Dolby CP750 - level drop disease recommendations?

[Ronda Love]

I have several Dolby CP750 units with level drop disease. What would you do to fix the problem? Just replace them? What would you replace them with? Is there a repair/exchange program?

Thanks for any suggestions.

[Leo Enticknap]

I'm sorry, but unless you can find "used but OK' ones, you will need to replace them with another model (coincidentally, I'm currently on a field call doing exactly that). Dolby can no longer supply replacement motherboards: there was a fire at the factory in Japan that made a DAC chip used on the board, and production of them was not restarted afterwards, apparently.

Even if you can find "used but OK" ones, you'll just be kicking the can a few yards down the road.

Most of the replacements I've done have been either to a CP950 if a six channel analog input is not needed, or a Trinnov OV2 if it is. Annoyingly, neither is a 30-minute plug 'n play upgrade. Installing a 950 or a Trinnov in place of a 750 may require adapters for the input from your media block: if it has dual RJ45 outputs you can plug and play into the back of the new processor, but if it has a DB25 out, you'll need an adapter. The power input on the 950 is on the other side of the unit, so be prepared to snip zip ties and reroute the cables, etc. You will need to retune the room, because you can't import EQ settings from a 750. Both models have internal crossover capabilities, and the 950 has a built-in monitor speaker, too, thereby enabling you to do away with a separate monitor/crossover unit. However, in order actually to use that crossover, you have to prepare a Dolby Atmos Designer file, which involves taking room measurements and a few other steps.

The bottom line is that when a CP750 dies, there isn't a fix that avoids significant time and expense.

[Ronda Love]

Of course not! LoL. Thank you for the info.

[Marco Giustini]

Has anybody tried fixing those boards?

[Carsten Kurz]

I think Marcel did. But it's painful. The question is - would someone be able to track down specific components that cause the level drop, instead of replacing all caps. I can't imagine that all those caps are actually responsible. I know that a few people were able to get rid of the level drop at least temporarily by opening the case and cooling the board with additional fans.

[Marcel Birgelen]

Like Carsten indicated, I successfully recapped the motherboard of a CP750 a couple of years ago. Please notice that the problem here wasn't a level drop, the unit was "flat dead". While I at first suspected the PSU, the problem eventually was in the motherboard. I replaced all the "big" elecrolyte based caps and that fixed it. I measured all caps with an ESR meter, but found none to be out-of-spec, just some to be close to being out-of-spec. It was peak pandemic and I had some time to spare. I've shared the experience on this forum a few times. There were a few discussions here that the procedure also helped to get rid of the level drop.
In any case, the outcome isn't guaranteed and the amount of hours you're going to spend on it, is in most cases, not economically viable. I eventually ended up donating the processor to a struggling theater and it's still working fine (knock on wood).

Before Dolby declared the CP750 EOL, the only supported fix was to replace the motherboard. While there may still be new-in-box and refurbished motherboards out there, if you can afford it, I'd follow Leo's advice and look for a more recent replacement. While I've not yet installed one myself, I've heard great things about the Trinnov OV2 and while not a drop-in replacement, the CP950 may be a good match if the CP750 worked well for you. If you frequently require alternate inputs for alternate content or events, a Datasat AP25 may be a good choice, though Datasat hardware nowadays comes with notoriously inconsistent lead times. If you need a replacement NOW and you can't find one in stock, it's not really an option, I'm afraid.

Edit: Added some more background info about the nature of the fix.

[Dave Macaulay]

Somewhere I have information from Dolby about the caps that are apparently the problem. There's a LOT of them. I started sourcing replacements thinking it could be a service I could offer to make some $$$. First, the initial parts cost was pretty high as we would need to to purchase a large quantity of many to get any and lead times were long for some (several months). I wanted to try a recap to see if it actually fixed a CP750 so needing to buy 500 was not ideal. Then I looked into the rework tools needed. Not a fortune but not cheap.
I suppose it's possible but for me... just not worth the time and cash investment required. Plus not sure if such a service would be sellable since we're in Canada and experience has taught me that many USA people are not willing or able to do international shipping paperwork correctly.

[Devin McNamara]

I’d love to get a hold of one to fiddle with it. A lot of cinema equipment that I’ve gotten my hands on so far has been problematic because of bad capacitors. They have a bad habit of using under spec caps. Particularly temperature and of poor brands. Teapo seems to be a big one. Seen a lot of these in some of these power supplies. Mostly in power supplies. One power supply was a dry solder joint. Most of our Doremi power supplies went down due to bad caps on the 5 volt standby circuit. I’ve replaced them then thoroughly tested them. We have a couple in service now and it’s been about a year now and they are still kicking. The rest are still spares at the moment.

[Marco Giustini]

Replacing SMD capacitors on a multi-layer PCB is not super easy but I feel one could offer a replacement and take the affected 750 to a workshop. There, on a bench, with proper tools, lighting and likely a pre-heater, the process could be completed in a reasonable time.

SMD caps could be removed with a plier - I am not a fan of the process but it's a possibility. Better, with hot tweezers. I'd say it's easier to replace SMD caps than through hole - the ground plane of a PCB can be a royal PITA.

Obviously this is not something I'd do while on site. But considering the option is a full replacement, I do see an opportunity here.

There is also the option of checking the capacitors in-circuit with a proper meter. The bad ones MIGHT show up that way.

[Marcel Birgelen]

SMD caps could be removed with a plier - I am not a fan of the process but it's a possibility. Better, with hot tweezers. I'd say it's easier to replace SMD caps than through hole - the ground plane of a PCB can be a royal PITA.

Just before the pandemic hit, I invested in a new and shiny soldering station, primarily for some upcoming personal hobby projects (most of which were put on-hold primarily due to the pandemic), so that came in handy for this particular job. Hot tweezers are pretty effective, but it takes a bit of practice to get used to them when you're only used to work with a "single" soldering iron, but it's the method I primarily used. I'm not such a fan of the destructive method of removing them with pliers, but if you're going to replace them all anyway, that's probably the most effective aproach.

I was lucky enough to get the replacement capacitors within a reasonable time frame. It's not really all that cheap though, if you buy them in low quantities.

I didn't replace any of the ceramic capacitors, as they're not really prone to failure. It's mostly the usually bigger, electrolytic capacitors that have the tendency to dry out, especially if they're of bad quality or are being overstressed.

There is also the option of checking the capacitors in-circuit with a proper meter. The bad ones MIGHT show up that way.

I remember ending up measuring all of the electrolytic caps both in-circuit and out-of-circuit after I removed them. As indicated, none of them were detected as faulty, according to the specs, a bunch of them were pretty borderline though.

My theory for my particular fault is that the board was designed with low tolerance in mind. Designing the analog circuitry of modern PCBs has since become somewhat of a lost art, most people designing those boards simply rely on their automated design tools. Over time, those capacitors start to slowly degrade, especially in hot environments. They tend to dry out a bit and slowly, some circuits go out-of-spec or become unstable and that's when stuff starts failing. It's probably not a single capacitor "failing", but it's most likely a design defect, where too little tollerance was calculated between components.

I follow Louis Rossmann on YouTube. He's not only a big factor in the "Rights to Repair" movement, but he also happens to run a repair shop that does a lot of board-level repairs, primarily on Mac hardware, which is at least as complicated as a CP750 mainboard. Maybe a shop like his can make it work. They do have the necessairy tools to do this in quantities and he also employs a few wizzards that can do complex board repairs at reasonable speeds. If you could collect enough of those broken boards and know a "sure-to-be-good" method to fix them, using a shop like this, may make it feasible to repair those boards at a reasonable price. Still, shipping this stuff around also comes at a cost, especially overseas.

[Marco Giustini]

I'm familiar with Rossmann as well, cool guy.

Electrolytics have 20% tolerance, I'm not so sure that a borderline one could cause issues - but can't say that for sure of course.
Guts feelings is
1. Caps can be tested at various frequencies, maybe your tool was using some test signal which was more forgiving compared to the 750s?
2. Some caps would misbehave when under voltage. I've seen a few myself. That said, I'd imagine it's very low voltage there so...

I was searching for a picture of the 750's MB and found this: https://www.chinahao.com/product/40964349364/

Yes, that's a lot of capacitors indeed. Knowing which one actually caused the issue would help. I suppose someone could replace them one by one and test afterwards. Once the bad ones are identified, then it could be easier?

[Steve Guttag]

Since heat seems to play a factor in the MB (and power supply) degrade, I'm wondering if a suitable fan on the CP750 could help prevent the problem (it won't fix an affected unit, however). Even more annoying, Dolby, clearly, thought about a fan as there is an unpopulated fan terminal block area near the power supply input.

[Dave Macaulay]

If a CP750 is installed with another unit directly above it or, worse, above and below it... the power supply is likely to fail pretty quickly from overheating and the chassis interior will get very warm. The convective cooling is OK with 1RU airspace above and below but with the top and bottom vent holes blocked... no good. Some fail where the DC voltages go high with some AC on them, and heat output increases. I doubt that does the board components any good at all.

If anyone is interested here's the list of capacitors Dolby suggested replacing, 128 capacitors total of 10 different types:

C323 - Panasonic / EEE-HD1H1R0R or Rubycon / 50SLV1M4X6.1 - CAP 1U 20% 50V AE SMD 4mm
x1

C180, C181, C1007 - Nichicon / UUL1HR7MCL1GS or Panasonic / EEE-HD1D4R7R or Rubycon / 50SLV4R7M5X6.1 - CAP 4U7 20% 50V AE SMD 5mm
x34

C133-C174, C909, C993-C1000 - Nichicon / UUL1V100MCL1GS or Panasonic / EEE-HD1V100AR or Rubycon / 35SLC10M5X6.1 - CAP 10U 20% 35V AE SMD 5mm
x50

C89-C104, C131-C132, C385, C404-C405, C408-C415, C900-C901, C959-C966 - Nichicon / UCW1E220MCL1GS or United CHEMI-CON / EMZE250ADA220ME7EG - CAP 22U 20% 25V AE SMD 5mm
x29

C297 - C299 - Nichicon / UUL1C470MCL1GS or Panasonic / EEE-HD1C470AP - CAP 47U 20% 16V AE SMD 6.3mm
x3

C175, C968-C970 - Kemet / EDT107M025A9HAA or Rubycon 25TXV100M6.3X8 - CAP 100U 20% 25V AE SMD 6.3mm
x4

C341-C345, C351-C354 - Nichicon / UUL0J221MCL1GS or Wurth Electronik / 865230145005 - CAP 220U 20% 6,3V AE SMD 6.3mm
x9

C596-C599 - Nichicon / UUA0J102MNL1GS or Panasonic / EEE-HC0J102P - CAP 1000U 20% 6.3V AE SMD 10mm
x4

C364, C366 - Nichicon / UCW1E331MNL1GS - CAP 330U 20% 25V AE SMD 8mm
x2

C904 - Nichicon / UUD1V470MCL1GS or Panasonic / EEE-FK1V470P or Rubycon / 35TZv47M6.2X6.1 or Wurth Elektronik / 865080543009 - CAP 47U 20% 35V AE SMD 6.3mm
x1

To whoever at Dolby gave me the list: if I was not supposed to publicly post this... well, sorry.

[Steve Guttag]

It would be interesting is to know a bit more about the list. That is, is a particular group of capacitors known to fix the level issue versus those that may cause other afflictions of the motherboard. I would presume that some of them are much more prone to the heat issue based on where they are located as well as the make/model/batch. I'm also not quite following the Capacitor numbers versus the quantity listed. For instance, the second one needing 34 pieces but only 3 designations are listed.

I suspect that, given the cost of a new cinema processor, even changing all of the ones listed, would be significantly cheaper. And, with enough seed-stock, one could have units ready, on the shelf, for an RX program.

[Steve Guttag]

Double-posted.