so ive been building this direct conversion receiver for 40m on and off for like 3 months now, mostly following the NorCal 40A schematic but i made some changes because i couldnt get the right toroids locally and had to substitute. anyway got it all put together last weekend and it actually receives signals which honestly surprised me considering how messy the wiring looks inside the enclosure.

the problem is the audio sounds kind of hollow and there's this low frequency rumble that comes and goes. like if i tune to a CW signal it sounds like the tone is inside a tin can. SSB is barely intelligible. i checked my LO and its sitting pretty stable on the scope, the NE602 seems to be getting the right supply voltage. audio stage is just an LM386 feeding a small 8 ohm speaker. im wondering if its an impedance mismatch between the mixer output and the audio amp input or maybe my bypass caps are insufficient somewhere but honestly im just guessing at this point

anyone built something similar and had audio issues like this? would really appreciate some direction before i start randomly swapping components

the hollow sound on a direct conversion receiver is almost always one of two things in my experience — either your audio filtering is letting through a ton of RF that the LM386 is partially rectifying, or you have a ground loop somewhere in the enclosure. the LM386 is notorious for being sensitive to layout, pins 1 and 8 especially. did you put a cap between those for extra gain? if you did and your layout isnt tight it will oscillate or just sound terrible.

also check what value you have from the NE602 output to the LM386 input. a lot of people use too low a value coupling cap and it rolls off the low end making everything sound thin and hollow. try something like 1uF there and see if it helps. and make sure pin 7 on the LM386 has a bypass cap to ground, 10uF usually does it. the rumble you're hearing could be hum getting in through the supply if that isnt well bypassed

yeah direct conversion receivers can be a pain with audio, ive built a few and they always need tweaking. one thing i dont see mentioned often is that the physical layout of a DC receiver really matters a lot more than you'd think, the LO leaking back into the audio stage is a real thing and can cause all kinds of weird artifacts. shielding the oscillator section from the audio section even with just a piece of scrap PCB material soldered in as a divider can make a pretty noticeable difference.

what toroids did you end up substituting btw? the bandpass filter on the input side has a big effect on what garbage gets through to the mixer and if the impedance transformation is off you might just be letting in way more signal than the NE602 wants to see