[Heathkit] SB-101 AVC problem

[Chris Codella, W2PA]

I recently finished restoring an SB-101 that I originally bought as a 
junker (for parts), by completely disassembling it, cleaning everything, 
replacing bad parts, and reassembling (except I didn't remove components 
I suspected were good from the PC boards).  It worked out well and was 
fun - almost like building a new Heathkit.  Everything seems to be 
working fine except for some funniness with the AVC.

Strong signals seem abnormally loud in the audio stage.  For example, a 
carrier at S9+20dB is much louder in the headphones than one that is, 
say, S5.  This should not be the case if the AVC is functioning 
properly.  I also notice some abnormal behavior in the S-meter.  While 
strong stations register at S-meter levels similar to what I get for the 
same signals on my other rigs, weaker signals and background noise that 
would normally be around S-4 to S-7 don't register at all on the 
SB-101.  It's also been my experience with Heathkit transceivers that 
when you turn the RF gain down all the way, the S-meter pins high.  This 
one doesn't - it stops near full scale.  No one of these things would 
raise a flag with me, but taken together they point to (probably) a 
problem in the AVC function.

I plan to do some measurements to try to figure this out - for example I 
would imagine that measuring the input signal to the product detector 
should show nearly constant level for various input signal strengths, if 
the AVC was operating properly.  But thought I'd pick the collective 
brain on the Heathkit reflector first.  Any thoughts?

73,
Chris, W2PA