The HeCd laser can produce a high quality beam at 442 nm (violet-blue) and 325 nm (UV) depending on the optics. Typical power output is in the 10s to 100s of milliwatts range. In terms of popularity, the HeCd laser probably ranks behind HeNe, Ar/Kr ion, and CO2 gas lasers. Although its wavelengths may be highly desirable for some forms of spectroscopy and non-destructive testing, they are pretty useless for laser shows and other common hobbyist applications. For that reason, as well as the higher complexity (and cost), one doesn't see these lasers nearly as often as the more common types.
HeCd laser tubes ARE more complex than those used for HeNe, Ar/Kr ion, and CO2 lasers. In addition to often using a heated filament/cathode, they also include a reservoir for the cadmium metal and a heater to control its vapor pressure, an overall heater and thermal insulation to control tube and helium temperature/pressure, and various sensors inside the envelope to monitor these parameters for use by several feedback loops in the power supply. Even if the mirrors are internal, they are often adjustable to some extent.
Some photos showing typical HeCd laser heads, tubes, and power supplies are available in the Laser Equipment Gallery.
WARNING: Since the temperature of the tube affects helium and cadmium pressure, both critical for proper operation, the entire tube is usually covered with thermal insulation. In older lasers, THIS WAS ASBESTOS! Thus, HAZMAT handling procedures apply any time maintenance or modifications are being done inside the laser head!
This is one big Helium Cadmium Laser tube, measures 32.5" long. Made for the typical 3 rod resonator frame. Note the typical bright gold/yellow color of the discharge with no pink from air leakage visible.
This tube goes in the model 374 laser head which is a larger version of the 456 head. The head's model number indicates the lasing wavelength: A 374 is 325 nm (near-UV) and a 474 or 456 is the 442 nm (violet). Omnichrome is now a part of Melles Griot. I am not sure if they still make this design since they also bought out Omnichrome's competitor, Liconix.
Here are some of the specifications for the tube:
(From: Daniel Ames (dlames3@msn.com).)
What are the symptoms?:
The Cd reservoir is a small glass ampoule located either in the center of the tube's length or at either end. It will have two small wires attached to a small cylindrical aluminum heater and will be covered with some type of insulation WHICH COULD BE ASBESTOS (fluffy fuzzy stuff - take care)! If it is asbestos, wet down the insulation first to minimize dust. You will have to carefully remove or at least ~ lower the the heater to see inside this Cd reservoir. A careful slit in the insulation near the top of the heater should allow you to gently slide the heater downward enough to see inside. If the reservoir is empty (cadmium looks sort of like aluminum) then the tube will need to be replaced. If not empty, than that is another story for a follow-up posting.
Where the tube does not start, how long has it been since the last time it did work? Has it been more than 6 months? Is the power supply producing the correct starting voltage? Also, check your connections, especially that the circuit board inside the head is plugged in securely.
If you are getting the high voltage, try chilling the head for awhile before trying to start it - like by putting it in a plastic bag in the fridge or outdoors overnight. Then bring it in and try starting as soon as possible before it warms up. This might lower the helium pressure enough to help it start, (depending on the weather, or the refrigerator.)
If the tube is actually cracked or broken, I would advise patching up the damaged areas with room temperature RTV Silicon (e.g., bathtub caulk) or sealing the entire thing in a strong plastic bag, and storing it (outside) until you find out where to send it for repair or salvage, or how to dispose of it safely. Cadmium is considered to be a toxic heavy metal, even the vapor is not fun and tastes terrible.
(From: R. J. Zimmerman (safed_chuha@yahoo.com).)
HeCd lasers have a filament/cathode. Near the output-end of the laser are three leads on metal feed-throughs in the glass. These lead to two filaments. You can measure if they are open circuit with a multimeter.
If they are both open circuit (with the center pole between them) then the cathode is dead. Still, the laser would probably start if it weren't up to air. This would kill it very quickly from a condition that Omnichrome called "cold cathode." The plasma would search for electrons anywhere it could and eventually find them on the nickel ion shield on the inside of the laser just before the output mirror. In the process of stripping out electrons, nickel gets sputtered all over the inside of the glass near the output and it looks much like an old burnt out vacuum tube with a mirrored finish. The mirror gets coated too and this condition which is very rare, destroys the tube as there is no reasonable way to replace the optic.
Of course opening the glass would also require you to reprocess the laser and backfill it with a near vacuum of Helium, so it will be hard to find someone to do this indeed.
(From: Don Klipstein (Don@Misty.com).)
I have found that white cotton underwear and white cotton socks fluoresce blue from wavelengths as low as the 253.7 nm mercury line and as high as a scandium line that I believe is 391 nm. Works on everything in between such as the 313 nm and 365 to 366 nm line clusters of high pressure mercury. That's my best bet for 325 nm.
I doubt that most fluorescent lamp phosphors would work. But the phosphor from a high pressure mercury lamp is supposed to work from 313 nm and I give it a good chance at 325 nm. Most mercury lamp phosphors don't work at 365 nm but one does at least somewhat - from the oddball and probably obsolete Westinghouse "Standard White".
(From: Steve Roberts (osteven@akrobiz.com).)
Hey, we have a UV HeCd laser here at the University. I just tweaked the system it's in. However, I learned the hard way about UV Hecds - the UV they pump out is kind of hard to detect and at 13 mW was swamped on my fluorescent paper by the bore light. I had to eat some humble pie when after proclaiming that it wasn't lasing. But, if you move the paper way out from the laser, the beam is there and the bore light isn't. The beam diameter was very large (1.1 mm) compared to your average HeNe or air-cooled argon. Trying a bunch of different fluorescent materials resulted in disappointment as very few seemed to be pumped well by the 325 nm wavelength. Over the years I've been contacted by many people who said they had surplus Hecds that didn't lase, even with low hours, no matter what they did. I have a funny feeling quite a few of those were actually UV machines. It does show up on the power meter however.
I once had a long positive column self heated HeCd laser where the gain curve was centered around one temperature and fell off to near zero 5 degrees C either way from that point. That point varied from tube to tube, but was not mentioned in the manual, only the fact that if you didn't have the temperature spot on, you had no gain. Modern units are designed for a slightly wider operating range and usually have a glass frit and a molecular sieve to adjust the helium pressure as well. So don't dispare if you don't see any lasing for a while, HeCd lasers are difficult and the PSU may have to be matched to a given laser. Your not just looking for the CD melting point, your looking for the place where it has a high vapor pressure, which may be considerably hotter. Ask the manufacturer, they sometimes will tell what you need on a obsolete model, especially if your a student.
(From: Daniel Ames (dlames3@msn.com).)
Ya, these are some really weird and overly complicated animals. I have spent probably a hundred hours or more trying to diagnose my non-working HeCd PSUs still not working).
I am a little uncertain about some of the problems that are mentioned below.
Power output is lower than expected:
Are you sure it is dying? How does the cadmium supply in the reservoir look? You should be able to see the (silver to aluminum in color) Cd metal from either end of the reservoir without having to remove its heater and covering. What color is the side discharge of the tube (gold, white, or blue)?
Is the power supply the right one for this tube? These MDL100-A's can be set up to run the larger Omnichrome series 74 heads also by changing the two large resistors near the fan in the PSU and readjusting the settings.
The He and or Cd heater control circuit may be out of adjustment. Also the reference voltage and current adjust trim pots could possibly not be optimized for your tube and power supply. From what I understand, they need to be tuned to each other to work their best, not just any MDL100 on any 56 head. It could still work, but could cause either a shortened tube life or low power output.
I would NOT recommend trying to adjust any of the electronics as they ALL interact with each other to allow the tube to operate correctly. 0r incorrectly!
NOTE: If you are going to try adjusting anything, trimpots or optical alignment, Don't try to adjust anything without first making two reference marks, one for the stationary position and one on the item being adjusted, i.e.: optical alignment nuts and/or the trimpot adjust slots. Plus, you would need to know what the ball-park settings should be. I do not remember what they should be set at, sorry.
Could the optic's alignment need a little fine tunning? Yikes, for UV you will definitely need a power meter!
Maybe you can just apply some physical pressure in different directions (after the tube is warmed up and stabilized) to the resonator rods separately, or on the mirror plates and watch the power meter. (WARNING: Take care with respect to the high voltage at the HR end!) If the output power increases, then the optical alignment needs some fine tuning.
And, for the low output power: Did you measure this output power with a properly calibrated power meter that is designed for this wavelength? If you used some other power meter, or a generic meter, most likely its response will not be accurate at the violet-blue or especially the UV wavelength. Most likely it will read lower that the actual output. If you know what type of photodiode it uses, then if you can find the optical response specs for it, you will be able to figure out the response % or ratio for the 442 or 325 nm wavelengths (or the combination if running multiline).
Fan doesn't run:
Is this the fan in the laser head or the power supply?
Filament lights up but the discharge will not come on:
There is probably no high voltage. see below.
The high voltage cable arcs inside:
I hate it when this happens. :(
WARNING: From experience, DO NOT TRY TO USE THIS HV CABLE, or you could damage the electronics in both the psu and the head!
Unable to locate a replacement for the SG3527J driver chip:
Right... Silicon General was bought out by Unitrode but they discontinued this chip or it was extinct before they took over SG's line and did not bring this chip back. Are you sure it is not an SG3627J (not that this helps much)? All my MDL100As use this chip for the HV control chip. The SG3627J and it's interchangeable cousins, SG1627 and SG2627, became extinct back in 1991 and Omnichrome no longer has them in stock, but they do make an after market circuit board that will substitute for it. Last time I checked (approximately a year ago), it cost about $70.00. Yikes!!!! I hear that there is a way around this, but I am not sure of how it is done.
The high voltage transistors (MJ10007) cross reference to a NTE97. The SG1524/SG2524/SG3524 cross reference's to a NTE1720. Note: These 3 SG chips are the same chip but with different operating temperature ranges, either will work just fine with these PSUs.
(From: Sam.)
I believe the SG3627 is a multiple channel MOSFET driver so this would imply that some sort of transistor network be installed in its place (for each channel). This shouldn't be that difficult to implement or find a substitute part (though not pin compatible).
The SG1524/SG2524/SG3524 is still available for less than $1 from Mouser and other suppliers. I expect that and original MJ10007 replacement can be found as well. (ECG/NTE/SK parts can be used in a pinch but may be of variable quality and are almost always much more expensive than the real thing).
(From: Daniel.)
I have now found a distributor that most likely still has some SG1627 Chips for Omnichrome HeCd (and other) switchmode PSUs. I had contacted SGS Thomson through their web site, who referred me to a distributor, so I contacted them back in March, 1999, and they had 97+ of these chips. Omnichrome used many of the SG1627, SG2627, and SG3627 for the same chip. However this distributor does have a minimum order, (*&%$##*) which at that time was $100.00. The more you (or we together spend) the more chips for the dollar we get. Contact: Ron Holmes (ronh@accgrp.com).
High voltage is not present:
Many different things can cause the high voltage DC to not work or shut down: