so ive been working on this double sideband transmitter for 40m on and off for like three months now, mostly following the W7ZOI designs from Experimental Methods but tweaking a few things based on what i had in the junk box. got it actually putting out a signal last weekend and made a couple contacts which was pretty wild honestly.

my balanced modulator is a diode ring using 1N4148s, nothing fancy, and the carrier null is sitting around 40dB down which i think is decent for what it is. the final is a single IRF510 running class E, getting somewhere between 3 and 4 watts out depending on the supply voltage i feed it.

my question is about the low pass filter after the final. i just threw in a 7-element chebyshev i calculated with one of those online tools and it seems to be working but ive never actually verified the harmonic suppression with anything better than my cheapo rtlsdr. does anyone have experience eyeballing whether a lpf design is trustworthy or do i really need to throw it on a real spectrum analyzer before i start using it on the air regularly. i dont want to be that guy lighting up harmonics all over the place

the RTL-SDR honestly isn't terrible for a rough check if you know its limitations — you can at least confirm the second harmonic is way down relative to your fundamental. that said, if you know anyone local with a real SA, even a 20 year old HP or Rigol or whatever, it's worth an afternoon trip just to see what's actually going on. the 7-element chebyshev should be more than adequate on paper for 40m harmonics but component tolerances matter, especially if you wound your own toroids and they're a bit off. what core did you use for the inductors?

also 40dB carrier suppression on a diode ring with 1N4148s is actually pretty solid, ive seen kits ship worse than that. you might get another few dB if you hand-match the diodes but honestly at that level it's probably fine for casual operating.

yeah get it on a real spectrum analyzer if you can, even just once. a buddy of mine built a similar rig and was convinced his LPF was fine based on sdr checks and then borrowed someones SA at a hamfest and found a third harmonic that was only like 25dB down. not catastrophic but not great either. he ended up adding another stage and it cleaned right up.

the IRF510 in class E is kind of sensitive to the exact load impedance too so if your output network is slightly off the harmonic performance can shift from what you calculated. anyway sounds like a cool build, post some pics if you get a chance