so ive been sitting on a pile of parts for like two years now and i finally decided to just do it and build a simple direct conversion receiver for 40m. been going back and forth between the NE602 based designs and just rolling my own with a diode ring mixer but honestly im leaning toward the NE602 just to keep things manageable since this is my first real homebrew RX project.

my plan is something loosely based on the Sudden receiver design but ive made a few changes to the audio chain because the original felt like it was going to be a bit deaf in the AF department. added a second stage of audio gain after the LM386 section. not sure if thats going to cause oscillation issues or what, i havent breadboarded it yet.

the VFO is the part im most nervous about. i want something stable enough to actually copy CW without the thing drifting all over the place every time i breathe on it. thinking about a Hartley or maybe a Colpitts — any strong opinions on which holds up better thermally in these kinds of low power homebrew situations? ive read a bunch of posts going back and forth on it and nobody seems to fully agree.

anyway if anyones built something similar and has gotchas to warn me about id really appreciate it before i start committing parts to the board

colpitts all the way for 40m in my experience. the hartley can work fine but youre adding an inductor tap and that just means one more thing to wind and get wrong. with a colpitts you can tune the ratio of the capacitive divider to control the feedback and it tends to be more predictable thermally if you use decent NP0 caps in the tank. i built three or four of these over the years and the ones that drifted were always the ones where i got cheap on the capacitors.

for the audio gain thing, yeah two stages can definitely oscillate if you're not careful about decoupling. put a good RC filter between the supply rails for each stage and keep the high gain stage physically away from anything thats picking up RF. even a couple inches on a piece of ugly construction can make a huge difference. the LM386 is notorious for doing weird things when the supply isnt clean.

the NE602 choice is fine for a first build, just watch the input impedance. a lot of people forget to match their antenna to the front end properly and then wonder why the thing is deaf. little LC filter at the input does wonders.

i did almost exactly this project last winter, also 40m, also based loosely on the Sudden. mine ended up working but i had a really frustrating issue with images for a while that turned out to be just RF getting in through the power lead — completely obvious in hindsight but i spent like a week thinking my mixer was blown or something. ferrite bead on the supply line fixed it immediately.

one thing i'll say is dont underestimate how much work the enclosure does for stability. i breadboarded mine and it drifted like crazy and then i moved it into a little aluminum project box and it calmed way down just from the thermal mass. nothing fancy, just one of those hammond boxes from the usual suppliers.