Brian wrote:

I'd like to see a schematic of this amp posted at ABSE or somewhere,
so we could see it.

I wish I had a diagram. This is an early Knight kit, deco styling, no
model number, just says "KNIGHT HI-FI 10 WATT AMPLIFIER." Works fine,
with or without the funny feedback.

Since the gain effect of the funny feedback is only 3 dB, I'm still
guessing that it is simply intended as a cheap way to bypass a cathode
resistor. Bypassing that stage with an actual capacitor yields the
same gain increase.

Next time I have the amp out, I'll run it through my spectrum analyzer
and see whether the feedback affects distortion. I couldn't see any
time-domain effects on the sine wave I was using to determine the
feedback gain effect.

Brian

Part of good tube craft and service work on many items of
electronics for me
includes tracing out the schematic in my work book,
and this makes it clear what the total picture of the amp looks like,
so I can calculate and draw in all the instantaneous working signal
voltages
with their approximate phase relative to the input.
Its slow and laborious, but this works for me,
so I under stand what the feedback is achieveing
in empiracle terms, and if it appears to be what I think
is good practice, then all is well.
I can then share this with others,
with a scan and send, but only if I have kept the drawing neat enough.

That measuring with the analysing might well be interesting.

Patrick Turner.