[larry.said]

I build the PPA v2 but noticed only after turning it on that I had used an AD843 (not unity gain stables) in the ground channel. Maybe that screwed up the ground channel because now my Q33 (in the buffer section) heats up to the point where I can't even touch it anymore. I replaced all the transistors and tried again (with an OPA134 in OPAG) and the same problem arose...I checked the traces on the board and everything is being routed to the right place. The buffer section in the right and left channels does not have this problem, only the ground channel...Anyone know what could be going on?

 

[FallenAngel]

AD843 IS unity-gain stable but there have been quite a few reports of it not working in PPAv2 for some reason.

Q33 is in no way related to the opamp so it heating up is a sign of another problem. Post a detailed photo of the top and bottom of your board, another pair (or lots of pairs) of eyes should help catch any problems.

 

[TzeYang]

because the current draw is rather high.

You have to somehow swap the rail isolation JFETs to JFETS with a much higher IDSS value.

 

[mwofsi]

I was thinking of trying the AD843 in a PPA. Tangent recommends against it.

I think the explanation was that the 843's have a tendency to lock-up, slamming to one power rail and causing high dc offset on the output, if they don't get enough current as they start up. As TzeYang said, in the PPA the jfets are what cause the problem.

I don't know if it's related to your Q33 problem tho'. Sounds like your getting too much current thro' Q33. Looking at the schematic check R33 and perhaps R32 that there the right values and that they haven't failed to a short.

 

[larry.said]

Quote:

Originally Posted by TzeYang /img/forum/go_quote.gif because the current draw is rather high. You have to somehow swap the rail isolation JFETs to JFETS with a much higher IDSS value.

I'm using 2N4392s with IDSS of 20 mA