Film-Tech Forum ARCHIVE: Diode Failure on Christie

This is topic Diode Failure on Christie in forum Film Handlers' Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=1;t=000745

Posted by Joshua Waaland (Member # 287) on 03-13-2000, 06:54 PM:

I was working Saturday and halfway through Mission To Mars the bulb shut off. We couldn't get the breaker to stay on and the bulb had a lot of hours so we decided to try to change it. No such luck. We couldn't find anything obviously burnt so we called a tech and he said it was probably a diode or the igniter. It ended up being one of the diodes.

I was fairly surprised to see that there wasn't much to the rectifier. When I worked with some older lamp houses that were on bases they had the big brown floor mount rectifiers. I always thought that there was more to them. Does anyone know what I am talking about? Why are they so big? Do they also contain the transformer? Does the Christie recifier just consist of diodes with big heat dissipation plates?

Joshua Waaland

Posted by Gordon McLeod (Member # 33) on 03-13-2000, 08:00 PM:

All no switching style rectifiers consit of a combination of the following
step down transformer
diodes (bridge rectifier)
filter capacitor
filter choke
(in some cases a seperate boost supply circuit)
and a contactor
some use rotar switchs or shorting bars to change the taps on the transformer to select the output while others use either magnetic amplifier to saturate the core or actually move the core to control the output

Posted by Scott Hicks (Member # 376) on 03-13-2000, 09:15 PM:

Joshua,

In almost every case, when the breaker blows, a diode has shorted in a non-switching rectifier. We advise changing all 4 or six diodes at the same time, from a future reliability standpoint, but if you can determine the defective one, replacing just that one diode will get you back on screen. Changing diodes is fairly easy, I've seen many a projection booth stocked with a couple spares. Of course, if you're close to a local tech, it's better to have an expert do the dirty deed.

Posted by Zach Zagar (Member # 379) on 03-13-2000, 09:30 PM:

Hi, sorry, I'm new here, but I have worked on Christie's for about 3 years now (no, not a master, but can do some amount of tricky stuff). Anyways, yes, diodes aren't very hard on the Christie lamphouses. You could change one at a time if you have an OHMS meter, but its much quicker to change them all and not to mention, yes, keep it uniform, may as well change em all if you're in there.

As for the rectifier being smaller, well, aside from the technical aspects pointed out earlier, there has also been some degree of miniaturization in EVERYTHING (not just projection booth), over the years, which could account for a change in the size.

Ok, enough rambling from a newbie.

Zach

Posted by Randy Stankey (Member # 64) on 03-14-2000, 10:32 PM:

Joshua:

I use CFS lamphouses @ TT-17. I've got LOTS of experience changing diodes. CFS lamphouses **EAT** diodes for lunch!!! After a while, you get pretty good.

Honestly, it isn't really hard. For the CFS, I can name all the tools you need right off the top of my head:

1 - 9/16 socket wrench
1 - 7/16 mini-socket wrench
1 - 7/16 open end (dog bone) wrench
1 - flat/ regular size screwdriver
3 - 150K120A diodes (RCA / Mouser preferred)
3 - 150KR120A diodes
1 tube of heat-sink grease
some alcohol and a wiping cloth
an ohmmeter/DMM

I don't say all that to show off. I say it to show that it isn't all that hard, once you learn. After the first 9,000,000 times we just say, "Oh Sh**!, Where's the tools?"

We had this old broken-down make-up table that we stripped for replacemant parts so we turned it into a "Crash Cart". We just put all the tools and diodes, etc inside it and leave it in a "central location". On a good day we can fix a blown diode in 20 minutes or less!

Posted by Gordon McLeod (Member # 33) on 03-15-2000, 10:37 AM:

Try using higher voltage rated diodes
I always had them failing in the XPS-45 rectifiers whether they were CFS ORC or LP's
I think some highvoltage from the ignitor finds its way back
I also solder a .1uf 10,000v ceramic disk cap accros each one and the replacements have become few and far between

Posted by Randy Stankey (Member # 64) on 03-15-2000, 12:01 PM:

We use RCA diodes, now from Mouser. Our chief Tech recommended this. He also told me what the ratings/markings mean but I don't think what he told me is right. I didn't want to argue with him so I just let it drop.

Here's what I understand:

150K?120A = 150,000v Peak Inverse Voltage (PIV) and 120A max. forward current (I-max) if "?" = "R" then it's a reverse polarity diode.

Our tech says it's the other way around.

Since diodes are supposed to be rated 2.5 times the I-max, that means we should use 150k?150A at minimum.

Which one of us is crazy??

Posted by John Pytlak (Member # 331) on 03-15-2000, 12:03 PM:

I agree with Gordon. Using higher PIV rated diodes is cheap insurance, as random voltage spikes can easily exceed several hundred volts. The 0.1uf 10kv cap helps suppress any spikes too.

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Professional Motion Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com

Posted by John Pytlak (Member # 331) on 03-15-2000, 12:27 PM:

Randy:

A diode with a 150,000 volt rating would be quite an accomplishment! The International Rectifier 150K100A rectifier has a forward current rating of 150 amperes at 150 degrees Celsius maximum, and a Vrrm of 1000 volts. So by inference, the 150K120A would have a maximum forward current of 150 amperes and a Vrrm of 1200 volts.

If your rectifiers "eat" diodes, are you sure you have enought air flow over the heat sinks?

I knew that BSEE degree would come in handy some day!

Posted by Randy Stankey (Member # 64) on 03-15-2000, 03:01 PM:

Okay, now I get it.

Still, everybody I've talked to says that CFS are well known diode eaters.

We had one of our guys take the schematics in to his tech school and his prof. said that this design shouldn't do that. Since we started buying better diodes it's happened a lot less but it still happens occasionally.

There are blower fans down in the power supply that are SUPPOSED to keep the rectifiers cool. From my experience it has limited success. When the fans stop working, you've probably got a week or two to replace/fix it before diodes start blowing. (That happens a lot, too!)

I guess the moral of the story is, "If you use CFS you'll have to change a lot of diodes!"

Posted by Gordon McLeod (Member # 33) on 03-15-2000, 03:22 PM:

Get better blowers Grainger has good dayton ones
Since it is in a console is it possible you don't have enough exhaust suction to bring the internal temperature of the console down

Posted by Randy Stankey (Member # 64) on 03-15-2000, 03:43 PM:

They are Dayton blowers. We get the replacements from Grainger. In fact, I can remember the model number right off the top of my head. It's 4C446.

I think you're right about the lack of air flow. A few of the times we had fans blow, I just take the side panels off and go downstairs and get 3 Tensa-Barriers to keep people the hell away from the proj. Just having the sides off seems to make a difference.

Part of the problem with the blowers dying is that they are down inside the power supply in an almost completely inaccessible place. These fans (as well as many other types) have to be cleaned and oiled a couple of times a year. The way they are mounted, that's impossible unless you remove them and re-install them. NO WAY, am I doing that! It's tough enough to replace them when they blow!

Posted by Gordon McLeod (Member # 33) on 03-15-2000, 05:11 PM:

A few years ago I got pissed of with a fan that was in a difficult place to service so I got some vacume cleaner hose the size of the outlet nozzel on the fan and mounted the fan in a easy place to get at and ran the hose down to the unit
If you do this you will have to increase the cfm of the fan as the ribbed hose will increase the static pressure
It is interesting as I haven't really had much of a problem with the Dayton blowers failing in the ORC/LP XPS45's and they are the sme design as the CFS

Posted by John Eickhof (Member # 323) on 03-15-2000, 05:47 PM:

I am surprise to hear of the diode troubles
in the CFS consoles. Are these the 2 KW or the 4KW? I have found that when neede, I install CEHCO 85A60 types, 85amp, 600piv
I've had no returns on these!
(The 85 is the largest available in a 1/4"
stud.)
For the larger ones, the 150A100 diodes seem to hold up good. I think you have a serious
air-flow or cooling problem! Heat will destroy diodes of any size! The 85s work for years in the christies!

Posted by Patrick S. Chaffins (Member # 348) on 03-15-2000, 08:42 PM:

Power supply cooling is VERY important!!!! Most power diodes will start to fail at about 80°C (176°F). (That number comes from looking at data sheets for the last five hours today) That's the case temperature, not the temperature of the heat sink! There is always some heat transfer loss between the diode and the heat sink. If it feels like the heat of a hair dryer coming from the P.S. exhaust then there is not enough air flow. Putting in higher rated (Current) diodes will help somewhat, however it generally just puts the replacement timing farther apart, (they will eventually fail). From my experience in working with high current power supplies in the last 30 years there has never been excessive heat sinking or airflow incorporated into the design. (Just put in enough aluminum and fan to keep it below the fail temperature of the device.)
Moral of this story:
1. Keep the openings clear
2. Keep the blowers cleaned and oiled
3. Keep the heat sinks clean (all them fuzzies rob the heat transfer severely)
4. Make sure the diodes are bolted securely (especially after moving the Power Supply and during installs)

---------------

may your lights be bright,
your screens clean
your prints scratch free
and your sound warm and not fuzzy
Pat

PS is there anyone else out there from the Detroit area?

Posted by Randy Stankey (Member # 64) on 03-15-2000, 10:12 PM:

No, these babies get hot. That's the way they've always been. If you take the panel off the side and wave your hand around, (Careful not to touch anything live!), It's pretty warm. Diodes and all metal surfaces get too hot to touch. Right after a show is over, you could get a nasty burn. That's where you have to develop talent. How do you change HOT diodes inside a HOT lamphouse on a sold-out Saturday night on opening weekend?

1) open the panels first. (shut off breakers!)
2) go get your tools. let things cool off.
3) put shop rags on the hot metal parts. (dont' forget to take them out!)
4) learn to endure pain!

I tell everybody it'll take 1/2 hour. When it takes me 15 minutes they're all impressed at how good I am. I learned that from watching Scotty on that episode of Star Trek!

Posted by John Pytlak (Member # 331) on 03-16-2000, 06:13 AM:

Randy:

If the heat sinks get so hot that you burn your fingers when you have to touch them, you definitely have a cooling problem. Increase that air flow! Despite the difficulty in getting at them, you will need to maintain the fans on a more regular basis. You've noted the quick failure of the diodes when the fans fail.

Posted by Tim Reed (Member # 137) on 03-16-2000, 09:28 AM:

First things first when replacing diodes:

(1) USE HEAT SINK COMPOUND

All the cooling in the world won't help, if the diode can't get the heat conducted away from it in the first place.

Heat kills 'em for sure, but I have a hypothesis that the PIV gets 'em more than you might imagine, when the power supply is turned OFF. That's where Gordon's caps across the individual diodes help (that, and it also shunts the ignition voltage).

When the power is shut off, all that heavy current flow comes crashing to a halt, which induces a pretty good reverse voltage spike as the field collapses. This, of course, is why you can't shut down a rectifier by opening the DC line (like we used to do with motor-generator sets). The change is too great, and you'll pop the diodes. Shutting the primary power down at the AC input, as we do now, is better, but I believe it is not perfect. It is probably the source of a lot of diode failures.

They really need to build a power supply that shuts down in quick steps; gradually decreasing output over a second or two until the arc breaks, and the input voltage returns to zero.

Just a thought, from my meanderings...

Posted by Randy Stankey (Member # 64) on 03-17-2000, 03:08 PM:

I oughtta' buy stock in a silicone factory! (Yes we do use heat sink compound) I also use alcohol and a clean rag to get all the old stuff off the heat sink too.

As for the caps on the diodes, I like the idea but I'd NEVER get people to go along with that one! I have a hard enough time getting people to keep using FilmGuard. (It was too much work to rewind those pads! -- When they found out they didn't have to clean projectors every day, they changed their tune!)

Posted by Gordon McLeod (Member # 33) on 03-18-2000, 10:19 PM:

Just a possibility at the top of the two heat sinks there should be a cap bridging the two assemblies is one of the lead broken or burnt off
I just had a ORC console loose a diode and I found the cap of

Posted by Randy Stankey (Member # 64) on 03-20-2000, 02:08 PM:

I'll look for the cap. to see if it's missing.

BTW, they are mostly 4500W but there are some 2500W. We rarely have trouble with the 2500's

I'm pretty sure it's heat. The little blower down by the rect. never seemed like it was enough to do the job. In fact one person told me that the blowers were an afterthought by CFS. Until they added the blower, I'm told they REALLY went through diodes! Maybe cutting a hole in the side and sticking a blower in would help.

Posted by Gordon McLeod (Member # 33) on 03-20-2000, 04:45 PM:

The XP45 rectifier is "usable" from 2K-4500 watts
The higher the current occours the hotter the stacks will get
hence the initial question is there enough airflow and I think we have an awnser to that
also diodes are more transient voltage sensitive the hotter they get
Also the PIV part usually doesn't get exceeded as much as the DV/DT (change in voltage vs change in time) exceeds there hot rateing
The console should have a minimum of 500CFM of exhaust and of course that much air must also be able to get into it as well

Posted by Jason Burroughs (Member # 68) on 03-20-2000, 09:05 PM:

Randy,

On the 7000 Watt version of the CFS Console they did just that, they added an additional fan to blow on the backside of the heatsink, in addition to the one already mounted. I too have experienced the CFS consoles going through SOME diodes but never to the scale that Randy is reporting.

I have also seen the fans below the heatsinks on the new Christies fail, causing them to eat diodes as well, even roasting the connection wiring. I would have to agree with everyone here, biggest killer of diodes is heat.

Posted by Jon Bartow (Member # 241) on 03-20-2000, 10:36 PM:

Hey everyone,
this is only sort of related, but I had an interesting diode situation with a ORC console. The theater called me and said there were sparks coming from inside the console (near the rectifier) After telling them not to touch the thing and shut the power off I high-tailed it over there. Turns out the vibrations caused by the projector (Century SA) were causing just about every bolt in the console to work loose. Including the diodes. Four diodes were almost falling off. On those diodes the electricity was arcing from the heat sink through the washer and bolt and then onto the stud. Welded the stud, bolt, and washer together on all four. Intersting to get off.
I retighted most of the bolts on the whole console (and projector) and it has been fine ever since. (The motor mount bolts were loose causing the motor to vibrate, in turn vibrating everything else.)

Posted by Gordon McLeod (Member # 33) on 03-21-2000, 08:20 AM:

ORC's have always been notoriuous for loose connections. Some of there standalone rectifers for 1600watt had a plate that connected the top of several caps togather as a square bus well they used to vibrate loose and arc as did the connections on the breakers
It has always been our service policy if ORC (or the like) retighten every connection on every service
Some is due to vibration other loosenings can be due to thermal expansion and contraction of the bolt threads

Posted by Randy Stankey (Member # 64) on 03-21-2000, 12:40 PM:

Yes, tightening those connections is important. I always crank 'em down. It's also a good idea to check them all every time you change a lamp.

We've had two instances of the taps/jumpers smoking themselves because they came loose from repeated heating and cooling cycles. The whole terminal block had to be replaced!

The exhaust fans are good enough. All of our lamps are lasting to at least 1000 hours, now. We had to have the exhaust blowers cranked up right after we first opened. After that, no problem. I just think that down there in the bottom of the console there's too much metal blocking the air flow.