so ive been working on this for probably 3 months now, its a single band SSB exciter for 40m based loosely on the W7EL design but ive made some changes along the way mostly because i couldnt source the exact parts. built the balanced modulator first and got that working pretty well, carrier null is decent, maybe -45dB which im happy enough with for a first attempt.

the problem im running into now is theres this kind of low frequency feedback squeal that kicks in whenever i talk into the mic. its not constant, like it only happens when theres audio present and it seems worse when im running higher mic gain. ive tried swapping the electret element and that didnt change anything so i dont think its the capsule. the feedback loop seems like its coming from somewhere in the audio chain before the balanced modulator but i cant pin it down. ground loops maybe? ive got the whole thing in a steel chassis but the grounding is a bit of a rats nest honestly.

anyone dealt with this before on a homebrew SSB rig? running out of ideas here and i really want to get this thing on the air before winter

classic ground loop symptom honestly, especially with electret mics since they need that bias voltage and if the return path isnt clean you get exactly what youre describing. what does your mic preamp power supply look like — are you pulling it from the same rail as the RF sections or is it regulated separately? ive seen guys have this exact issue and the fix was just adding a small LC filter on the 8V mic bias line, something like 100uH and 10uF tantalum right at the preamp board. also check if the feedback changes at all when you touch the chassis, if it does thats almost definitely your ground scheme. the rats nest thing you mentioned is probably the culprit, i'd spend an afternoon tracing every ground return back to a single star point before chasing anything else.

might also wanna look at physical layout, like how close is your audio wiring running to the carrier oscillator or any of the RF stuff? even in a steel box you can get enough coupling to cause grief if a mic lead is running parallel to something hot. i had a similar thing on a 20m rig i built years ago and it turned out one of the audio leads was literally draped over the VFO can. moved it like 2 inches and the problem went away completely. also -45dB carrier null is pretty solid for a first build, mine was embarrassing when i first fired it up