[Jack[_11_]]

Dave Platt wrote:

If you use a lower frequency (e.g. 100 Hz) it'll be within the
specification range of almost any simple AC voltmeter, and should read
out properly... you can trace the signal through both channels, and
see where the differences first appear by comparing the voltage
readings.

At the output stage, playing the 100Hz tone, measuring across a single
transistor and ground, using the 200v AC meter, with the speakers turned
off, I read about 38-39 volts at maximum volume, so this may be useful.
Thanks for that tip. But since I turned the volume to max, the problem
has disappeared so I have to power down the unit until it returns. In
the meantime I will map out my probe points.