so ive been wanting to build something from scratch for a while now and finally just jumped in. went with a simple direct conversion design for 40m, mostly following the Minima schematic but i kind of went off road a bit with the BFO section because i couldn't source the exact xtal values locally and figured id just use a VFO instead.

got the thing mostly assembled on a piece of copper clad i had sitting around, point to point wiring, nothing fancy. it actually receives something which i was honestly surprised about for a first attempt. i can hear SSB signals but theres this really annoying carrier hum that im not sure how to track down. tried twisting the leads to the audio amp section and it helped a little but its still there. also the image rejection is obviously not great but i kind of expected that with a DC receiver.

the VFO is drifting a fair bit too, probably 500hz or more over the first 20 minutes. i used whatever coil former i had in the junk box and the cap is just a crappy ceramic. thinking maybe i need to use an NP0 cap and maybe pot the coil in wax or something? anyone done that with good results. also wondering if a jfet buffer after the oscillator would help with frequency pulling when i change bands.

the hum thing is almost certainly a ground loop or your power supply isn't filtered well enough. with DC receivers any ripple on the supply goes straight into the audio, its kind of brutal that way. try running it off a battery and see if the hum goes away — if it does thats your answer. also check that your AF gain pot is grounded at the same point as everything else, floating grounds in the audio section will mess you up every time.

on the VFO drift, yeah NP0 caps are basically mandatory if you want anything stable. and the wax potting helps a lot, i use regular paraffin melted down. some guys use hot glue but i find that actually introduces more temp coefficient weirdness. the jfet buffer is a good idea regardless, it isolates the oscillator from load changes. i built a 40m DC receiver about 3 years ago and it drifted horribly until i added a MPF102 buffer stage, after that it settled down to maybe 50hz drift which is livable for casual listening.

im in kind of the same boat, building a 20m dc rx right now and the image rejection thing is what gets me. i know its a fundamental limitation of the design but still annoying when you hear a strong signal bleeding in from the image. have you looked at the phasing method at all? i was reading about it, basically two mixers 90 degrees out and then you cancel the image in audio but its way more complex to build. might be worth it though.

also what did you use for the mixer, ring diode or something active? ive been going back and forth on that