If you can lay hands on a variac, use that to power the unit. This kind of
fault then becomes a breeze to locate (the basic cause of). If it turns
out to be related to the output stages - and I'm not quite sure of what
you are measuring to come up with your "800 ohms" figure - then how simple
or not the *actual* problem is, is a whole new ball game with DC coupled
amps like this ...



Use a large incandescent lightbulb wired in series with the unit for
testing. It's cheaper than a variac and in this case is a better solution.
200-300W should do it.