IRS20957S / ICE 2A165 based amp: Ampeg SVT 7pro +15 VDC rail short but why?

[frozenfrogz]

Dear all,

I am working on said SVT 7 pro bass amplifier. Already repaired a couple of these that all had an issue with overheating SMPS IC (ICE 2A165, thermal shutdown), that could be handled by replacing the IC with its next bigger counterpart 2A265 and mounting an additional heatsink on the DIP-8 housing.

The unit I am working on right now has some more issues though. I replaced the SMPS IC and found it to be running pretty hot in spite of the added heat sink. After some further evaluation I could pin point the fault on the +15 VDC rail produced by T3-C, D14.

The issue seems to be quit similar to what nickb stated in comment #14 on this music forums thread.
Also found this thread on our forums, but in my case the issue is different.

What (I think) I know:

• excessive current draw on +15 VDC rail after some time running
• +15 VDC rail measures 20Ω to analog ground AGND (tube heater) or 390ishΩ without preamp board connected when cold
• +15 VDC rail shorts to ground after some time running (measuring 0Ω from U37 pin 1 to AGND, bounces back to expected values as stated above after some time unpowered)
• checked U37 (+5 VDC regulator) -> working fine
• unsoldered R205, R110, R223 with no change in behaviour

Something is shorting +15 VDC to AGND after some time and right now I am literally thinking of shotgunning every electrolytic cap. There is a whole lot of silastic everywhere and working on this thing is not much fun.

I would gladly take some pointers from you guys and try some fresh ideas.

Thank you!

Edit:

Some temperature measurements taken with my thermal camera

U15, U18 around 50°C
R35, R36 -> 40-50°C
1W and 1/2 W resistors between 50°C and 70°C
ICE 2A265 @ 50ish °C until circuit runaway, then 90°C+ and I power the unit down.

[JKKDev]

You could try pulling the R285 off of the board to take one section out of the equation. TBH I don't know how the different parts of the amp would react if you start turning off the 15V rail so do this at your own risk An even wilder suggestion would be to cut the traces to every section using 15V rail and turning them back on one by one using solder bridges (you probably shouldn't do this but who knows how desperate the situation is )

Maybe you could even just run the PSU board on its own and load the 15V rail to around where it should be and see which board is causing the problem.

I hope someone gives you some better suggestions

[frozenfrogz]

Thank you for the suggestion.
I actually already pulled R285 after writing my initial post. Error is persistent, but we can rule out that part of the circuit.

[frozenfrogz]

Just in case someone of you has time to kill because of the coronavirus: I would kindly take any further advice!
Also, if any of you can manage to find a source for revision G00 schematics that I can buy, please point me into the right direction. Since Ampeg was bought by Yamaha, there is no chance to officially get schematics unless you are an "officially authorized" repair slave.

[frozenfrogz]

After unsoldering,testing them out of circuit and resoldering a bunch of electrolytic caps, the recurring error of excessive current draw after some runtime did not return. Somethings seems to be fixed, however I find that situation rather dissatisfying.
Base temperature of the new 2A265 with added heat sink is not pretty (heat sink stabilizes at around 90°C). Since I do not want to go through the trouble of redesigning the power supply for the 12AX7 heater current, I am going to stick the biggest heat sink I can on that chip and call it a day.
That amp including the miserable board design is a total clusterf*ck...

By the way, I also found a dying power cap on the input side. 200V 2200uF that was down to 990uF, starting to bulge at the top.
CapXon chinesium caps..