burnt capacitors on faderboard

[ilmenator]

First time post here, I am totally new to the d8b. Located in Cologne/Germany. So I got me a d8b recently because I thought this might be a nice project - turns out the desk is not fully working as it has the flickering lights and stuck motorfader problem that seems to be caused by the infamous rail capacitors failing.

But, all of a sudden, on top of that capacitors number C501 and C801 on the 9 fader board started to release the magic smoke. First it was one of them, I managed to switch off the board quite quickly, but on the next try the next one became overly hot...

IMG_20200906_172337_kl.jpg (Array KiB) Viewed 1819 times

Though I can probably fix that, I start wondering if burnt capacitors are a known problem caused by/related to the rail capacitor problem in general, or if there must be yet another reason for these capacitors to actually burn!? I can easily replace all these Cx01 and Cx02, but will this really fix the problem or do I have to look somewhere else for a reason why these capacitors get so hot?

Thanks for any input on this!
Kind regards, ilmenator

[Jondav1120]

Hi ilmenator,

Those capacitors are indeed the rail caps. There is a lot of info in the forum on the issue, but the essence is that the original caps were (on the affected batch of boards) incorrectly marked as a higher voltage rating.
They generally seem to fail in a low resistance / short circuit mode which pulls the power rail down to a low voltage level and causes the legendary weirdness....
In your case it looks as though the two that have failed have gone completely short circuit, which will produce a lot of current (and heat) in the cap itself.
The board itself looks to be quite badly damaged, although probably repairable. You may need to check continuity to various points around the area of damage to determine whether the copper tracks have gone with the capacitor. Circuit diagram of the 9 fader board is in the database here. When looking for problems of this nature, I have found that a thermal imaging camera is very useful tracking down these overheating capacitors.

Regards

John

[ilmenator]

Thanks John,

that is very useful information! It seems like no other copper traces around that area are damaged, but for C801 a bodge wire job will be required. The underside of the PCB also shows a 3mm x 3mm black spot, but fortunately there are no copper traces in that exact area.

It will take me a few weeks to get this job done, but for now I am happy that it seems to be a problem that I can fix.

Best, ilmenator

[Jondav1120]

Hi Ilmenator,

Another thought has just come to mind....The rail cap problem generally seems to confined to the +/- 16v rails. It might therefore just be worth checking that the voltage coming from the CPU is actually +/-16v! There is a possibility that one or other of the LM317/LM337 regulators in the linear power supply has failed and is applying excess voltage to the pcbs in the console. Check this with the big power plug to the console disconnected (don't want any more capacitors burning up!)
Hope you get it sorted!

Regards

John

[ilmenator]

Hi John,

will the power supply deliver exactly +/- 16V even without a load connected?

Thanks, ilmenator

[Jondav1120]

Hi Ilmenator,

The power supply is regulated, so yes, it should deliver +/-16 volts with or without a load. The actual level is probably not that critical (from memory mine runs at about +/- 16.2v), what's more critical is that they are balanced. So providing the two voltages are within approx 5% of the correct value and within 2% of balance, you should be ok!
If you want to try testing the voltages on load, I'd pull the ribbon cables that go to the nine fader board and power it up like that. If the voltages are significantly lower than you measured off-load, then it's likely another capacitor is going somewhere!

Regards

John

[ilmenator]

So I finally got round to checking the voltages without a load: I get +15.8V and -16.2V

The PC power supply has been modified at some point, there seems to be a newer, more silent unit in there - I can tell because it is not mounted the regular way and it has an extra power switch next to the mains lead. However, this part of the PSU should not affect the +/-16V rail voltages at all, right?

Now I got me a second d8b which was said to throw the "System Error 43 - Host did not boot!" message, but apparently the previous owner just forgot to connect the 25 pin cable going from host PC to the desk...

I did a little experiment by exchanging the PC/PSU units between consoles, here is what I found:

1) PC/PSU No1 with desk No1: seems to work ok, but the desk doesn't have any faderboard installed as I need to repair them first; audio is ok
2) PC/PSU No1 with desk No2 (the "new" one): almost all of the faders are stuck on the -infinity position; the area where some of the capacitors were burnt on desk No1 gets rather hot so I switched the desk off quickly; audio is ok
3) PC/PSU No2 with desk No1 or desk No2: no audio at all, and the meter bridge doesn't show any sign of life; on desk No2 the faders are working ok, although if ganged (stereo) the other channel lags behind (not sure if this is considered normal?) half a second or more

So PC/PSU No1 seems to cause a rail voltage issue, and PC/PSU No2 seems to cause an audio issue (although all the voltages measure ok without a load).

What would be your next step?

Thanks, ilmenator

[Jondav1120]

Hi Ilmenator,

As you now have two D8Bs, I'd try and get at least one working correctly so we have a known good from which to work.
Just a thought on the no audio issue on Unit 2.....is a clock card installed in the sync slot? Might be worth swapping a known good clock card from Unit 1 to check this...
Ok, so we seem to have a power supply issue in PC/PSU No1 and possibly a computer issue in PC/PSU No2.
Bit of a bizarre idea, but you could use both PC/PSUs with one console for testing purposes. By this I mean use the power supply in one PC/PSU to power the console and the computer in the other PC/PSU to run the console. To do this, connect the BFC from PC/PSU No2 (good power supply) to the console ( this powers the console). Do not connect the data cable from PC/PSU No2 to the console. Now connect the data cable from PC/PSU No1 (good computer) to the console. Switch on PC/PSU No2 first and then No1. I can't say for sure that this will work as I haven't got two D8Bs to try it on, but I can't see any reason why it shouldn't, so please don't feel pressured to try this if you aren't comfortable with the idea!

Regards

John

[ilmenator]

Hi John,

so I did as you suggested: console 2 was powered by PC/PSU No2 and connected to PC/PSU No1 via the data cable. Boot up went normal, the console did not notice any difference .

However, I get the same behaviour as with power from PC/PSU No1: audio works fine, but the faders are stuck at -infinity and the fader boards get quite hot.

The two consoles have pretty similar serial numbers and the same manufacturing date code (9M1811) so I think it's time to replace the rail capacitors on the brain board and on the fader boards. However, one of the consoles has a blue brain board, the other one has a green board - I had hoped that maybe one had been replaced in the past already, and I would get away with that...

I have also ordered the AD8804 chips, I think I read somewhere that these could be the source of the stuck fader problem. I'll look into the capacitors first, that's a cheap fix, and if that doesn't solve the issue then the AD8804s are due.

Thanks, ilmenator

[Bruce Graham]

Hi ilmenator;
The faders being stuck at infiniti is a chip on the faders that controls the motors. There are 2 different types, On the 8-way fader boards it is ADA803 and on the 9-way it is ADA804. It think! There is a post in here and info im the data base.
Once installed, it's not hard, run fader calabration a few times. I ordered a few spares as they do go. Ebay or "Funk" here in the forum as a few to sell I think!
Cheers
Bruce