Diagram at:
http://home.netcom.com/~wa2ise/radios/aa5dpm5.jpg
web page:
http://home.netcom.com/~wa2ise/radio...ni.html#infdet

Well, I had moved this set from one particle board speaker cabinet to
a better and bigger cabinet. And found that there was some distortion that
I hadn't noticed before. "I broke something" I first thought. Well,
got out
the scope to hunt around. The detector triode's output looked fine, but the
other triode's output looked kinda crappy. Even after disabling the
negative
feedback from the speaker feeding the cathode. Found that the grid and
cathode had some supersonic radio frequency on them. Which didn't vary
much if at all with the setting of the volume control unless the volume was
completely zeroed. A grid stopper only allowed the RF to persist even
with the volume control at zero. What did get rid of it was a cap connected
to the grid and cathode. 0.001uF. What this does is make the gain of
the triode zero at the RF frequency. And doesn't seem to impact the
audio much if at all (It's an AM tuner so there's nothing above 10KHz
anyway).

Where did the RF come from? Seems that there was stray coupling from
the detector triode. The RF didn't seem to be an oscillation, but related
to the IF frequency.

This stray RF was making the second triode go nonlinear. Now with this
fixed, I can hear really clean audio from AM stations, both local and weak
out of town ones. Clean as in low distortion. There's background noise,
but you get other forms of background noise from records and turntables.