Fluke/Philips PM2534 Error 4, low inguard power supply output

[ogdento]

Hi all,
I bought a pair of non-working PM2534s on fleabay... one just needed some mechanical help and now works great, but the other is definitely bad.  Both are pcb versions 8051.4 and appear to have been built the same time.

The bad unit boots up and beeps continuously... the self test fails with an Error-4 (adc error).  Per the service manual I started with the power supplies and the outguard (+5v) is fine but the inguard supply +/- A (8v), and -L (-6v) were low.  A thermal image of both meters showed that some inguard regulators on the bad unit were getting too hot... the "bad" meter shows a 110mA draw when plugged into a kill-a-watt meter vs 60mA for the working unit.  I'm not very experienced with troubleshooting such complex stuff, but I figure there's a blown component somewhere in the analog section - I just can't find it!

I pulled the piezo and all the electrolytic caps in the inguard... the caps all tested ok on an esr meter but I temporarily replaced them anyway - no change.  I then pulled the output legs of the abnormally hot regulators (+A, -A) and the draw of course dropped to ~60mA and the outputs were good except +A was 7.7 (still .3v low).  I then tested each regulator independently and they all seem to work.  The lm337 for -L (-6v) felt hot, but it gets hot on the working meter too. 

Since something is drawing too much current, I was afraid to run both meters side by side long enough to check/compare voltages for every component - figuring I'd smoke something else.  So I in-circuit checked all the diodes and many of the transistors, and checked all the other caps for shorts (using continuity check), and compared some of the readings to my good meter.  The ptc R1152, resistor R1103, and triac V1153 on the input stage checked ok as well.

The manual doesn't have much as far as troubleshooting, but it does outline which components go with each analog block (and parts are numbered according to block) - I'm thinking the next step is to de-power each analog block until I find the culprit but that likely involves cutting traces on the board.  I'm on the fence of whether it's worth the time, but before I start that process...

1. Am I overlooking something glaringly obvious?
2. Besides the electrolytic caps, is there another typical analog block or component failure in these things that would cause a high draw on +A/-A? 
3. Do the non-electrolytic caps on these fail often (i didn't find any shorts using continuity, but i didn't pull them all and test them)?

Thanks!

I attached a photo of the board, focusing on the inguard section.