so i finally got around to building a direct conversion receiver for 40m, been wanting to do this for like two years and just kept putting it off. used the NE602 as the mixer/oscillator and its actually working way better than i expected for a first attempt. picking up SSB signals no problem, can hear CW fine too.

the issue im running into is the audio stage is kinda hissy and there's this low frequency rumble that i cant seem to get rid of. i've got an LM386 in there for the final audio amp which i know isnt the most elegant solution but it was what i had on hand. ive tried bypassing various points with caps and moving wires around but nothing really seems to help the hiss much. the rumble gets worse when i touch the enclosure which makes me think grounding maybe? the whole thing is built on a piece of perfboard in an old project box, not exactly ideal for RF stuff.

also the VFO drifts quite a bit when it first powers up, like maybe 2-3 kHz over the first 10-15 minutes before it settles down. i assume thats thermal drift from the oscillator? the coil is wound on a piece of PVC which i've read is not great thermally. wondering if i'd be better off rewinding on a toroid or just living with the warmup drift for now since its not the end of the world for casual listening.

anyone been down this road with the 602 and LM386 combo? curious what people did about the audio noise specifically

yeah the LM386 is notorious for being hissy, thats kind of just what it does especially when you've got the gain cranked up with the cap between pins 1 and 8. a lot of guys building these things swap it out for an LM4562 or even just a TL072 based stage with a bit more thought put into the gain structure. the 602 itself adds some noise too, its not exactly a low noise figure device but for 40m its usually fine since the band has plenty of signal above the noise floor anyway.

for your grounding problem, what youre describing with the rumble getting worse when you touch the enclosure sounds classic ground loop stuff or maybe your audio ground and RF ground are fighting each other. i had basically the same thing on a 80m receiver i built a few years ago and what fixed it was running a single point ground back to the power supply rather than just relying on the board traces to tie everything together.

the VFO drift on PVC is real, that stuff moves a lot with temperature. if you want a quick cheap fix try winding the coil on an FT-50-61 or similar, even a small ferrite toroid will be way more stable thermally than the PVC form. or if you want to get fancy, using NP0 ceramic caps in the oscillator tank circuit helps a lot with drift compensation, the negative tempco of the coil and the positive tempco of most caps kind of cancel out if you mix the values right. its one of those things you end up tweaking by feel mostly.

dont overlook the power supply decoupling, i built something real similar last winter and had all kinds of audio garbage until i added a bigger filter cap right at the LM386 supply pin, like 100uF right at the chip. also make sure you have a cap on pin 7 of the 602, the datasheet recommends something there and i forgot it the first time around and got all kinds of weirdness. small thing but easy to miss when youre just winging it on perfboard.