Sometimes my job can become quite interesting, due to either crappy quality control from an equipment manufacturer or some really bizarre karma. The job I'm currently wrapping up is a good case in point.
First, bear with me as I give some background on what a "Xenon Power Supply" is. Modern cinema projectors use a Xenon Short-Arc lamp as a light source. Wattages of the lamp range from 500 - 7,000 watts (up to 10,000 watts for "IMAX"). The average cinema uses either 2k or 4k lamps. The lamps operate as a constant-current arc, at low DC voltages. A typical case such as this one is 2,000 watts, 70-85 amps and roughly 24-35 volts DC.
Early designs for these power supplies were based on a transformer/rectifier combination, not unlike a typical arc welder/forklift battery charger. (In fact, the very first supplies for cinema use were modified forklift supplies made by Christie Electric.) The advantages are simplicity and ease of repair (most common failure being the diodes), but have the disadvantages of weight, noise, and heat generation.
Modern supplies, "switchers", like the ones in this caper, are built like computer supplies on steroids, using switching PWM technology and much smaller (and lighter) transformers. Light weight, fairly efficient, but not repairable in the field.
That brings us to the caper at hand:
The room is equipped with two projectors using identical lamps. Part of the job was to provide a cutoff when the fire alarm was activated. Easy enough, I just had relays interrupt the 120vac control voltage to each power supply. I had replaced the old supplies as one was going bad (early generation switcher) and the other was the old magnetic style without a housing!! (Leaving exposed terminals at the supply feed of 208v 3 phase.)
Both of the new supplies worked right off the bat, and my tests of the fire alarm cutoffs went well. I also test ran some film for about a total of 3 hours running time on both units.
The first hint of trouble was this past Tuesday. The fire alarm folks had scheduled the final test for 6am. I go in with my helper, we fire up both lamps and get ready for the test. About ten minutes later, the lamp on #1 projector goes out. Huh. I try to relight it, no dice. The control lines were still hot, and the breaker for the supply was still on. I crawl under the floor to the power supply and it has a fault lamp lit, indicating a bad module (the main power module.) Oops, defective supply. No big deal, we only needed to have one to prove the fire cutoff works. So we do the test with the #2 machine.
It shut down, just as advertised. But it wouldn't relight after the alarm was reset!! Geez!!! Out comes the Fluke..good control voltage to the supply but it wouldn't turn back on!! Now I'm getting a bit torqued. So we kill the breakers to it, I recheck everything, no dice. To heck with it, I gotta go back to change out #1 anyway.
On a whim, about an hour later, just before we left, we tried #2 again. It fired right up as if nothing was wrong!!
Fast forward to today. We go back and replace the #1 supply with a new one. Fire it off, it works!! Yea! Let's let it cook for an hour and call it a day. At the hour, I tell my helper to shut the lamp off, to let it cool for ten minutes and then we'll turn it back on. He tells me it won't shut off!! Sure enough, the only way to kill the lamp is to shut off the breaker!!
Out comes the Fluke again. Lamp control leads read as .02 volts OFF, 119.6 volts ON. All normal there. But the power supply's "GO" light stays on anyways. Disconnected the control leads, same result. Power supply says "GO" and lamp lights.
So now the factory (record so far 0-for-2) gets to send me another power supply.
Did I mention that these power supplies cost $3,800 EACH?
Did I also mention that they sent the first replacement Next Day Air? (52lbs and oversize.) From Omaha, Nebraska to Southern California?
Just glad I'm not paying the bill for all that freight.
The saga will continue.........
edited smiley code
[This message has been edited by mxslick (edited 11-17-2006).]
Yeah, the Avalon is a great place. I should be going back there by Xmas for their P.M. checkup.
Generators such as the one in the Avalon would actually be better for the life of the Xenon bulbs, as they have extremely low ripple current. (Xenon bulb electrodes erode rapidly when exposed to ANY type of A.C. or ripple in the D.C. input voltage.)
The only real reason one cannot use a generator for Xenons? (At least for practicality sake.) Generators for carbon arcs do not have a high enough open circuit voltage. (Most only manage around 2 times the running voltage, or about 75-90 volts open circuit.)
Xenon arcs require an open circuit voltage of at least 125-175 volts D.C. and are struck using a high-voltage, high-frequency pulse ignitor. Typical ignitor output pulse is 35-65kV, somewhere in the R.F. band in frequency. The ignitor's H.V. output transformer is driven by a spark gap R/C circuit at about 400 volts A.C.
And I can tell you that the pulse from an ignitor is extremely painful.
Stay tuned for the outcome of the caper....
Stupid should be painful.
Re: The great Xenon power supply caper....#72052 11/19/0603:50 PM11/19/0603:50 PM
It's hard to say without any schematics. Is it possible that your FACP alarm contacts are bouncing the control power? How are they normally controlled? Snap action switch? Is the 120VAC from ctl transformers derived from the 208 or a separate circuit? Could there be a zero crossing circuit that is never seeing a zero current? I just dealt with little projectors shining into a film chain but I sure do miss the old days of "Wait Til Your Father Gets Home" and "Lancelot Link, Secret Chimp". Now if I can only stop tapping my foot to cue dots and pointing at the screen on reel changes. Joe
Re: The great Xenon power supply caper....#72054 11/21/0612:44 AM11/21/0612:44 AM
Wow I feel like a meathead. That old avalon theatre must be a great place to work. I posted in the nostalgia section, see if you can get a permanent job there. If for no other reason than to provide me with many many neat pictures to look at. I love old buildings. Yeah, I know, i'm wierd.
I showed this thread to a friend and he reminded me of a problem that he had had. We both worked on a PA system installation (massive stadium) that had many, many amplifiers. I think it was 21, but I can't remember and he wasn't sure. Due to the required output wattage needed, all of the amplifiers were large, 3 phase units. He kept having issues with controlling them, as they would just kinda work whenever they felt like it. I never really dealt much with it as my job was just hand soldering microphone wall plates.
I was finished with my part of the job long before he had figured out what was wrong, so I moved onto another project in a different building. 2 years later, he still remembers what the problem was. Each amplifier had a 30 amp 3 phase 208 volt mains connection. They each also had a 120 volt "control loop" which drew about 10 watts. being as though this is practically no current, he just kinda wired them wherever was convenient. It turns out, however, all control lines must be wired to the same phase. He ended up ordering a 10 amp QO breaker, and wired all the control lines to it through a contactor. the contactor was then hooked into the large, computer controlled system. Sure enough, this solved the problem and left many people scratching their heads. 2 years later, they have had no problems with the system. Go figure. I guess thats why we haven't used amplifiers from that manufacturer since then.
Re: The great Xenon power supply caper....#72055 11/21/0601:01 AM11/21/0601:01 AM
It turns out, however, all control lines must be wired to the same phase. He ended up ordering a 10 amp QO breaker, and wired all the control lines to it through a contactor.
Funny you should mention that. My next plan of attack was to be to add a relay with a 120vac coil, drive that from the control line from the lamp and then drive the power supply from one of its own phases. But the factory engineer swears that the control can be driven with 120vac from anywhere, it's supposedly completely isolated from the power supply's mains. And he agrees that there is a control board problem. Go figure.
Yeah, this thing is a stumper. To explain the hookup:
Each lamphouse is powered on a 20amp 120v circuit. This feeds power to the lamp's blower, high voltage ignitor and also the control circuit, after satisfying two safety interlocks. There is a double-pole switch on the lamphouse that kills power to the control circuit and ignitor when it is off, this is the switch the projectionist uses to control the lamps.
The control circuit sends the 120vac to the power supply "contactor" (in this "new and improved" version, it is a solid state circuit driving first an scr then a high-power triac array to heat up the supply.) The control circuit is designed to turn on with an input voltage range of min. 90 volts and max of 135 volts. See my first post for the measured voltages with the lamphouse switch "on" and "off".
The supply itself has a 208v, 3 phase input on a 30 amp 3pole breaker. This breaker in some installs is left "on" all the time.
So in summary, each lamphouse has its' own 120vac feed and each power supply has its' own 208 vac input. Edited to add: All these circuits are in the same panel.
The fire alarm cutoff is a simple set-up. Inside each lamphouse I have a 24vdc 4pdt relay. The normally closed contacts of this relay are in series with the 120vac hot control line. When the alarm is activated, this relay pulls in and interrupts the control circuit.
The 24vdc comes from the master projection control system (another major PITA, it is a homebrew from a former studio engineer and has lots of bugs.) The control system DOES NOT control the lamps.
The fire alarm system has dry contacts which are normally OPEN, and CLOSE and maintain in an ALARM condition. Resetting the alarm opens these contacts.
So the alarm system isn't part of the problem. If it were to chatter, the result would be that the xenon power supplies would shut down and lock out on their "multiple ignition failure" mode.
Bear in mind that even after I pulled my fire alarm cutoff relay, and even disconnected the control leads altogether, that blasted supply wouldn't shut off!!
***Update: I was back there today on another issue, and when tested the power supply worked properly for the first several rapid on-off cycles. But after it was left "ON" for about five minutes, it would fail to shut off again.***
The sad part of all this is that, despite the cost of this equipment, this manufacturer has always had serious quality control issues. The early generation switchers I mentioned above always had one of four "operation" modes:
It would work great, for about a few hundred hours or less, then quit; It would work great and run virtually forever; (I actually have one on my home projector) It would fail right out of the box; or It would work for a few minutes, then literally go down in flames.
Now if I can only stop tapping my foot to cue dots and pointing at the screen on reel changes.
I can totally relate!! I can't go to movies anymore, I'm so programmed to watch for the wretched cues even when I'm sitting in a full house of people. That, and if there are ANY technical flaws in the presentation, I go ballistic.
I once got a house of over three hundred people refunded due to the total ineptitude of the alleged "projectionist" at a very major and famous theatre in Hollywood.
But that's a whole 'nother story.
edited to respond to napervillesoundtech
and to add panel comment
[This message has been edited by mxslick (edited 11-21-2006).]
Stupid should be painful.
Re: The great Xenon power supply caper....#72056 11/21/0610:53 PM11/21/0610:53 PM
Those home brew specials can be kinda icky. This bay be too obvious, but, is there an emergency shutoff switch on the PSU anywhere? paralell wire it if there is. Sometimes, it isn't as obvious. Get an old PC with an ATX supply. Pin #14 is the power switch control line through the motherboard. short it to ground, and the PC turns on. Remove short, and it turns off. There may be something like that. I dunno. Do you have any website links for the Arc PSU's? It could be neat to look at.
Re: The great Xenon power supply caper....#72057 11/23/0602:41 AM11/23/0602:41 AM
Nope, these things don't connect like any type of PC switching supply. The available connections are:
Nema L15-30 3 phase, 4 wire flanged inlet, 30amps, 208/230 volt (250 volt plug, Hubble P/N HBL2723)
2 pole screw terminal block, accomodates up to #2 wire
4-position screw terminal barrier strip, 10/32 screws, two for supplying 120vac from an internal transformer for that Manuf. own lamphouses, and the other two are the 120vac control return.
And before anyone brings it up, yes, in most cases the lamphouse itself would be getting the power from the PSU's internal transformer, BUT it is allowed by the manufacturer that other brands of lamphouse can be connected as I have done here. In fact, the previous swithcers have been connected exactly as I have done, but were being replaced due to the one being improperly changed out and it's mate (which was still operational but unstable) being changed to keep the equipment matched. In fact, I have used these supplies with a variety of lamphouses and connecting them in this same manner with no prior problems.
They do have a fairly complex microprocessor based control board to regulate output current and monitor the status of the unit. And there are several DIP switch options to enable/disable certain protection modes as well.
As for website links, no luck with this particular maker. They don't even supply (pun intended) tech data to thier own dealers and techs!! They act as if these supplies are a matter of national security.
However, anyone who wants a look at the beast can send me a PM and I'll send back some pics.