so after about 3 weekends of fumbling around i finally got my 40m direct conversion receiver actually receiving something. using a NE602 mixer and an LM386 for the audio stage, pretty standard stuff. the problem is the audio quality is just awful — huge hum, kind of a motorboating sound, and the volume is way too low even with the gain cranked up. i can barely make out a signal under all the noise.

i followed the basic schematic from Experimental Methods in RF Design which i know is kind of old but figured it was a good starting point. my power supply is a 12v wall wart, i wonder if thats the issue. the decoupling caps i used are just whatever i had in the junk box, like a mix of 0.1uF ceramics and some old electrolytics. ive been chasing this problem for days and honestly starting to wonder if the LM386 is just a garbage part or if i messed something up in the layout.

anyone been through this? what did you do to clean up the audio on a simple DC receiver like this?

the LM386 is not garbage but it is incredibly sensitive to layout and power supply noise, almost to a frustrating degree. the motorboating almost certainly points at your power supply — wall warts are notoriously dirty and a DC receiver is going to pick up every little ripple on the supply rail. what i'd try first is throw a 100uF electrolytic right across the supply pins close to the chip, and then add a small RC filter on the input to the LM386, something like a 10 ohm resistor in series and another 100uF to ground. that kills a lot of the feedback loop that causes motorboating.

also check pin 7 on the LM386, thats the bypass pin — it wants a 10uF cap to ground right there. if thats missing or the cap is too small youll get oscillation and generally nasty behavior. the NE602 also has pretty low output drive so make sure your audio interstage impedance matching isnt killing your level before it even gets to the amp. half the DC receiver builds i've seen have this issue and its almost always the power supply or a missing bypass cap somewhere.

yeah i went through basically the same thing with an almost identical circuit a couple years back. ended up running the receiver off a 9v battery just to prove to myself the wall wart was the culprit and the hum dropped way down. after that i added a proper LC filter on the supply line and it cleaned up enough to actually use. still not hi-fi but good enough for CW which is mostly what i use it for anyway.

one other thing — the physical layout of the LM386 stage matters more than you'd think for this kind of build. if your audio wiring is anywhere near the oscillator section you're gonna have problems. i had to basically rebuild mine on a second piece of copper board and keep the RF and audio as far apart as i could get them.