Yes, both sets of jacks are isolated from the chassis. However, one amp had it's inputs tied to chassis ground so in effect one set of inputs was connected to chassis and one was not.
Either amp, used by itself, is completely quiet.
Yes, it is two paths to ground when one amp has a connection from input to chassis ground and one does not. Because chassis ground is relative to the transformer ground on each amp, and the transformers don't tap at the exact same voltage, so there is a potential created between the two amps. The slight ground current is running from one amp to the other via the input cables. Two paths to ground.
I agree, amp A was creating the potential through that input chassis ground. Lifting the ground on the inputs on amp A did break the loop.
As I said, this does fix the problem safely. I don't mind continuing the discussion in the hope that someone else will find this helpful. Ground loops can be very difficult to solve and I know how much time I spent on this.
This stuff is only fun after you find the solution. :-)
Either amp, used by itself, is completely quiet.
Yes, it is two paths to ground when one amp has a connection from input to chassis ground and one does not. Because chassis ground is relative to the transformer ground on each amp, and the transformers don't tap at the exact same voltage, so there is a potential created between the two amps. The slight ground current is running from one amp to the other via the input cables. Two paths to ground.
I agree, amp A was creating the potential through that input chassis ground. Lifting the ground on the inputs on amp A did break the loop.
As I said, this does fix the problem safely. I don't mind continuing the discussion in the hope that someone else will find this helpful. Ground loops can be very difficult to solve and I know how much time I spent on this.
This stuff is only fun after you find the solution. :-)