so ive been building a little 40m CW transmitter from scratch, nothing fancy just a crystal controlled thing based loosely on a design i found in an old ARRL handbook from like the 80s. got it oscillating fine and im getting maybe 2-3 watts out which is honestly more than i expected for a first attempt.

the problem is when i hooked up my cheap SDR to poke around the spectrum a bit i could see what looks like harmonics showing up pretty prominently, one around 14mhz and another fainter one around 21. i figured some harmonics were inevitable but these seem kind of strong? the 14mhz one is maybe 30db down from fundamental which i guess is okayish but im not totally sure what's acceptable here or what i should be looking at to bring that down.

i threw a lowpass filter together on a piece of ugly construction copper clad but im not sure i got the component values right, used what i had in the junk box which was close but not exact. should i just buckle down and wind the right inductors or is there something else i could be missing in the transmitter stage itself that would cause stronger harmonics than expected

30dB down on the second harmonic is actually not terrible for a raw oscillator stage but yeah you definitely want more attenuation before you put that on the air, FCC wants 43dB down on harmonics for anything under 5 watts i believe, could be slightly off on that number but its in that ballpark.

the filter is almost certainly your issue if you cobbled it together with random junk box values. the cutoff frequency and impedance matching matters a lot more than people think with lowpass filters for TX, if the impedance is off the attenuation curve shifts and you can end up with less suppression than the math says you should get. id wind the inductors properly, doesnt take that long honestly and makes a big difference. toroids from the right mix for 40m, probably 2 or T50-2 cores depending on what you can get. also check that your PA stage isnt running into compression, a transistor being pushed too hard will generate way more harmonics than one operating in its linear region.

what transistor are you using in the final? i built something similar a while back and had harmonic problems that turned out to be because i was using a 2N2222 way outside its comfortable operating range for that frequency. switched to a proper RF transistor and things cleaned up noticeably even before i got the filter sorted out.