so ive been messing around with this direct conversion receiver for 40m that i breadboarded out over the past few weekends. started from a design i found in EMRFD, modified it a bit because i couldnt source the exact NE602 pinout board someone linked in another thread and ended up using a SA612 which is basically the same thing i think.

anyway it works, sort of. im picking up signals fine, mostly SSB and some CW, but theres this really annoying hum that comes and goes. its not 60hz, sounds more like 120 to me but i havent scoped it yet. also the audio stage is way hotter than i expected — i just threw in a LM386 because i had a pile of them in my junk box and it clips pretty bad even with the gain pin left floating.

also wondering if anyone has tried winding their own toroids for the bandpass filter ahead of the mixer stage. i did mine on T50-2 cores with what i think is the right number of turns but the thing kind of loads down if i touch a longer piece of wire to the antenna terminal, like resonance is way off. might be my L calculations were off or just sloppy winding.

not looking for a kit recommendation, just want to get this thing actually working before i move on to the tx side of the project

the 120hz hum is almost certainly your power supply ripple getting into the audio somewhere, classic LM386 problem actually. that chip is notorious for picking up supply noise especially if you're running it off anything less than a well-filtered bench supply. try putting a 10uf electrolytic and a 0.1uf ceramic right at pin 6 to ground and see if that cleans it up. also make sure your audio ground and your RF ground arent sharing a long trace or wire because that can couple all kinds of garbage in.

on the toroid question — yeah T50-2 is fine for 40m but if your resonance is drifting when you connect an antenna it usually means your Q is lower than it should be, probably from loose or unevenly spaced windings. i wind mine as evenly as i can around the core and then use a dab of Q-dope to lock them. also double check your capacitor values in the filter, i once spent two hours debugging a BPF before i realized i had grabbed a 100pf cap from the wrong bin and it was actually 10pf. easy mistake.

SA612 is fine, ive used them interchangeably with the NE602 plenty of times, no real difference in practice at HF.

the LM386 clipping thing — gain pin floating still gives you 20dB which is actually a lot if your mixer output is already decent level. you might just need a voltage divider between the mixer audio output and the amp input to pad it down a bit. like even a simple resistor divider, 10k and 1k or so, gives you a lot of headroom to play with without adding much noise. i know it feels wrong to attenuate before the amp but with the 386 sometimes thats just how it goes.