Chess,

I hear you … I'll be sending my ADA1000-8 in for service too and hopefully will not experience your repeated "return" loop. I’m also spending some time to make sure they will be able to reproduce the problem (which is taking way longer than I planned). I don’t want it returned as “not reproducible” or “must be a ground loop”.

My amp has a very low level buzz on all its connected channels 6 heights/tops. It’s loud enough to be unmistakable but low enough to be drowned out when the projector is on.

Note, not a ground loop ... first test I did. But while I was doing that I hit something interesting. If only one of the ADA inputs is connected to the preamp (two if the same ADA board) ...no buzz. If two or more channels were connected across amp boards ... it buzzed. This happened even if nothing was plugged into the prepro (even the power power cord)... But I needed reproducible proof without the prepro in the equation ...

I don't expect anyone to do this (even if you're an EE) ... but
After a few iterations, I ended up simulating the prepro by terminating the ADA inputs with a resistor (valued close to the prepro's output impedance) and tying all the input grounds together (didn't matter if I tied signal ground to earth ground). This seemed to reliably reproduce the buzz. I'll include that test cable when I send the unit in ... but just in case, I also took a few FFTs so they could see what I was getting here.

I'll send it in over vacation to minimize time without it.