimulated by being told by another laser tech why my customer's laser that had excess argon added by mistake should not be repumped. The tube isn't at quite high enough pressure to do major damage and the excess argon will burn off eventually. Hydrogen, on the other hand, tends to reappear for some reason, often without warning. I think the answer to that would only appear if we were to torture a ticklish Spectra-Physics or Coherent engineer for hours with a feather :-).
You see, I had thought all along that argon was what needed burying by use, and some of it does need to be buried, but what argon that actually outgases gets reburied in the first 15 to 20 minutes, unless a fill valve from a reservoir is leaking or is opened too much by mistake on a big argon. It's not the argon pressure that prevents ignition, but the H2! Hydrogen seeps into the metal during manufacture of the tube, and can in some cases either leak out slowly over time or suddenly appear, most likely from thermal shock.
Hydrogen causes a large increase in tube voltage, makes starting harder and depletes the population inversion, thus reducing power. When hydrogen or nitrogen is in the plasma, it generates large amounts of heat, thus damaging the tube, hydrogen also reacts with the cathode materials, reducing output and damaging the cathode. If a tube goes high in hydrogen and the getters wont clean it up over time, it's time for a rapid repump, to recover the tube from quick failure.
I learned this the hard way today on a customer's 10 watt laser. The previous tech, not knowing he had a PSU problem, assumed it was low tube voltage (discharge), when in fact the tube voltage meter on the front panel was 25 volts out of calibration from a fried resistor. Each fill brought the laser tube up about 3 to 5 >volts on this model, and he did 5 or 6 fills. Thus, the nominal 354 V tube >drop was now actually 374 V and the PSU could still ignite it. But, the blue >and weaker green lines were suppressed by the high pressure. However, the >528 nm line really blossoms when the pressure is high. This is why a new tube >is so hot in the green but then goes blue as it is used and pressure goes down. >The overfilling resulted in a 2.5 watt drop in lasing power.
The net result was at low currents the tube would not lase at all, and it pushed the lasing threshold (current) through the roof. And, getting the unit to its rated power required a current of 34 A instead of 31 A! We were lucky enough to have a factory test sheet on this one with all the detailed measurements. A high pressure tube manifests itself in two ways, it won't lase or lases at greatly reduced power, and the voltage across the tube increases. I have spent many hours on some units thinking I had a alignment problem, when in fact it was real high pressure.
Once I fixed the ignite boost circuit - which was the original problem - the tube did start up but getting it aligned well enough to get any power was a factor of 10 harder then what it would have been at normal pressure.
The gas fill on this unit is not by solenoid, but by two manual valves with a small space between them, you open one, fill the space, close it then open the other to add the small known quantity of gas. The previous tech thought it was necessary to close the fragile refill valves with vice grips, thus ruining the needle valves, he didn't realize the tube takes time to stabilize as the fresh gas has to be pumped to one end of the tube, heated, and then the new pressure would stabilize. This actually takes 15 to 30 minutes per fill.
Some 500 hours of operation from now, it will be where it should be. All this was heard when I hauled the unit to the refurb shop for a more thorough evaluation. However if this were hydrogen outgassing, which raises the firing voltage and damages the cathode, a repump would be required. A repump would have also been required if the pressure was so high that lasing would not be possible, in this case about another 10 volts. When in doubt, call in a expert!
Tube voltage drop is directly related to pressure in a Ion laser, charting the tube voltage is a easy way to monitor your laser, gradual increases in pressure over time suggest a leaky fill valve. Sudden increases in tube pressure suggest a tube that has possibly overheated and been damaged, hydrogen out gassing, or a recently activated autofill system. An autofill system getting out of control could rapidly overfill a tube. If you hear a sudden burst of solenoid valve clicks, something is going on. Most modern fill systems have a double solenoid valve with a small chamber between them, and a larger high pressure argon reservoir. Opening valve number one fills the chamber with a known amount of gas, then valve number one closes and valve number two opens the chamber to the tube. It then takes 5 to 20 minutes before you see a real stable change in tube voltage. Although you will see a sudden jump when the valve opens, it takes awhile for the gas to go against the discharge and flow down the tube. It is wise to let a laser run for 20 to 30 minutes before enabling fill, as well as checking across the tube with a multimeter once in a while to check the panel meter calibration.
Keeping an eye on where the autofill enable switch is set is a good idea. On some older lasers it's a third key-switch position, and the operator is responsible for just pulsing the key momentarily to open the fill system solenoid - until the fill system buzzer stops beeping or the light goes out. That little "fill" button is a tempting thing for a visitor to push on some lasers, especially if the fill card has been rewired or bypassed by a tech tired of having his morning coffee ruined by a loud beeping while his recalcitrant laser stabilized. If you do anything to the fill circuit, carefully recalibrate it to the factory specs or add a extra key-switch to prevent it from firing.
Some older mixed gas and krypton lasers have a bizarre version of the fill system called a "pressure pump". Since some of the the Kr red, yellow and green lines are fussy about lasing pressure, the pressure pump has a third larger chamber besides the fill chamber to pump the tube down. You increase the current, which raises the plasma temp and thus the gas pressure thanks to the laws of Boyle and Charles (remember P = VT?). A series of valves then open to the larger third chamber, which fills with hot gas acting to reduce the bore pressure. Later when you wish to raise the pressure, the third chamber is slowly dumped into the tube. These lasers tend to have a series of special markings on the tube voltage meter, to guide the operator in choosing the best pressure for a given line. The 647 nm red and 568 nm yellow lines share the same upper state. Whether you get red, or yellow, or a mix of both is determined by the gas pressure which favors a certain lower state for the transition. The 647 nm line likes high pressure, while the 568 nm line likes a much lower pressure. Running one of these requires some operator skill, patience, and a good reading of the manual.
Remember, when working with fill systems, patience and care is needed, and time must be given for the tube to stabilize.
Incidentally, if you do much work on lasers, it is wise to buy a cheap DMM, as when working on the laser you can often forget it's still across the tube, then when the tube winks out and restarts, the ignite pulse fries the meter into oblivion. (Or always put the meter before the igniter if you have access to the power supply and account for any wiring/other voltage drops.)
And for those poor souls who would wish to repump a CR52 or other ancient coherent graphite with the vacuum access valves mounted on the reservoir, the valve actuators to hook onto the tube are available from CPC-Cryolab. Now I can't reccommend equipping modern ion laser tubes with O-ring sealed anything, having bad experience with various valves etc. Orings cannot be depended on at the low levels of vacuum used in a ion laser. Graphite tubes have such horrible gas cleanup and release cycles that something was needed to recover the tube if it got over or under pressured. On more modern tubes, a oring or most types of bellows valve is NOT going to stay sealed for more then a couple of days of operation. Nothing beats a copper pinch off or a molten glass seal off. However for a DIY sealed CO2 laser, where air leakage is not going to do much harm, these things are great!
(From: Dan Glassburn (dan@niteliteproducts.com).)
Concerning tube pressures and discharges....
While under normal pressure range, the voltage at a particular current will be related to tube pressure - i.e., lower voltage equates to less pressure and higher voltage equates to higher pressure. However this is only within a particular range of pressure. Many tubes when running out of the normal operating range will reduce voltage then at some point start to increase tube voltage required to keep them ionized. This is more prevalent with solid bore tubes. The disk tubes just won't sustain ionization at lower pressures and higher currents. Many of the solid bore tubes will go into a "chirp" or oscillation which usually will cause havoc with the pass-bank. In the case of an air cooled, higher currents to deliver greater outputs will cause the tube to outgas more, especially if you don't have enough cooling. An easy way to see if the tube is low pressure in a 60X is to drive the current to it's maximum from and idle and see if the tube winks out. Many of the systems on the market do exhibit this symptom, although this is usually not how the customer normally uses the system. To get additional life from a low pressure air cooled, ramp the current up slowly. This allows the gas pressure in the tube to increase slowly and help sustain higher currents.
Bad plasma oscillations usually occur if the tube pressure has decreased to the inversion point, where the impedance transforms from slightly positive to negative. Passing through zero is the trouble point. Oscillations again start occurring if the pressure gets almost too high to light or stay lit. For example, if you have a leak on the pumping station while lit. I recently have had 2 I90s on station and one of them had burned so low that it wasn't lighting. Adding a few puffs of gas made it unstable and only able to stabilize at high current. Puffing in about two volts of gas made it appear stable for a while, but to get ignition in the next few weeks operating at low currents I had to puff in more gas, even though the tube seemed to be at the correct voltage. (For these tubes, the inversion point seems to be a plateau in the curve.) So I finally over puffed it by about 3 volts, and now she always fires on the first pulse. So the worst plasma chirp seems to occur at very low pressures.
Why does a healthy laser cathode climb in power (positive delta-T) when the plasma is lit? Rough explanation: Electron emission is an endothermic cooling process for the cathode. Sure you have to heat it to get it to emit, but it looses electrons, which by definition are a packet of energy. This results in cooling and a lower resistance of the cathode wire so there is an increase in current till the system reaches equilibrium. The cooling is perhaps roughly proportional to the cathode work function times the emission current.
In a healthy ion laser tube with good gas pressure, you have many more positive ions then electrons floating around the tube. Thus, the tube voltage climbs as the beam current increases. There is plenty of gas to ionize, electrons are rapidly adsorbed and used, they have a short mean free path from atom to atom, but lots of atoms to ionize. That short mean free path and higher probability of collision means they have a longer path down the tube, hence the higher voltage.
In a mediocre tube, there are less gas atoms due to lower pressure from plasma driven adsorption, and contaminates floating around the tube to adsorb gas and poison the cathode. But there is still healthy cathode emission, so you have a more of a break even condition - less gain, less ionization occurring, but still the tube is happy, just its failure mechanisms are starting to accelerate because of the lowered pressure. This is observable as less tube voltage climb with current and is the "Inversion Point"
In a low pressure tube, there is a big electron cloud with just a few ions. The ions remaining are energetically accelerated and plasma etch tube structures. The electrons are also highly energetic due to a lack of collisions and thus do even more damage as they pass through the tube. By this time the cathodes emission is probably shot from contamination and plasma etching. Impurities don't get buried if there is no gas to accelerate them into the walls. The plasma is unstable and probably moving around by this point. As you increase tube current, you have a excess of electrons floating around and so the tube voltage actually goes lowers.
So the first thing you do when you get a new laser, is chart the tube voltage versus current and check for a delta-T on the cathode. Delta-T is best observed by using a clamp-on AC ammeter around a cathode lead. Note that a few more modern high tech lasers have constant current transformers or autoswitching PSUs for their cathode and this may not apply to them. Also, if your cathode transformer is not correctly tapped or the buck boost on a older laser is misadjusted, you may not see a delta-T of either sort.
Lowering the cathode current moves the first dark space in the discharge back toward the cathode and can reduce the amount of ions generated, resulting in etching of the cathode's active emitting surface, as well as running the risk of cooling the cathode into the 600 to 700 °C region where cathode sag occurs. Excessively raising the cathode voltage (say .5-.7 volts) results in the outer emitting layer "boiling off" and emission falls off fast, resulting in short tube life. The remaining emitting layer will tend to "spot", i.e., most of the emission will come from the remaining emissive material at a cool spot and the cathode will eventually melt or burn at that region. Its is however better to err on the warm side then then the cold.
Some modern tubes with cathodes that are NOT tungsten-rhenium with a borate matrix or NOT a "dispenser" emitter run at about half the current of older cathodes, i.e., many Coherents now have the recommended cathode amperage stamped on the gas reservoir, and I have received reports from credible people that some Lexels are only consuming 15 to 17 amps instead of the normal 25 to 35 amps. This would be characteristic of a more modern Nitride ceramic cathode. So if in doubt, call the factory with the tube serial number and manufacturing date in hand. I have not yet heard of air cooleds using the expensive lower current cathodes, only the large and medium frames. However the ceramic cathode has a much longer life and greater electron emission.
Thanks to Kim, Dan, and Dale for their contributions on this subject, responsibility for errors or incorrectness is my own. --- Steve)
One contributing cause to cathode sag is something called the "Miller-Larson effect". It's a momentary plastic state in the filament caused by a change in the tungsten grain structure as its temperature passes through the range 600 to 675 °C. This results in a lengthening of the filament, and each time the filament goes through that range on warmup, it gets a bit thinner and a little longer. So, it might make sense to keep your filament on (or at reduced power but above 700 °C) during standby rather than shutting it off completely.
Start plasma can suppress emission from the cathode, and if you have a flakey cathode transformer or one that's borderline, the filament can still be cold. One symptom of a cold cathode is the tube catches and starts for a few seconds but winks out as a stable hot spot won't develop on the cathode until its heated more and the barium kicks in. There is a point as you heat the cathode hotter that there is no gain in emission. So you want to be above the Larson-Miller region and yet not too hot. I just found a excellent book on the lifetime and operation of large radio frequency vacuum tubes in broadcast duty, that's where this is coming from. In fact, the broadcast folks don't shut their cathodes down if the downtime is going to be less then 4 hours, they run them at what's called "Black Heat" at about 800 DegC to avoid the stresses of startup.
(From: Steve Roberts (osteven@akrobiz.com).)
If you just try to run at higher than normal temperature, it will be more likely to destroy the cathode permanently by blowing the emissive barium layer off, depleting it possibly in seconds!
Part of cathode initial processing and reprocessing involves heating the cathode to drive the barium out of the porous tungsten matrix at hotter then operational temperatures, but it must be done in a vacuum at 10-7 Torr. It DOES NOT work in the inert gas.
Thats what the famed car battery trick tries to do, blow the dirty non emitting regions off as dust into the tube and expose a little fresh barium. Thats bad too, its much better to have the tube reprocessed at that point then fill it with dust that will erode the bore and optics.
The more correct thing to try is undercurrent the cathode and invert it. Pure Tungsten has a known sagging temperature region where it is butter soft at 600 to 700 °C. However, the cathode is not pure tungsten - it has rhenium, barium, aluminum and calcium - so the exact softening point is known only to the cathode manufacturers. Normally, the cathode runs at 1,080 to 1,120 °C by design. At 800 at °C it actually starts getting tougher in shear and you're not going to do that well because of the additives. You'll end up cooking and outgassing if not actually melting the nickel or molyB section that joins the tungsten to the kovar in the lead throughs.
So if you want to try inverting it, get it to barely a warm soft red glow visible only in the dark and stay and watch it.
Operation at high currents especially, with improper power supply regulation (resulting in plasma oscillations), overheating, or just long hours of use, can lead to failure of the filament structure, excessive sputtering with degradation of the Brewster windows or mirrors, and other irreversible damage. And, accidents can happen as ion tube are relatively fragile structures. "You dropped the wrench where?" :( Any of these would require a more extensive amount rebuilding commonly called a refurb.
Whether a gas refill or total refurb is involved significantly affects the expenses involved. Depending on how much needs to be done and the size of the tube, the cost could be anywhere from a few hundred to a few thousand dollars. Where it is closer to the low end of this range, either of these is still likely to be a bargain compared to the prices of new ion tubes. For example, for a typical air-cooled laser like the Spectra-Physics 161B, typical list prices for replacement tubes are: 15 to 18 mW - $1,900; 20 to 25 mW - $2,750. Just think of what the biggies cost! OK, I'll give you a hint: You can buy a nice new automobile for the same amount of money!
However, On a small air-cooled ion tube (say less then 350 mW), the cost in materials and labor to do a complete refurb exceeds the cost of buying an entire replacement system surplus (about $1,000 if it's done properly with a new cathode, decent welds, etc.).
For just the refill, there will still probably be many hours of actual work over several days involved to gain access, pump it down, backfill, bake, pump it down again, backfill, bake, etc., and then seal it, with an extended burn-in using an ion laser power supply. During a major portion of this time, the tube will be the sole occupant of the expensive equipment required to perform the refill. If all they do is just put in some new gas and pinch the exhaust tube shut, only half a job was done and it probably won't last long. However, you could get lucky.
The amount of gas used is much greater than just what is needed to finally fill the tube prior to seal-off since each backfill/pump down cycle wastes a lot. However, even where krypton is involved (which is more expensive than argon), the cost of the actual gas is still probably not a major part of the total expense.
In addition to labor and gas that is used, you're paying for overhead in the way of the cost of the vacuum equipment, gas manifolds, inventory of gases and gas bottle rental, power supplies needed to test the tubes, rent, etc.
Refurbs include much more work on top of the gas refill possibly including: cutting out the old cathode (filament and possibly a getter) assembly, cleaning the bore, replacing glass or ceramic to metal seals, installing a new cathode assembly, replacing Brewster windows or mirrors, etc. These steps are even more labor intensive and the cost of the replacement parts in this case IS significant.
How much do you think an aspirin tablet costs to manufacture versus what you pay?? :)
Also see the section: Ion Tube Rebuilding in Your Basement?.
Now, for slightly more details:
(Portions from: L. Michael Roberts (NewsMail@laserfx.com).)
The procedure is actually quite complex. First the laser must be disassembled to remove the tube from the head - this process can take 3 to 6 hours depending on the type of laser head. If the cathode, glass, Brewster windows, etc. are in good shape then the tube can just be re-gassed. If not, these and other parts may need to be replaced - which would require disassembly of the tube. This involves glass blowing and/or replacement of the glass to metal seals (it isn't like you just remove a few screws and replace an O-ring!).
The actual regassing requires that the tube be hooked to a vacuum system with an airtight seal. Then, typically the tube is pumped down in a temperature controlled oven for around 24 hours. The heat helps to get the last remnants of the old gas out of the tube. The tube is then typically flushed with a new gas fill and the procedure repeated. This 'rinse and repeat' is used to remove any trace of contaminants in the tube. After the tube is pumped down and baked for the second time, it is refilled with gas and the seal pinched off. Even the pinching off and sealing of the tube takes great precision and special equipment. The tube is then tested and reassembled into the laser head where it typically undergoes a 24 to 98 hour burn-in to assure that all is well.
The procedure can consume 2 to 3 working days as well as many kilowatt-hours of electricity and involves a major investment in specialized and precision equipment - as well as a few dollars worth of gas. You also have to take into account that it is not always successful the first time and may have to be repeated. It is also possible that the tube can be damaged in the process and additional repairs will be needed - all of this factors into the cost. The refurbishing market is quite comparative and if prices could be any lower, they would!
(From: Steve Roberts (osteven@akrobiz.com).)
Hum, as one who does this from time-to-time, I can perhaps enlighten you a little more. Most lasers are sealed with a pinched off copper fill stem It looks like copper refrigeration tubing, but is actually oxygen free ultra high purity copper, about $60 a foot, and this is pinched off by a $3,500 tool that pinches off the copper and cold-welds it with about 30,000 PSI of force, using carbide dies. Here's the rub, the tube is filled with argon at a few Torr - a very small fraction of atmospheric pressure with argon. Should that pinch operation leak even slowly with a microscopic leak, the tube is shot and the rebuilder will be supplying the laser owner with a new tube, or having to completely rebuild the old one again. One speck of hard dust or a void in the copper is all it takes to ruin that pinch. Why not stick a valve on there you ask? Well, even the best valves leak or worse, will be turned by an overly inquisitive customer, a technician grasping at straws for some unrelated laser problem, or even by vibration of the cooling water. When a laser does have fill valves, they are attached to a another sealed low pressure chamber of argon - they don't exit into air.
When you have tricky steps like that in the process, you have to bill extra to pad your losses when you have parts of the process that are really beyond your control.
The header with lead through seals that holds the cathode has a series of glass to metal seals, each one of which is about $200 in quantity one, with a $250 cathode assembly mounted on it. All of these have to be installed with either a moly-mag or gold-indium braze because the use of any other welding process introduces flux into the tube which poisons the cathode and shortens life. The tube looks like iron, but it's actually nickel or something else exotic and it's a pain in the neck to weld. You have to do a weld with no pits or holes in it, even ones that don't go through the weld, as these trap pockets of gas or flux materials that can come back to haunt you. Indium and gold are two of the most expensive metals on the planet, and a typical large tube braze can contain enough of them in one or two joints to make a very respectable wedding ring. This whole thing has to be baked to 700 °F or more under vacuum, leading to possible cracks from stress where the metal bonds the ceramic. Then, the cathode has to be processed so it activates its electron emitting materials, and you risk burning it out at that point. It's also a risk that you may say, repair a leak, try to repump and then find out the cathode is shot. And then finding yourself cutting into the other end of the tube you just repaired. This can turn into a couple of days of work. Keep in mind that some or all of this must be done in a glove box and clean room to avoid contamination.
The whole vacuum pump and fill valve assembly on the work station must be baked out to drive out water and other organics, using turbo pumps or ion pumps, and if you're doing it right, you also have a $20,000 helium leak tester to spot those tiny leaks that could stop the tube from functioning weeks or months down the road, as well as an equally expensive residual gas analyc-zer to make sure you have a clean vacuum.
Even at $1,500 a repump and $4,000 to $5,000 a rebuild, it takes a lot of rebuilds and repumps to pay for that kind of equipment as well as feed your family and put money in the bank for hard times. This is why many rebuilders offer services beyond fixing lasers or also work on much less fragile YAGs or CO2s to help pay the bills.
If you're in the USA, figure at least about $5000 to $10,000 a year per employee in taxes and health care costs as well. And, the process is tricky enough without having to do your own sales, shipping and answering the phone, so you need some help. When you have lasers needing 30 kW and 6 gallons of water per minute, your facility costs are not cheap. It's about 25$ an hour to run a large frame - much more if you figure in the initial costs and rebuild costs and PSU parts. Heck, some of the fuses in a big one are $40 each!
Don't even get me started on the precision glasswork and polishing it takes to make a hard sealed Brewster window. Now if your making thousands of them, the price does come down a little, but one-offs are time consuming and not cheap.
As for gas prices, 250 liters of moderate purity krypton was quoted to me at $269 (plus hazmat plus the cost of the bottle), with research grade about $60 more. I could only find one place in the US that would sell it in small quantities, as it is now a commodity item for semiconductor manufacture. Everybody else wanted me to take out a contract and the price changed daily as the stock market went up and down. As far as I'm concerned that's not cheap compared to argon at $10 a fill for 2,500 liters.
I used to moan and groan about rebuild costs too until I saw what a cheap 'chop and pump' quickie rebuild does to tube life. Figure a large tube is 8,000 to 30,000 dollars from the factory without installation. That can easily justify the cost for rebuilding a commercial or scientific laser.
(From: Dean Glassburn (Dean@niteliteproducts.com).)
As one of those people that supposedly fleece others on refilling. Along with spending my entire weekend refilling a white light Lexel that just will not stabilize in pressure, I can attest to the difficulties involved.
I always have a problem with people saying that it's just too easy. This is why tube rebuilders and reworkers have such a hard time. Until you do this for a living and find out all the shortcomings and have to explain to your customers why your reprocess only lasted 2 months or why you completely ruined their tube, one should be a little less flippant about saying that there is nothing to it and all the parts can be found cheaply. It must be nice to have a cushy job.
And, if properly done, the equipment alone will cost more than a new 1999 Mustang GT.
Acquiring all the equipment and supplies if you don't already have them set up or have access to a lab just to refill a tube or two is at best justifiable only for its educational value. Just sealing the tube off again without *any* leakage isn't as trivial as it might appear for those with metal fill pipes like the ALC-60X. To do that properly, requires a special pinch-off tool that alone would run you the cost of several replacement tubes or maybe even a new laser!
If you have a few gassed out tubes sitting around heading for the dumpster and access to the proper facilities, by all means go ahead and experiment. You might get one of them to lase and last an hour or two after it is sealed. Or more if you're really lucky. Just don't make stupid mistakes like running the filament when the tube is up to air, letting in air suddenly blowing all sorts of crud over the Brewster windows, and zapping holes in the glass parts with your Oudin/Tesla coil! And, forgetting to cool it in the excitement once there is significant current flowing!
For additional comments on ion laser rebuilding, see the sections: About the High Cost of Refills and Refurbs, Caveat Emptor, and Britt Pulsed Argon Ion Lasers. To get an idea of what is involved in getting started before actually cracking the seal on the tube, see the chapter: Amateur Laser Construction since much of the equipment and techniques that are discussed there will be required
(From: Brian Bohan (camlaser@cambridgelasers.com).)
If you think Laser repair prices are high, you should have seen the price of plasma tubes before reprocessing companies existed!! Over the years we have talked a few customers though basic regassing, if you are seriously interested in trying it yourself, I would be more happy to give you some pointers.
(From: Steve Roberts (osteven@akrobiz.com).)
If you're trying to pump one at home, you need to get the pressure to at least 10-6 Torr during cathode outgassing or they die about 2 minutes after you first light them. Lifetime is proportional to how well you get that cathode matrix sucked out at pump, we try for at least 10-7 Torr or better. Most laser companies and rebuilders use a big ion pump roughed by a cryopump with cryotraps. Death is by loss of cathode emission from surface poisoning and then you get a tube that wont restart. The list of things that will poison a active matrix cathode include nickel, carbon, copper, sulfer, iron, kovar, any gas other then hydrogen and the rare gases, most metals and just about any organic or nitrogen compound. welding and brazing fluxes do it quicker then anything else. everything needs to be TIG welded without flux or else hydrogen furnace brazed. Outgassing the cathode properly takes about 48 hours of slowly turning up the variac, waiting for the vacuum to go back down each time. Cathodes are incredibly spongy - they hold enormous amounts of gas.
The gas is five nines or better purity. Ordinary tank argon may be used if you trap out the water with a getter. It depends on your gas companies refill practices, but often tank industrial grade argon is quite good. The difference between good industrial and pure grades in the US is whether the tank has been vacuum cleaned and heated in a oven before refill and then certified on a chromatograph or residual gas analyzer of some sort. Otherwise the tank is just roughly purged then filled. For experimenting its fine. The right grade of single isotope argon might pick up some more gain, but who could afford it?.
On the krypton however, I would not skimp, its very sensitive to impurities.
One would also like large amounts of decent helium or argon or xenon on one's station to clean the tube, but beware of using pure xenon in a narrow tube, it can and will explode. 20% AR 80% Xe is a good stable mix. One needs a lot of gas to "hot purge" the tube, i.e., fill it, run it on station, then suck it down quickly to pull out impurities. One must reach and maintain 1x10-6 Torr or better during cathode processing when you outgas huge amounts of gas. If you have a Coherent tube and are going in through the reservoir you may not need to reprocess the cathode if the tube is in decent health.
Start at high pressure and work your way down. Very low pressure fills can get really conductive (extreme negative resistance or thyratron effect) and blow the pass-bank. (Just did that one last week! ouch, about $140 worth of fuses and diodes.) This is difficult if you can't run awhile on station.
Also be aware of your tube's need for gas - is it a slow flow tube like a BeO bore or a fast flow tube such as a tungsten disk tube where the return bore may be 40 times the area of the arc bore. It's not fun to light off a fast flow tube and watch the plasma race down the bore and then wink out in a second or two from lack of gas.
You also need really good gauging, at least a well calibrated thermocouple gauge or preferably a zero to one Torr capacitive manometer gauge (expensive).
Your best weapon in getting a clean vacuum is a little heat! Heat tapes are wonderful, extruder band heaters are better yet and are not that expensive, and they clamp around standard pipe sizes. Heat the tube too. Crud then flows to the coolest part of the vacuum system.
Change the pump oil OFTEN if you're pumping crud.
Also, it might be wise to have some pure or inert waste gas flowing out of the connector when you go down onto the tube refill. That way it doesn't see air or water vapor, which makes getting a clean vacuum easier.
I can't afford to do this at home, but backfilling the whole vacuum system with argon when you're not using it makes getting back down to high vacuum a lot easier.
You'll need a good cathode transformer on a Variac. preferably a variac that can exceed normal line voltage by 30% or so.
I hope you have a turbopump or ion pump or titanium sublimation pump as a diffusion won't do it due to oil contamination. Oil gets to the coolest part of the tube fast an sucks in gas, only to release it when disturbed.
(From: Anonymous (localnet1@yahoo.com).)
I have actually SEEN a medical I-90 ion laser pumped down by a Welch Duo-Seal rotary pump. It put out right around 5 W after the guy pumped it down, then the power rapidly fell of after a few hundred hours. That might not sound like a lot of life time, but to an experimenter, a few hundred hours is a few hundred nights of fiddling with the laser for an hour at a time. :) If you ever want to go that route, I guess the key is to pump down, and seal fast. Don't give hydrocarbons time to diffuse into the tube. Or, alternately, you could set up a argon purge system so that there is continual gas flow through the tube to the pump, thereby helping to prevent back-streaming. You can pump out through the tip off (exhaust tube) if you have a tip off sealing tool and put some high grade argon in the fill port. Just a guess though. Don't hold me liable if this doesn't work in practice. One more hint: If you go this way, use diffusion pump oil in your mechanical pump (but first make sure the viscosity is what your pump calls for). Diffusion pump oil is more expensive but back-streams a lot less. If you have access to some cheap Fomblin (and it normally isn't cheap!!!) use that, for next to no back-streaming at all.
There are of course safety issues with going anywhere near a part that may have several hundred volts with high current available as in an ion laser, or several thousand volts for a HeNe laser. However, the typical commercial Tesla/Oudin coil is encased in a well insulated shell and the tip is isolated from the power cord, though by exactly how much isn't known. Doing this with a home-built Tesla coil or other RF generator may be much more risky though!
This is only likely to work on B-windows mounted on glass or ceramic extensions, not the type that Melles Griot uses with metal end-caps and frit seals on Brewster HeNe laser tubes or Cyonics/Uniphase air-cooled ion laser tubes. Use the Tesla coil only near the lowest settings or else there is risk of arcing right through the glass or ceramic, which isn't going to be very good.
(From: Phil Bergeron.)
You can bet I WAS nervous that the RF might cause the power supply current to arc to ground through the Tesla Coil(!) when I did this to the Control Laser 10 watt argon laser with the tube lit at 550 V at 35 A. I am here today so it did not do that. These small Tesla coils have like a 1/2 spark gap inside and I guess even with the Rf there was just not enough voltage thank god. I was sweating for sure and imagined I was about to become black and crunchy like on the Saturday morning cartoons. Hey. I did say there was some risk didn't I? :)
And it did work. The output went from 3 or 4 watts to over 8 watts in a minute. Alas the tube was so dirty the window just got coated again (cathode puke syndrome) and again until I got sick of it and declared the tube was a lost cause.
For a detailed description and diagrams, refer to U.S. Patent #4,625,317:
Internal Mirror Laser (Cyonics).
Depending on model, these tubes should do 5 to 50 mW or more of 488 blue,
514 nm green, multiline, or some other combination at 9 A.
It's designed for a 104 volt drop, so you can get it running with just a 7 to
10 ohm adjustable resistor and a bunch of filter caps.
The tube wants to see about 2.8 to 3.0 V for the filament at 15 to 25 A,
magnetron transformers with the HV winding whacked off are a good start.
Adjust for 2.7 to 3.0 V across the filament after a 30 second warmup. Make
sure it does not exceed 3.2 V after it is running.
Tube current is 9 A maximum, 4.5 minimum. A simple 'heater' based power
supply (115 VAC feeding 400 V, 20 A bridge, a few thousand uF of filter
capacitance, and a fan cooled heating element for a ballast resistor) will do
for a few hours until you buy or build a more sophisticated unit. But you
must include an accurate means of monitoring tube current and will have to
watch it like a hawk.
The recommended fan is 225 cfm minimum, a bit less then for an ALC-60X. The
patriot P2B3s are $22.95 from Marlin P. Jones and Associates quantity 1. Make
sure flow is sucking air out of the vee shaped aluminum shroud around the
tube.
I (Sam) built the laser head for a Cyonics tube out of the aluminum box
formerly from a defunct audio amplifier project, a fan ripped from a DEC BA11K
expansion box power supply, and an igniter based on a flyback core. The entire
power supply was also constructed from scrap parts. In fact, the only
component that had any association with lasers in its former life was the
tube itself. See the sections starting with:
Sam's Even Simpler Ar/Kr Ion Laser Test Power
Supply (SG-IX1/SG-IY1) for details including the purpose of the bicycle
tire inner tube and the successful revival of my originally non-lasing Cyonics
tube. :)
These tubes have mirror mounts designed very much like those of modern
HeNe lasers, with restricted regions that permit adjustment by careful
bending. However, this isn't something to do with Vise-Grips(tm).
Aside from being electrically live, this is guaranteed to ruin your
entire day. For minor alignment to peak power where gentle pressing
side-ways on the mirror mount is sufficient, adjusters can be installed
on these tubes to do the gentle pressing in a totally reversible manner.
Sometimes, these tubes will be shipped with three-screw locking collars
installed on the mirror mounts. They look similar to those found on many
HeNe laser tubes. See Three-Screw Locking Collars
Adjusters on Melles Griot HeNe Laser Tubes. Where locking collars
are present, the use of properly insulated hex wrenches will permit correction
of slight alignment errors. Where locking collars are not present, it may be
possible to obtain a set from Uniphase or from a dead tube. Otherwise, it
should be a relatively simple matter to fabricate a set. I have found that
Melles Griot HeNe laser full size three-screw locking collars actually do fit
the Uniphase argon ion tube. These adjusters were present on virtually all
Melles Griot (non-barcode scanner) HeNe laser tubes (as shown above) until
relatively recently and there are plenty of dead HeNe laser tubes floating
around. (Melles Griot has eliminated adjusters on most new tubes.) The
fit isn't perfect on on the tube in my 2214-30SLB - the adjuster's center
hole is too wide - but they do work. In fact, there are already four
access holes covered with plastic plugs at the correct longitudinal position
around the cylindrical head. (Maybe these were designed for a locking collar
with 4 holes!) It isn't possible to get three holes to line up
quite perfectly but it is possible to orient the adjuster so a well insulated
hex wrench with a ball (swivel?) tip can get to all three screws at not
so terribly steep angles through three of the holes. Where your head doesn't
have access holes, it would be very desirable to drill some.
Slightly more forceful use of the hex wrench will permit permanent adjustment
alignment such that the locking collar isn't needed, or is only needed for the
fine tuning. Determine which screw(s) increase output power when tightened
and then adjust them in very small increments, backing off afterwards to see
if the power stays at an increased level. The idea is to go just far enough
and no more. However, it's easy to go too far and mess up lasing entirely.
Use with care!
I had found that gently pressing on the HR-end mirror mount (with an
insulated tool!) of my 2214-30SLB it was possible to increase power from around
12 mW to nearly 30 mW or more at 10 amps (according to the power monitor
assuming 10 mW/V sensitivity, which tracks quite closely with a
Newport laser power meter). By adding a Melles Griot three-screw locking
collar to the OC (cathode-end) mirror mount, it's now possible to get 25
to 30 mW out of this old high mileage tube.
I also installed a similar locking collar loosely on the mirror mount at
the front of the head just in case I do any fine tuning. This required
removing the light pickup and running time meter PCB and photodetector
assembly as well as widening the hole in the plastic insulating plate.
The trickiest part was that the 2 wires to the photodiode need to be
carefully unplugged from sockets in the PCB before removal. Attempting
to tweak this collar may require a super insulated wrench as the front
mirror mount is probably the anode and would have multiple kV pulses on
it when the tube is started.
The non-lasing tests can be performed even before constructing any sort of
real power supply for the tube - to determine if it is worth proceeding.
The lasing tests can be done with either a simple pulsed power supply (See the
section: Pulsed Operation of an Ar/Kr Ion
Tube or a proper DC ion laser power supply).
For most tubes made by NEC and some Spectra-Physics models, the visual
inspections should be even easier. The gas return tube may be a coiled
capillary (possibly very fragile) external to the tube. And, for those with
glass end-bells, the filament will be clearly visible from the side.
For tubes with Brewster windows, check these for dirt, a metallic coating
(due to sputtering), or other damage.
A test for a gassy/up to air tube can be made using a low current high voltage
power supply capable of at least 5 kV (e.g., the start voltage from a HeNe
laser power supply would work). The idea is to determine if it ionizes with
the correct color (purple-white) and breakdown voltage (around 2 to 3 kV for
the Cyonics/Uniphase tube, possibly somewhat higher for some of the others).
Attach the power supply positive to the anode and the negative to one of the
filament leads. Make sure to use a low current supply (a few mA at
most)! (A Tesla or Oudin coil can be used to check that the gas ionizes and
that the color is correct, but not the voltage.) If these tests fail, at
least you have the option of getting the tube rebuilt or using the tube for
your own experiments in tube refilling! The filament and optics will still
be fine (unless someone before you powered it up in its current state or it
was running when the leak occurred.)
The discharge glow isn't very bright when excited by a low current source like
a HeNe laser power supply or Oudin coil - view in a darkened room (but watch
out for the HV terminals!). For lasers with external mirrors, it may be
easier to see the glow if one of the optics is removed. For example, on
the Lexel-88, stick a coin in the slot on the black or silver (depends on
model)end of the HR optics mount and turn counter clockwise to remove plug.
Then grab the optic with tweezers and gently pull. Note location of O-ring
which can be either behind or in front of the optic. Set the optic face down
on lens tissue so it won't pick up dust and cover the cavity with a plastic
bag when open. The glow is soft violet.
Bring up the voltage slowly while looking in through the cathode-end mirror
The filament is a thick coil - a single layer solenoid of a few turns - which
you will be viewing end-on. Watch as it starts glowing - all turns should
stabilize at approximately the same brightness.
At 2.5 V the filament should be glowing strongly with all turns more or
less equally bright. If some turns are dark, the filament is shorted in
spots. The tube may still work somewhat, though it may be necessary to limit
maximum tube current so as not to overheat the remainder of the filament.
The entire filament can be safely brought up to 3 V or till its brightest
parts are orange-yellow but don't take it any further.
For a newer tube, the filament will be well formed, symmetric, and centered.
An older or abused filament may exhibit 'droop' or distortion but as long as
the central bore is clear, the tube should be capable of lasing. You should
be able to see clear through the bore in the center of the filament and out
the other end of the tube. In the Cyonics/Uniphase tube, the three (3) gas
return holes around the periphery may also be visible.
If these tests all pass, it is probably worth proceeding with the construction
or acquisition of a proper power supply for the tube.
(Note that some of suggestions below only apply to the Cyonics/Uniphase
tube - some others may not survive!)
You then want to fire it up and run it, say 1 minute on or so. When all other
causes of failure are accounted for, the igniter pulse becomes the biggest
factor in the life of the tube. It (a) buries gas, (b) blows chunks out of the
cathode until the cathode spot forms and (c) deposits junk on the internal
optics. I'd say don't worry about it, as the Cyonics/Uniphase tube is really
overbuilt. It has the same cathode and anode as a 5 W tube, and has so much
metal that you need some time to heat it up. Unlike an ALC-60X tube which is
critically cooled, this one will take some punishment, and no amount of energy
you can dump into it with a 115 VAC line will rupture or implode or melt it.
I have one sitting in the garage that contained all the fragments when the
tungsten disks and cathodes melted, it held the vacuum as well.
You could run the Cyonics/Uniphase tube at 14 amps for short periods of say 1
minute and do no damage at all, provided you have proper cooling and the
cathode is no higher then 2.8 V. It's when you log hours and hours of
operating with an improper supply that you will damage it, not during short
term testing. In fact tubes are sometimes ran without cooling to recondition
them a little (but I probably wouldn't recommend doing this at home unless the
next stop would be the dumpster!)
It's a good idea to power up and see what is going on then switch off, but it
may take 5 to 7 seconds after ignition for the cathode to come up to final
temp as pressure rises from the discharge heat. Hook a voltmeter across a
current shunt in series with the tube, if it goes way out of range, just
switch off. Don't panic. This thing has a much larger thermal mass then a
HeNe tube, and the metal will desorb the pure gas over time, unlike the HeNe
glass which destabilizes the discharge when it heats up enough for ion
migration and the bore conducts/outgases CO2 and N2.
I (Steve) do my initial checks for 30 to 60 seconds with the lid and fan off,
often taking my time to stick the voltmeter leads across the tube and then the
cathode, and my DVM is slow. Then I simply slide on the fan and cover and run
it at 4 to 5 A for a while or shut down for 5 mins, which is about a 1% duty
cycle. :-)
You will pop a weak ceramic-to-metal seal long before you ever risk a
catastrophic failure. The main safety hazard is reaching in and touching line
side. If it fails the discharge will go out when the cathode opens up.
There is a BeO (beryllium oxide) warning on the tube for those who would
actually grind or smash a tube. Recent discussions with a shop that laser
machines BeO have revealed that the hazard is only for a certain particle
size. They cut it with CO2 lasers every day and set it up so the particles
are outside a certain range without wearing space suits and using only a
simple HEPA filter to recover the BeO dust for remelting as it's an expensive
material with resale value.
(From: Bob.)
I remember a fellow who used an Oudin coil and two slice toaster to
demonstrate 60X heads that were pulled from photocopiers. The Oudin was used
to strike the tube and the toaster for current limiting. (I think he used a
battery for the filament but am not sure as this was some time ago).
He had a rectifier but no filter capacitor, and when he struck the tube by
allowing a spark to jump form the oudin coil to the tube lead, the TV and
radio got a ton of white noise interference. Current control was digitally
determined by selecting one slice or two. :)
(From: Sam.)
Well, I would suggest enhancing this kludge with a filter capacitor but
such an approach would certainly work for really quick and dirty testing.
No, I am not recommending it! :) The minimum to be used for anything serious
would be something along the lines of SG-IT1 or SG-IX1. These aren't really
that much more complex electrically but would be a lot safer for you and your
tube. See the chapter: Complete Ar/Kr Ion Laser
Power Supply Schematics for details.
You can use a HeNe supply between the anode and filament. There should
be a weak ionization flash, and perhaps a sustained discharge glow if your
HeNe supply has a strong enough start pulse to break down the resistance
between the two electrodes. If the voltages produced by the HeNe supply
are too small, you'll see nothing if the tube is good or bad.
If the tube is glass, and you can see the filament and "getter" area
inside it, a sure indication of a dead tube is a whitish or oxidized
appearance on the normally silver metallic flash evaporated off of the
getter when the tube is first started. If this area has a shiny
metallic mirror-like appearance, things are good. If it looks like
white powder on the inside of the tube, then it is probably up to air
pressure and is dead. If the tube casing is opaque ceramic, you won't
be able to see inside to use this test.
You can positively test for gas integrity with a small Tesla coil,
such as the Model BD-10A made by
Electro-Technic
Products (Go to "Leak and Pinhole Detection"). When it is
turned on, and the tip is brought close to one
of the clear glass sections of the laser tube, the high frequency
radiation from this unit will ionize the gas inside the tube, which
will produce a dim bluish glow. If you can see this, then the tube is
probably okay. If you can't see any ionization glow, then the tube is
probably up to air pressure and is dead.
If you are attempting to test the tube with a Tesla coil or similar
device, it is important for you to stay away from the windows, the
filament, or any other glass-to-glass or glass-to-metal seals on the
laser tube. A spark from the Tesla coil can penetrate and weaken the
seals, and possibly cause a leak. Try this test only on sections of
the tube which are clear glass, and preferably 100 mm (or four inches)
away from the windows or any glass-metal seals.
However, even if the tube is gas intact, it could still be at high pressure
and be impossible to operate with a good power supply. You'll have to try
it to find out for sure.
If you have an oscilloscope and a HV probe or are willing to risk the normal
probe monitoring the voltage on the power supply side of the igniter
transformer or blocking diode, see what happens at the instant the igniter
fires. I haven't tried this but would expect that if there were a power
supply problem, the voltage should dip to well below 140 V - down to the
tube sustaining voltage at the peak current of the igniter pulse, perhaps
110 or 120 V, then drop lower, before cutting out. This would also occur
if the light preamp were full on due to a circuit problem since in this case
the power supply is also turning off the current. However, if
the tube is high pressure, the voltage would not get down this low,
perhaps not even to below 140 V, then climb up to the point where the
power supply can't maintain adequate current through the tube.
An Oudin coil is a sort of hand-held Tesla coil that is adjustable in output
and runs off AC line. It may also be called a Tesla coil or spark coil. A
new one costs about $160 from neon sign equipment suppliers but these are
often found in high school physics labs and other places working with vacuum
systems. You may be able to borrow one for for an afternoon (or longer).
A home-built alternative is also possible. See:
Home-Built Substitute for Oudin Coil.
The Oudin coil is fun as sometimes you get all the energy stored in the PSU
caps to flow through the air in a inch or two spark, pzzzzzap-flash-bang!
(But see WARNING below if using a regulated power supply that might not be
happy having its output shorted.) Using it around plastic bags of pure noble
gases also is fun. You get 6" long sparks or glows. The only glitch is if
you turn it up too high around your ion laser tube, it can puncture the
Brewster stems which would be bad news indeed. The high voltage low current
RF field doesn't damage the PSU or support electronics, even when you see
corona off the leads of solid state components (though I wouldn't press my
luck with MOSFETs or CMOS devices and especially keep clear of the light
sense amp in the laser head!), yet it will light a nearby fluorescent lamp to
the point you can read by it.)
WARNING: The idea is to get the ion tube to start conducting. If an arc
develops outside the tube to the HV return, it may damage expensive
power supply components not as a result of the RF but due to the high current
discharge pulse that follows. The PSU expects to see the normal ion tube
voltage drop - an arc in air could be much less than this.
Plug the coil into the wall and set the knob on the coil for a 2 to 4 cm spark
to metal. Then hold it at the cathode-end of the tube where it can hit the
air riser (metal box below the fan) and the bare ceramic of the tube with the
power supply on (and trying to start). This will totally ionize the whole
tube volume and may allow enough current to flow for the discharge to start.
WARNING: Beware of turning up the spark coil so high you punch holes in the
glass to metal seals on the Brewster stems.
Note: When using a power supply like the Omni-150R with light feedback, a
failure of the preamp in the head or component in the control loop can also
result in the laser refusing to start with the dreaded 'tick-tick-tick' sound.
If anything in the light control chain is pegged to the supply rail (or even
if the control is turned down too low), the power supply will reduce current
below the minimum required to maintain the discharge as soon as the tube starts
thus effectively aborting. The light preamp can be damaged if you mishandled
the sensor - it has piezoelectric properties and can blow the op-amp input if
banged or dropped.
If it succeeds, running the laser for a few hours will eventually bury enough
gas that the igniter will handle starting on its own the next time.
This problem results when you don't run the laser for a long time and it
outgases, I shudder to think how many perfectly good tubes have been thrown
away that just needed a little assist.
A cure for this is to run these lasers at least 30 minutes every three weeks,
like HeNe lasers, they like to run, it takes a couple of months of setting
there for the pressure to really build. On some tubes this doesn't happen
but it is a common problem even with the large frame water-cooled units and is
worse with a krypton fill.
The Krypton red and yellow lines share the same upper state, and what
determines where they fall to ground state is the gas pressure. If your
tube is on the low end, you get a lot of yellow; if it is high you get a lot
of red. Large krypton lasers have a pressure control pump built into the tube.
Two solenoid valves and a gas reservoir form the pump.
Lower tube pressures cause more gas to be buried, leading to a runaway cycle.
An argon tube with clean multi-line optics installed that only emits the 488
blue line is a dead giveaway for low pressure and end of life. However,
it could just be dirty optics as the 488 is the highest gain line, it will
lase with a 90% reflector, so try proper cleaning before throwing in the
towel.
Also see the section: Tips for Maximizing Ion
Laser Tube Life.
(From: Steve Roberts (osteven@akrobiz.com).)
9 Amps is flat out max for an NEC tube. They go on forever at 6 to 7 amps.
I've seen over 40,000 hours when most others only last about 6,000.
They only do 25 mW or so when new, and measuring blue light accurately without
a thermocouple meter or very high quality silicon meter is difficult. If you
can power up the light card with a regulated +/-12 or +/-15 VDC supply and
the laser has a calibration sticker on the light jacks, it should be fairly
accurate, +/-10% or better. Warning, light card and ignite circuit are on
the same board - watch for ground loops. NEC brings out the light signal to
their PSU for display on a front panel meter, so your unit should have been
calibrated somehow.
NEC tubes want to have a cooler filament than tubes like the ALC-60X - about
2.6 VRMS (compared to 3.0 VRMS for the others). They are really fussy about
this - overtemping the filament causes it to outgas which can lead to hard
starting.
NECs are also particularly susceptible to ripple and overcurrent. They have
some bizarre start up circuitry in their PSUs to insure ignition without
damaging the tube by actually current limiting their start pulse.
Sometimes for seemingly no reason at all, NEC tubes go south (probably they
outgas some hydrogen - not good for the discharge) and need a little starting
help. However, first check all the zeners and SCRs on the ignite board as well
as for shorted/leaky caps. Since the NEC laser head generates its own boost
voltage, something may be wrong on the ignite board. Make sure the filament
volts/amps are adequate - cold filaments won't ignite!
If this doesn't turn up anything, just getting some Tesla type HV near the
glass bell may help. Leaking a spark across to the metal fins on the body may
do it as well. Set the laser on a wood board along with a battery powered
Tesla coil. Snake a single conductor lead over to the laser with the top
cover off, then let the wood act as the return to the HV, if conditions are
right, you might get some faint sparks to the fins to pre-ionize the gas.
As soon as the tube starts, you have about 30 seconds to get the hood on (to
provide cooling), so don't panic. It needs to run to get the pressure down,
DO NOT interrupt the current flow for any reason. (Well, at least not anything
less than life or limb threatening!)
Assuming none of this works, you can consider the following two procedures
but they ONLY apply to air-cooled NECs due to a unique structure in the tube.
DO NOT attempt this with other lasers:
There is a lot more to a Oudin coil then a simple interrupter driven ignition
coil circuit (though this is what it may appear to be from the outside). They
have some unique windings for enhancing resonance and some caps in critical
places. Mine is from the 1930s, insulated by wax. A few years ago I had to
rewax it - paraffin doesn't work, only pure beeswax for some reason (so I had
to rewax it again). I wish I had drawn out the details, but that is a true
resonating Tesla coil in every sense, not just 60 hz pulsed AC.
I've never got a start boost off anything but a pure RF spark, not a sharp
pulse! - hence the confounded wood block setup I suggested.
Oudin coils source so little current they generally do not harm the start
boards, even the UJTs. It's not uncommon to stick one on a laser for 30 to 45
seconds with no harm to the electronics, its extremely high peak voltage,
but NO current, but its broad spectrum RF that seems to contain (and perhaps
adapt) to the natural frequency of whatever it's trying to excite, gas-wise.
When we hit a tube with a true Tesla discharge (not just a few sparks here
and there), it ionizes a large percentage of the atom in the tube, as the
broad spectrum RF finds and crawls down every possible waveguide mode in the
bore to the anode and beyond. This provides a path for the main start pulse
through the hydrogen. On some tubes, the getters (if the tube has them) will
start reacting and bite down on some hydrogen when they heat up and hot
energetic ions are flying around to react with them.
<begin humor>
The basic BD-10 Oudin coil, Tesla coil, spark coil - whatever it's called -
is adequate. One supplier is
Electro-Technic
Products. Another one is EGL -
Your Neon source (click on "accessories", "spark coil"). They both sell
5 or 6 versions. The cheapest is perfectly adequate.
These also turn up on eBay quite frequently more likely to be called
"Hand-held Tesla coils" or something similar. Sometimes, the listings
associate them with quack medicine of the early 20th century but that's
fine as long as they do what you want. :)
The rear and front sections are attached via two heavy wires. The front
section is fully potted so I don't know what's inside but at its end is
the output high voltage contact.
The only user serviceable parts inside a BD-10 are the power cord and
interrupter contacts. The transformer in the bottom section could
conceivably be rewound if needed but it doesn't appear as though that
would be much fun. Only Steve knows what's inside the top section but he's
not telling (see the previous section).
I acquired a vintage BD-10 (original General Electric) off of eBay for $30
that certainly looks and feels exactly like the one we used to test for
vacuuum leaks in the cyclotron at our high school. Oh, and also for chasing
unwanted visitors. :) (See the sections starting with:
The Central High School Cyclotron.)
See Electro-Technic
Products for a sample photo (next to "Pinhole and Leak Detection").
It worked fine generating sparks over an inch and a half long, but had
a frayed and rather stiff cord, which I decided to replace. Everything
inside appears to be in pristine condition, so maybe it's not really
that old. However, there may be some minor differences in
construction on really modern ones. Here's a step-by-step disassembly
procedure for reference:
Reassemble in reverse order.
The typical hand-held Oudin coil (may also be called a Tesla coil or spark
coil) from external appearances would seem to be just a line powered high
voltage transformer driven like an ignition coil with a buzzer type of primary
interrupter. However, from the section: More
on Oudin Coils this is not the case (at least for some types) and produces
significant radio frequency (RF) energy - not just a stream of pulses at 60 or
120 Hz.
A high voltage high frequency source can be constructed from a TV or monitor
flyback transformer using a transistor or two for drive. Alternatively, an
interrupter (relay/buzzer) or low pulse rate transistor driven circuit using a
flyback or ignition coil can be enhanced with suitable resonant components
(additional capacitors) to promote the generation of RF energy. Note that if
a flyback is used, it must NOT have an internal rectifier as that would simply
result in
Cyonics Tube, Testing, Hard-to-Start Tubes, Tube and Igniter Problems
Cyonics Argon Ion Tube
This is typical of a well-built small sealed internal mirror argon ion tube.
Uniphase Corporation acquired
Cyonics in the late 1990s.
Cyonics/Uniphase Mirror Alignment
After many hours of use and thermal cycles or abuse (e.g., running at high
current or loss of cooling), alignment may change enough to significantly
reduce output power, possibly to 0 mW. Even if there is no output, all that
may be needed to restore full power is a very slight adjustment of mirror
alignment.
Testing a Small Ar/Kr Ion Tube
This was originally written for the Cyonics/Uniphase 22XX style internal
mirror argon ion tubes but portions apply to others like the ALC 60X or
Omnichrome 532 with minor and obvious changes. Comments on these and some
other types are also noted below.
Preliminary Tests
Even before attempting to get a tube to lase, there are some very basic tests
that can be performed to determine if it is even worth building or acquiring
a power supply:
Initial Power Tests
For initial tests with a home-built or commercial (DC) ion laser power supply,
the tube can be run at a very low duty cycle without the cover/fan in place.
Power up for say 4 to 5 seconds and power down, just don't rush, unlike HeNe
tubes, these do not stablize instantly. This will determine that there is
nothing catastrophically wrong with the tube or power supply.
The Ultimate in a Bare Bones Test Supply
This almost qualifies for inclusion in the laser humor department.
Though, it's not as simple as the one using a set of 18 car batteries
in series for a "portable" system. :)
Carl's Notes on Testing an Ion Laser Tube for the First
Time
(From: Carl Hannigan (carl@ppsfx.com).)
High Pressure Tube Versus Power Supply Problems
One of the common symptoms found when testing a (possibly newly configured)
system for the first time is that the igniter fires, the tube flashes, but
the discharge won't stay lit. (If the mirrors are properly aligned, there
will also be flashes of laser output but mirror alignment may be another
unknown so these won't always be present.) There may be a relatively easy
test to determine whether the tube or power supply is at fault. The following
assumes a small air-cooled ion tube running at around 100 V but obviously
can be modified for other lasers.
Hard-to-Start Ar/Kr Ion Tubes - Outgassing and Keeping
Your Laser Healthy
If you don't run an ion laser periodically, gas is released from the tube walls
and then it can build to such a level that the tube won't start. So, you then
have to spark them in the right place with an Oudin coil of the neon sign
shop/vacuum lab variety till they start and bake them at full current for a
few hours to drive the pressure down by burying gas into the metal parts of
the tube. This is easy when you have a plasma core that is hotter then the
surface of the sun and has 170 nm extreme ultraviolet photons bouncing around
from every ionized atom in the tube!
Notes on NEC Argon Ion Tubes
The comments below were prompted when the NEC-3030 tube in the laser head Ben
was using with his version of the home-built SG-IL1 power supply described in
the section: Ben's Linear Ar/Kr Laser Power
Supply (BJ/SG-IL1) refused to start for no apparent reason.
More on Oudin Coils
(From: Steve Roberts (osteven@akrobiz.com).)
{
<end humor>
Having your bore transported through wormhole space by a scalar RF field of
course helps. With any sort of luck, the hydrogen doesn't get beamed back
from the parallel universe. Oops, that was meant for alt.conspiracy...
Sorry, wrong forum. :)
}
BD-10 Anatomy and Repair
The BD-10 is the classic hand-held Oudin, Tesla, or spark coil that runs off
the AC line using a set of magnetically activated contacts as an interrupter.
There are no active electronic components inside. The BD-10 consists of two
sections. For lack of a better terminology, I'll call them the "rear"
and "front". The rear consists of a heavy primary coil and thin secondary
coil in what appears to be an autotransformer type of configuration. The
primary is in series with the AC line and the interrupter contacts. One of
the contacts is attached to a magnetic piece which is attracted to the
core of the transformer when current flows through it. The other contact's
position is adjustable via the knob on the bottom which thus determines
how much pressure is applied to close the contacts - the more pressure,
the longer the field builds up in the transformer core before current
is interrupted by the contacts opening. Thus, when the adjustment knob
is fully clockwise, the highest voltage is generated. Spacer washers can
be added under the knob to limit the maximum voltage if desired.
Home-Built Substitute for Oudin Coil
Oudin coils or hand-held Tesla coils aren't the sort of thing to turn up at
your typical garage sale (although you might find one at a high-tech flea
market or hamfest). However, it should be possible to construct something
suitable for dealing with hard-to-start Ar/Kr ion tubes at minimal cost.
