Pinout for this ancient xenon arc lamp power supply?
September 28, 2011 1:33 PM   Subscribe

Recently I obtained (for work-related use) an older arc lamp power supply. It's a Zeiss unit brand-wise but was manufactured by Ludl Electronic Products (LEP) on Zeiss' behalf. It may well turn out that this particular supply isn't suited to our needs (based on the lamp we want to use it with), but I would still like to find some way to determine the pinout of the output connector. Unfortunately, I've had a heck of a time finding ANY schematics for 75 W DC xenon short arc lamp power supplies, and despite my EE degree, I remain baffled following *mumble* weeks worth of ohming-out and circuit-tracing. And at this point I've been staring at the darned thing so long that it's become hard to tell if I'm missing something really obvious or whether it's just futile to expect to be able to determine what pin serves what purpose without a schematic. Long, convoluted explanation beyond the "more inside" button.

I should note that I have already contacted both Zeiss and Ludl regarding obtaining documentation for this supply (for which the part number is "91 01 44"). Zeiss couldn't provide any information at all other than the suggestion to contact Ludl, as they actually built the thing. Ludl, in turn, sent me a connection diagram for hooking up a Nikon arc lamp to one of their supplies, but the only useful thing that provided was confirmation that I had correctly identified the output connector on the supply (and the lamp we want to use is a Zeiss, so I certainly can't assume the same pinout applies).

The main bugbear in all this is the fact that the lamp module is a tad more sophisticated than the power supply. It appears to have its own built-in ignition/ballast circuitry, and the connecting cable (for the lamp - power supply interface) only has three conductors in it: +DC, -DC, and chassis ground. The power supply, meanwhile, has 7 pins in its output connector.

I have the proper mating connector to adapt the lamp cable so that's not an issue, but it's obviously useless if I don't know which pin goes in which hole.

The power supply, I'm guessing, has 7 pins to account for (a) safety interlock circuity (which suggests at least two pins are going to need to be shorted so that a connection is made when the lamp cable is plugged in), and (b) ignition/trigger circuitry, in addition to "regular" DC for the lamp once an arc is established. I am hoping that maybe I can just ignore the power supply's ignition connections since I think our lamp has its own stuff in that department, but I am not going to *assume* that's even possible without more information into the PS circuitry.

If it helps, here are some photographs I took of the power supply:

Back of supply (shows output connector):

Front of supply (with chassis lid removed):

Closeup of internal circuitry:

Closeup of wires from inside of PS to output connector:

And here is the lamp I am trying to hook up:

Housing exterior:

Lamp circuitry (partial/bottom view, with cover removed):

...but in any event, as for my actual question:

(a) Since I am fairly certain that the power supply won't output normally unless the interlock condition is satisfied, how might I identify which pins need to be shorted at the connector toward this end?

(b) Since the PS probably won't output normally unless the interlock condition is met, how can I conclusively identify which pins represent the power supply's +/- outputs (without being able to measure them with a multimeter)?

(c) What, if any, other useful tests might I perform (with the supply either powered or unpowered) toward the end of identifying pin functionality?

(Thus far I've only turned on the supply a few times, each time for less than a minute. This led to the interesting discovery that I end up with an approx. 30 VDC reading between the pin connected to one side of a large filter capacitor and *either* of two pins at the connector that each go to one side of a 1 ohm power resistor.)

(d) Finally, is this just a completely ridiculous quest in the first place?

I AM actively looking for a different power supply that will be more readily compatible with our lamp (Zeiss 44-80-12) and have found some candidates, but given the headaches I've been through already with a power supply that looked reasonably good "on paper" I want to make darned sure I KNOW the thing will work before I order another one. And when you're dealing with old/used/piecewise equipment that the manufacturers don't even support anymore, this is NOT necessarily a straightforward task!

Also, I know that technically it's best to use ONLY the particular power supply that a manufacturer has specced for a particular lamp, but Zeiss said the lamp was too old for them to even be able to recommend a matching supply (they don't even support it anymore) and I am unwilling to consign it to the landfill without at least *trying* to see if there's something out there to power it.

So yeah. Any insights from anyone experienced with xenon arc lamps and the power supplies that go with them would be appreciated. Thanks!
posted by aecorwin to Technology (8 answers total)
From the department of bad ideas and destructive testing:

1) Wear safety glasses and gloves. 2) Take a xenon tube you can afford to lose. 3) Plug one end into previously identified high voltage hole, 4) Plug other end into each of the remaining six holes.

See what happens.

You may wish to put the xenon tube in a clear case at the end of long wires, rather than use the tube leads directly, for this.
posted by zippy at 3:20 PM on September 28, 2011

Best answer: The photos are not particularly helpful as too much is obscured. Can you reconstruct any kind of schematic from what you can see? You don't have to completely understand the schematic in order to break it into functional blocks based on what each block almost-certainly needs to do.

It's likely, for example, that the capacitor is at or near the high-voltage output stage, so if there are wires going directly to the connector from it you'd just need to identify + and -. The interlock is harder but you'd want to trace wires back from the output connector until you found something resembling "logic" (a relay, transistor, actual logic IC).

Note, by the way, that there are a few failure modes of high-voltage capacitors wherein the cap attains its full charge when it's not plugged in -- so be careful around that thing until you (a) determine it's not a HV cap or (b) find a shunt resistor that would bleed off charge when it's unplugged.
posted by range at 6:22 PM on September 28, 2011

Best answer: Seconding when dealing with high voltage, use your EE training. Only one hand on device, other hand in pocket. Getting a zap that runs down your side is bad, but better than one across your chest.
posted by zippy at 7:33 PM on September 28, 2011

Response by poster: zippy: heh, part of the trouble here is that we *don't* have any "sacrificial" xenon bulbs. Those bulbs go for around $200 apiece, and you're supposed to stop using them anyway when they're nearing the end of their useful life (400 hours or thereabouts) as they can actually become dangerous to operate at that point.

range: yeah, I mostly just put the photos up as a way to communicate the general look/age of the supply. I've been working on a schematic but right now it's mainly pencil scribbles on paper (and wouldn't likely be very useful to anyone else even if it scanned well).

That aside, the capacitor I measured voltage across today is not actually a super high *voltage* cap -- it's the big blue can-shaped one in the pictures, and IIRC it's only 65V max. The capacity value IS pretty large, though, and it takes forever to discharge even when the supply is off, so I am fairly sure it is there for filtering/smoothing purposes rather than HV pulse generation. Also, I've at least determined that the blue wire coming off one side of that cap is connected to the primary DC rectifier's negative side (the pink wire coming off the other side goes to the positive side). The blue wire (negative rectified DC, if I am correct) also goes directly to the output connector, but the pink wire does not, which is interesting -- the pink wire goes through a bunch of other circuitry, mainly large power resistors and large switching transistors, before connecting with the white wire that does go to the connector.

I actually think the ignition pulse is created by the second of two transformers in the unit (which looks to be configured like a flyback driver or similar). So yeah. I've learned a lot by tracing out the circuit but not quite enough just yet. Thanks for the tips, though, and definitely I will keep in mind the "one hand in pocket" philosophy. If I *do* get the igniter circuitry working supposedly that can generate pulses in the 25 kV range, which isn't trivial!
posted by aecorwin at 10:10 PM on September 28, 2011

Response by poster: Oh and as far as functional blocks go -- thus far I've figured out the following:

AC input (from wall) --> stepdown/isolation transformer --> DC rectification --> current control/ignition/output control circuitry.

There are two PCBs in the unit: one looks to be mainly current control (there are 2 potentiometers, probably one fine adjust and one coarse adjust) and the other looks like it's controlling whether the flyback transformer is "allowed" to output or not (so I'm guessing there's circuitry there relevant to both ignition and safety interlock). FWIW.
posted by aecorwin at 10:13 PM on September 28, 2011

I'm curious about the circuitry in the lamp housing, too- seems like you might have some redundant parts there. I have a couple of Zeiss short-arc mercury lamps- they are much simpler, two-wire connections.
Howsomever, as an alternative, you might spend thirty bucks on an automotive HID ballast and see if it'll ignite your bulb. Automotive xenon HIDs are typically 35w, I believe, and that might be a good source for replacement bulbs, too.
posted by drhydro at 11:05 PM on September 28, 2011

Here is some information about somewhat larger gas discharge power supplies that might help you understand the basic circuit design of what you're looking at. Look at the section on testing with a dummy load and variac for some ideas on how to test safely.
posted by doctord at 7:05 AM on September 29, 2011

Response by poster: woot, thanks for the info, everyone!

Also, I went ahead and uploaded an image of my sketchy attempt at a schematic. It's FAR from perfect and some blocks currently lack detail but it's a start considering I have NO real documentation on this thing.

Schematic attempt, via flickr
posted by aecorwin at 2:58 PM on September 29, 2011

« Older Two Licenses, One Cup   |   I can't hear you over how awesome XBMC is (I wish) Newer »
This thread is closed to new comments.