so ive been building a direct conversion receiver for 40m based loosely on the NorCal 40 design but with some mods i found in an old QST article. got it mostly assembled over the last few weekends and it does technically receive — i can hear stations but i have to crank the AF gain all the way up and even then strong signals are maybe S3 or S4 on my external meter when my IC-7300 would show S7 or S8 on the same antenna.

the mixer is a SBL-1 i pulled from an old project and the LO is a colpitts oscillator running around 7.1 MHz. RF amp stage is a single 2N3904 with maybe 15dB of gain according to my napkin math. i did the layout on perf board which i know isnt ideal for RF but its what i had. ground plane is a mess honestly.

not sure if its the mixer conversion loss, the LO leaking back somehow, or just the AF stage not being great. anyone dealt with sensitivity issues like this on a simple DC receiver? wondering if i should try swapping to an NE602 mixer setup or if thats just trading one problem for another

the ground plane thing is probably hurting you more than anything else. perf board on HF is rough, RF just does whatever it wants without a solid ground reference. ive built a few DC receivers and every time i went back and added even a partial ground plane by soldering braid or copper tape on the bottom things improved noticeably.

that said the SBL-1 has around 6-7dB conversion loss which adds up fast when your preamp gain isnt great. a 2N3904 at HF can be kind of mediocre depending on your bias point and matching. if you have a BF998 or even a J310 FET sitting around those tend to work a lot better as a front end on 40m. the NE602 swap might help but keep in mind it has its own noise figure and dynamic range issues — its not magic, just different tradeoffs. id fix the ground situation first before changing the topology

yeah what he said about the grounding. also just worth double checking your LO level into the SBL-1, those diode ring mixers want something like +7dBm at the LO port to really work right and if youre feeding it less than that conversion loss gets worse. easy thing to check if you have even a cheap scope — LO signal should be a pretty clean sine wave and reasonably healthy amplitude. ive seen people blame the mixer when the oscillator was just barely putting out enough to tickle it