Hi folks-
I have an interesting version of this problem that I have not seen addressed - and am hoping one of you can provide guidance.
The problem is thermal drift in the offset of the a IG02940. Here's the sad tale:
(As context, I am a reasonably experienced amateur tech, having repaired/restored vintage gear for the past 15+ years.)
A friend brings me a Yamaha CR-840 complaining of sound quality issues. Sure enough, the left channel output shows a strange distortion when a sine wave is input (and one I have not seen before)- jagged "sawtooth" on the falling side (and only the falling side) of the sine wave at all reasonable levels. I swapped the left and right predrivers and the problem swapped with them, so ouala a bad IG02940. Lacking any alternative, I ordered a replacement on ebay from China, from the most promising sounding source I could find.
Three weeks later the part arrived. Slightly different package and VERY faint lettering, but whatever. Soldered it in, and ..... nothing. Complete brick. Took it out and checked various pins with an ohmeter. No chance it was a real 2940 (for example, diode drop between pins 1 and 4 has wrong polarity). Complained to supplier, who promised to send a replacement that "looked more like the original. Another three weeks pass and the replacement arrives. It does in fact look more like the original. And passes preliminary uninstalled checks. Installed, powered on, and within 15 seconds there is a loud, smoky 14 (between pins 7 & 8) has completely blown its can off and deposited threads of its internals all over the board. Wonderful. No more Chinese parts.
Was astonished to see a NOS IG02940 appear on ebay and snagged it. In the original Yamaha box with all legit markings, package looks great, long pins intact etc. The real deal. Installed it and, as a precaution, replaced all of the small electrolytics around it (C704, C708, C714, C716, and even C126) in case the bogus predriver had compromised them.
Now the problem: Initial measurements showed the infamous "pins 7, 8, and 10-14 all at 40+ volts". Rats. But as I watched, the situation changed. All of these slowly dropped into normal ranges. After about 15 minutes, everything looked more or less correct, with undistorted full amplitude outputs on pins 13 and 14. There is still significant offset at the junction of the power pak emitter resistors (a couple of volts) - but this is true on both channels so I am hoping this is operator error of some kind.
So something has major thermal drift. I hit the power pak with cool spray and nothing happened. Then I applied cool spray to the 2940 and within seconds it railed out again. Then, again, as it warmed up, the output reappeared and gradually sank to the proper range (centered near ground). So there is some kind of thermal issue with the predriver.
There is no obvious thermal compensation circuitry within the predriver or on the board. And I have been unable to find any mention of this problem on AK or elsewhere. I am really hoping that this miracle "last one on the planet" NOS part is not bad. Thanks in advance for any help!
By the way - the CR-840 service manual from hifiengine does not match this unit at all - different shsape amplifier board, different parts, etc. I found that the one for the CR-640 has the correct layout and schematics (and shows IG02940 and IG02970 as are in this unit).