Nick's video appears above.
First, let me say FB Nick. It is nice to see you making use of the AGC amp designed by Wes and Bob, using the board from Todd's Mostly DIY RF, using a mix of homebrew pads and Me-Squares from Rex, and finally the Franklin Oscillator that we spent so much time talking about on SolderSmoke.
But here are some comments:
-- I still don't think you need that RF amplifier in front of the mixer. And I suspect you would be better off without it. We did not use one in our high school 40 meter project, and never missed it. In fact, on one version of the high school receiver I even put in a simple 10k pot as an attenuator (no RF amp). Even up on 20 meters, I do not have an RF amplifier ahead of the diode ring mixer on either of the Mythbuster rigs I have built. Nick, maybe experiment a bit more and try the receiver just going from the BP filter into the mixer and see what happens.
Note that Wes W7ZOI DID NOT have an REF amp ahead of the diode ring mixer in his original 1968 40 meter Direct Conversion receiver (the one that launched the solid-state DC recevier revolution):
-- The Franklin oscillator is an interesting, but complicated circuit. The gimmick is, well, gimmicky. Here is the thing: You can achieve similar levels of stability using simple conventional, single transistor oscillators. We dispensed with the variable capacitors, and used PTO--style variable inductors. They worked fine. This Franklin oscillator still does seem to drift a bit, right? I would ground the board to the inside of the metal box.
I would also try putting all the stages on a single ground plane. This might help.