so ive been working on this double sideband transmitter for the past few months, mostly based on the old Solid State Design for the Radio Amateur book with a few tweaks. using a diode ring modulator built with 1N4148s, balanced with a trimpot across the secondary of the input transformer. been getting maybe 30-35dB of carrier suppression on a good day but when i fire it up and listen on another receiver it still sounds a bit carrier-heavy, like it wants to be AM almost.

the audio sounds fine honestly, my dummy load tests look reasonable on the scope, but when i try to have a qso the reports i get back are like 'sounds a little buzzy' or 'you there, you sound kinda rough'. i dont know if its the suppression thats the issue or if my audio chain has some problem, the mic preamp is just a junkbox lm741 stage which probably wasnt the best choice looking back at it now.

anyone built something similar and run into this? wondering if i need to go back and rebuild the modulator with matched diodes or if theres something else i should be checking first before i tear it apart again

30-35dB should be fine for casual ssb use honestly, thats not your problem. the 741 is almost definitely part of it though, those things have terrible slew rate for audio and will add their own distortion especially if youre pushing any kind of level through it. id swap that out first before you go touching the modulator. something like an NE5532 or even a TL072 would be a massive improvement and cost you like nothing.

also the 'buzzy' reports could be RF getting back into your audio chain. does it get worse when you key up vs just audio testing on the bench? if so id look at bypassing on the mic input and maybe a few ferrite beads on the mic cable leads. ive chased that exact problem for way too long on a couple different builds before i figured out it wasnt the modulator at all.

matched diodes make a real difference in my experience. i built a similar ring mod a while back, just used whatever 1N4148s i grabbed from the bin and the suppression was mediocre at best. bought a bag, measured forward voltage on like 30 of them with a DMM in diode mode and sorted them into matched pairs and the improvement was noticeable. got up around 40dB pretty easily after that with just minor trimmer adjustment. its tedious but worth the hour you spend doing it.