so ive been building this 40m direct conversion receiver for the past month or so, based loosely on the Sudden receiver design but i swapped out a few parts because i couldnt source the originals and honestly just wanted to try some things. the RF front end seems fine, i can hear signals, but the audio stage is just absolutely brutal. lots of hum, and theres this weird heterodyne-y kind of whine that follows me around the band when i tune the VFO. im using a LM386 for audio amp which i know is kind of a meme at this point but it was what i had on hand.

ive got bypass caps on the supply pins, or at least i thought i did. checked my layout and the ground return for the audio section is kind of a mess if im being honest. wondering if its a ground loop issue or if the LM386 is just oscillating. the whine tracks the VFO which is the part thats really bugging me, like it changes pitch when i tune. not totally sure if that's RF getting into the audio or something else going on. any ideas before i rip the whole thing apart again

that whine tracking the VFO is almost certainly audio-frequency beating from the VFO leaking into your audio chain somehow. happens all the time with direct conversion stuff because the whole thing is basically working at audio frequencies already and any little bit of VFO signal that sneaks into the wrong place will demodulate itself. check if you have any common impedance between the VFO ground and the audio ground on your board, and also look at whether your detector output line is shielded or at least short and away from the oscillator.

the LM386 oscillating thing is a separate issue usually, classic fix is 10 ohm resistor in series with the output and make sure pin 7 is bypassed to ground with something like 10uF, a lot of people miss that one. but honestly your main problem sounds like layout. DC receivers are unforgiving about that stuff in a way that superhet designs kind of mask.

yeah i built something similar last winter and had almost the exact same problem. ended up being that my ground plane was basically nonexistent, i was just using point to point wiring on perf board and every ground connection was kind of daisy chained which is a disaster for this stuff. redid it on copper clad with an actual ground plane and most of the hum went away. still had a little of the tuning-related whine but way more manageable. also wound up putting the audio board in a separate little tin from the RF/oscillator section which helped a lot with the RF getting in.