Is Q2 C to E shorted?
It's going to be something like that. Not much different to the output stages on TTi stuff.
I'd load the output with 50 ohms and then work back up the signal chain from the output until you get a signal that isn't sick.
Naaahhh... I already checked the obvious stuffs: MCP at output tests good with junction tester, no obviously shorted diodes, relays or tants (this thing is peppered with the little black ones, but all being run minimum of 10V below rated) in my original post before I even started searching for tech ref.
My money is on the AD811ANs; they're being run at +/- 15.5V on a 15V nominal/18V max rated part. They also have heat-sinks made from hand-cut scraps of U-channel bonded to the chip with silicone sealant; hardly the best choice for thermal conductivity.
My personal suspicion is that this unit was one of the dregs... the last few mainboards bought from Berkeley Nucleonics... whipped out with whatever they had in the lab that didn't require any cash outlay in an effort to stave off the creditors for another quarter.
Given all the other sins committed therein, and where they tapped into for +5V to run the thing (that requires a pretty intimate knowledge of the mainboard, as it is not an obvious first choice) off one of the solder pads for the RESET button, I wouldn't be surprised if the 30mm-fan-in-a-40mm-hole came out of the B&K Labs that way too.
That said... I think I'm going to step away from the bench for a afternoon... spend some time widda wife & kidzz. Bondene to fix my 54645A should be here soon, and I want to have my scope back in one piece before I go poking around this thing's innards. Using it nekkit to do triage on this thing was enough taunting Murphy for one project.
mnem
Everything is a learning opportunity.
Well, that was wholly unsatisfying... didn't even get to turn on my freshly-resurrected 54645A.
Did a little cleanup on the schematic I'd glommed to make it printable; noticed that the analog signal path is neatly divided at several points by a few resistors.
"I have a good feeling aboot R69..." I said to myself... so slid it off and with the soldering iron and BAM! (-12V) on the output goes bye-bye... so not the MCP, probably not U52 AD811AN feeding it. Remove R59 & R63, reinstall R69, -12V is back. Remove U48, (-12V) on the output goes away again... no DCV at open R59 or R63 from further up the chain.
Just to be sure, plug u48 back in the holes, -12V is back.
So... now shopping a couple AD8911ANs & a few resistors somewhere in Canada, and some proper heat-sinks with frag tape to put on 'em.
*sigh*
mnem
*toddles off to like... school the kiddzz or sumthin' *