Control/Dimming IPS Dimmer failures

My theatre has Entertainment Technology Intelligent Power System IGBT dimmers that are now about 11 years old. We are having failures that appear to be heat related on dimmers that are nowhere near capacity. For example, a single 2.4kw module with 1 575w fixture on each circuit running at 75%, will suddenly shut off after being on for 15 or 20 minutes. They will work again later in the same show. In some instances rather than failure the dimmer will "blink" occasionally. These problems are becoming quite pervasive. I would certainly not classify the system as heavily used, just average, steady academic use, and rarely are the dimmers loaded to anywhere near full capacity.

Anyone else out there have this experience? What is our best approach to dealing with this (short of replacing the system)?

I have to wait for an opportunity to do more scientific testing, but based on prior tests we found several modules failing with loads over 2.1k, so for the last year we have de-rated them to 2k, but now the issue seems to be worsening.
 
When's the last time that they were cleaned?
 
There really isn't much to clean. They are solid state sealed in a raceway that use cooling fins rather than fans and filters. On the modules I replaced last spring there was not much dust inside when I pulled them out. I am currently cross-referencing the ones that failed this week to see if any of them have been replaced before, in which case maybe there is a power issue.

I also forgot to mention that these are in a fixed grid, not rigged to fly (not that that should make any difference other than perhaps some additional abuse).

Here is the link to the current version. Intelligent Raceway - Philips Lighting Controls Entertainment Technology was acquired by Philips.
 
Your description certainly hints of a thermal problem, which could have external causes. What is the ambient temperature like for the dimmer strips? Is there good air flow around them? Are there a bunch of hot lights underneath them?
 
This problem is exactly why I do not believe in dimmer strips. Take something with electronics that generate a lot of heat that needs to stay cool and then put it in the hottest part of the theater and hope it works. I would take some compressed air and blow it on the heat sinks I bet there is a bit of magical theater dust on them.
 
Tim, While YOU may not believe in Dimmer strips, the Industry does. They are a tested, approved and viable technology. If I had to guess I'd say 75% of the issues one runs into with Dimmer strips is the result of abuse or misuse or poor maintenance. I reallt don't think anyone who brings a product like, oh, the ground-breaking IPS Strip, Throws it together and "Hopes it works". If you've never seen one the heat sinks don't tend to collect a lot of dust simply because of the way they are designed. ( See someone was thinking ahead) The fins are wide spaced so as to NOT catch dust.

LBC, Are the strips oriented correctly, Depending on the version there will or will not be a 'shelf' on one side of the cooling fins. The Shelf needs to be on the bottom, closest to the light. It is there to prevent hot air from rising through the fins and heating rather than cooling the dims. The Original version of the IPS strip didn't have the shelf and the bottoms of the heat sinks were expose to the heat rising from the instruments below. If this is the case with yours, you could help yourself a lot with making a shelf out of blackwrap or the like. you might even look into creating something more permanent that fixed to the Uni-strut channel and blocked the bad airflow.

If it's the same few units having this behaviour over and over, then it's possible this is a DMX issue < not likely, but possible> dble check the DMX lines running to those dims, Make sure someone didn't turn on the terminator on the dimmer before them in a daisy chain. It could also be a "Head-end" issue. You can , after unplugging the unit, remove the four screws holding the head end inplace and check for loose or heat damaged wires. Sometimes the molex plug that carries data to the individual dims get's heat damage or corroded simply needs reseating. < Specially in our lovely NW Climate>. One last issue is at the IBGT itself You can check the thermo-couple heat sensor. See if it has come loose from the IBGT. Occasionally they do in which case it might be picking up erronious heat signals from other components. If it is you can simply re- super glue it Or find an appropriate replacement part then re glue that in place. Or you can take the dim end out and ship it to Texas for Diagnostics.

Hope something in there helps. If not yell back at me, Our house happens to be a test bed sort of, it was like an IPS showroom for Rosco/ET.
 
Last edited:
Removed snarky comment replaced it with more adult chiding.
Who says we hafta be adults around here? :mad:

Now, respectfully Van, I have no dog in this fight, and don't wish to create an argument. But, it seems to me, your suggestion to use blackwrap to solve a thermal issue reeks of "Throw it together and 'Hope it works'".

... While YOU may not believe in Dimmer strips, the Industry does. ...
I would counter that "the Industry" has yet to adopt either distributed dimming OR the IGBT in any widespread basis. The equine was thrashed profusely four years ago in the thread http://www.controlbooth.com/forums/lighting-electrics/9344-sensor-vs-et-intelligent-raceway.html , so let's not go there again. As best I can recall, Van and DaveySimps are our only two other members with IPS installs in any large quantity. Perhaps there are more, but they're too embarrassed/ashamed to admit it. /non-adult snark

... Are there a bunch of hot lights underneath them?
Gee, I don't see how a manufacturer could ever foresee a situation where a bunch of hot lights would be hung directly below a dimming connector strip. /non-adult snark really OFF this time

LBCSeattle, I'd see what Texas has to say. I believe I heard something about ET only concentrating on architectural dimming/controls, [EDIT: See this post.] but there must be someone at Philips/ET/Strand who can help with IPS issues. Maybe it's heat-related; maybe the components are at the end of their useful life, who knows?

.
 
Last edited:
Having rechecked the dates, the system here was commissioned in 2004, so is only 8 years old.

Van, there is no "shelf" as suggested above, but that seems a pretty unlikely issue anyway since the electronics are contained inside the raceway. The free air outside the cooling fins would have to be hotter than the inside of the module. The fins are there because of the heat generated in the dimmer. The second law of thermodynamics would prevent the cooler free air making its way through the cooling fins and adding heat to the internal electronics. I would think adding a blockage would restrict airflow through the fins, making them less efficient at cooling. There doesn't appear to be a shelf indicated on the cooling fins of the current version either.

Given the nature of the failure, an issue with the sensing is more where I am suspicious. Since it is becoming more pervasive I suspect also suspect that to be a long-term reliability issue, which sucks for a system less than 10 years old. We will be sending in modules for testing/repair, but last time we did that ET was in the throws of being migrated to Philips and when we asked for more followup we were essentially abandoned with no contact for over a year. We are tackling it again now with Philips firmly in the driver's seat. There is very little evidence of the existence of ET as a company left on the Web. They were assimilated into the Philips Borg.
 
I too have the IBGT dimmer. Ours were commissioned the same year as yours, when they were first to the market. We had many issues with this when we first commissioned our system. A rep from ET was in our facility for months on a regular basis. They finally swapped the RCM for each of our electrics, and it really seemed to help. Did they ever do this to your system? Willow Creek Church, another church that was a big install like mine, had similar issues and it fixed some of their issues too I am told. Have you used the Raceway Control Module to see if there is an error listed when you have this problem? Often times, we got this issue and we would power cycle the RCM and it would clear the issue upon reboot. FYI, this works for many issues.

Unfortunately there is very little service available for the units outside of shipping them back to Philips now that ET is no longer manufactured. You cannot buy parts yourself either. Lord know I have tried. My last batch of dimmers I sent in took over 6 months to get back. The pair before was 4 months, so be prepared for that if you need to send anything in. they literally only have 1 tech that services them, I was told. And if you have to have them sent into their other plant for QC, it will take another month at least. Lucky I did not have to go through that.

Feel free to PM me if you have any specific questions I might be able to help you with. I will also make a point to check back here for additional posts.

~Dave
 
Thankfully I have not had many intermittent problems with our Rosco/ET IPS dimmers. When ours fail, they stay failed! We have both 3x20A and 3x60A units. Only once have I seen a problem with what is controlling the IGBTs. The one time was in a 60A unit with a bad transistor and resistor that carried the signal to the gate of one of the IGBTs.

Almost all of the problems that I have encountered are from one or both of the pair of IGBT modules failing. A failed IGBT is either stuck on, or off. The one that wasn't getting told to turn on or off quickly enough, it just blew up! In my experience they tend to stick on. This tends to show itself as about 30% output with a DMX signal of 0. It then either dims up from that ~30% up to full just fine, or flashes the output as the DMX increases, but will then will be fine at Full. Depending on the age of your dimmer, the repair may require soldering to replace the IGBT, or it can be done with just screws on the older style with miniBLOC IGBTs. There is information on the web, along with some good youtube videos that show how to test an IGBT. I use the 12v battery and lamp method. It is easy to take the miniBLOC out of the circuit to test. The soldered legs are more work to test out of circuit. I have found you can cheat and find which one is stuck on while it is still connected to the dimmer board. You just can't get a good one to turn on.
Either style requires knowledge and caution to change them out, as a static electricity shock can fry a good IGBT. I can write some more about it if anyone is interested. These repairs require experience dealing with electronics to complete safely and successfully.

Most challenging for me was finding parts since the IGBTs for the 20A models were well out of production. I was able to find current production replacements that have worked for us.
The older dimmers had miniBLOC IGBT (IXYS IXSN50N60U1) screwed onto the board, which I have replaced with IXYS IXXN100N60B3H1 components. You will also have to smear on new thermal pastebetween it and the heatsink.
The newer dimmers had IGBTs in a TO-264 package, with legs that are soldered directly onto the board were Toshiba - GT50J301, which I have replaced with IXYS IXGK-GX64N60B3D1 components. The thermal pad on these always rips on me when removing the component, so I have replaced it with Laird TGARD K52-2. I didn't find any precut to the size of the IGBT, so I got a sheet and cut it down. This provides thermal conductivity to the heatsink and also electrically isolates it from the heat sink, as the back of the chip is also the collector!

Thanks for bearing with me, I am a stagehand, not an electronic engineer! Best, tim
 
If you need some spares, I have four of the ET IPS dim bars that you can have. They worked great for us but the room has been updated. contact me directly if you're interested, they're in Waco, TX.
 
I too have the IBGT dimmer. Ours were commissioned the same year as yours, when they were first to the market. We had many issues with this when we first commissioned our system. A rep from ET was in our facility for months on a regular basis. They finally swapped the RCM for each of our electrics, and it really seemed to help. Did they ever do this to your system? Willow Creek Church, another church that was a big install like mine, had similar issues and it fixed some of their issues too I am told. Have you used the Raceway Control Module to see if there is an error listed when you have this problem? Often times, we got this issue and we would power cycle the RCM and it would clear the issue upon reboot. FYI, this works for many issues.

Unfortunately there is very little service available for the units outside of shipping them back to Philips now that ET is no longer manufactured. You cannot buy parts yourself either. Lord know I have tried. My last batch of dimmers I sent in took over 6 months to get back. The pair before was 4 months, so be prepared for that if you need to send anything in. they literally only have 1 tech that services them, I was told. And if you have to have them sent into their other plant for QC, it will take another month at least. Lucky I did not have to go through that.

Feel free to PM me if you have any specific questions I might be able to help you with. I will also make a point to check back here for additional posts.

~Dave
@DaveySimps "I too have the IBGT dimmer." OR IGBT, Isolated Gate Bipolar Transistor?
Toodleoo!
Ron Hebbard.
 
It was in theory a good idea.

Too bad it never lived up to its potential and I've yet to hear from anybody who used this system who didn't have major issues with it. Too bad

If it were me, I'd be tossing it all, adapt the power distro to something that would allow me to install Source 4 Dimmers. Or Leprecon Shoe box Dimmers.
 
Thankfully I have not had many intermittent problems with our Rosco/ET IPS dimmers. When ours fail, they stay failed! We have both 3x20A and 3x60A units. Only once have I seen a problem with what is controlling the IGBTs. The one time was in a 60A unit with a bad transistor and resistor that carried the signal to the gate of one of the IGBTs.

Almost all of the problems that I have encountered are from one or both of the pair of IGBT modules failing. A failed IGBT is either stuck on, or off. The one that wasn't getting told to turn on or off quickly enough, it just blew up! In my experience they tend to stick on. This tends to show itself as about 30% output with a DMX signal of 0. It then either dims up from that ~30% up to full just fine, or flashes the output as the DMX increases, but will then will be fine at Full. Depending on the age of your dimmer, the repair may require soldering to replace the IGBT, or it can be done with just screws on the older style with miniBLOC IGBTs. There is information on the web, along with some good youtube videos that show how to test an IGBT. I use the 12v battery and lamp method. It is easy to take the miniBLOC out of the circuit to test. The soldered legs are more work to test out of circuit. I have found you can cheat and find which one is stuck on while it is still connected to the dimmer board. You just can't get a good one to turn on.
Either style requires knowledge and caution to change them out, as a static electricity shock can fry a good IGBT. I can write some more about it if anyone is interested. These repairs require experience dealing with electronics to complete safely and successfully.

Most challenging for me was finding parts since the IGBTs for the 20A models were well out of production. I was able to find current production replacements that have worked for us.
The older dimmers had miniBLOC IGBT (IXYS IXSN50N60U1) screwed onto the board, which I have replaced with IXYS IXXN100N60B3H1 components. You will also have to smear on new thermal pastebetween it and the heatsink.
The newer dimmers had IGBTs in a TO-264 package, with legs that are soldered directly onto the board were Toshiba - GT50J301, which I have replaced with IXYS IXGK-GX64N60B3D1 components. The thermal pad on these always rips on me when removing the component, so I have replaced it with Laird TGARD K52-2. I didn't find any precut to the size of the IGBT, so I got a sheet and cut it down. This provides thermal conductivity to the heatsink and also electrically isolates it from the heat sink, as the back of the chip is also the collector!

Thanks for bearing with me, I am a stagehand, not an electronic engineer! Best, tim

Thanks for your post. I am having similar issues with a bunch of IPS-DS-2403's 20A dimmer modules, as the ones you describe. I have four of the screw on IGBT style modules stuck on, and one of those strobes, with increasing intensity, from about 5% to full. For this one, it looks like it's stuck on at around 50% then flashes up from there at increased intensity following the increase in the commanded DMX value. I also have one additional unit that does not turn on at all. I have identified, and replaced, bad IGBT's on a number of these modules, and this has resolved the issue for most, but on these remaining units, the Gate leg on the IGBT's seem incorrect, so I am looking further up the stream for another cause. I have tried swapping the CMOS microcontroller chip on all units, and while this worked in one case, on these remaining units this has been unsuccessful at resolving the issue. You mention that you found a bad transistor and resistor on a 60A module. Any thoughts as to where to look next on the 20A variety? There are many different transistors, and resistors, as well as capacitors, connected, seemingly in parallel, and in series to the gate of these IGBT's, and its very difficult to sort out where the issue might be. It's also extremely difficult to trace out the signal path of each small component on this tightly packed multi-layered circuit board. Any insight here would be greatly appreciated. Thanks!
 
I don't understand your reply. Yes... I know what IGBT stands for... Are you trying to cryptically hint at something that I should be looking at?
I think he was correcting a prior post that called them "Isolated Gate Bipolar Transistor"
 

Users who are viewing this thread

Back