Inspired by a previous post, thought id run some basic tests on PCBwizards encoder!
For anyone who doesnt know? it is based on the notorious BH1417F fm stereo transmitter IC!
For this encoder, the RF side is unused but could still be contributing problems? who knows?? Who cares!! lol
The PCB!
The 19khz squarey waveform!
The pilot in a forest of spikes!

Just out of curiosity, have you put a known, fairly high resolution square wave into your PicoScope in FFT mode? Those peaks in the V/f plot don't tally up with the trace in the V/t plot. For example, you would expect to see the 41 kHz signal at about -38 dB on the V/t plot (on the oscilloscope). Alright, it's smaller than the 19 kHz pilot, but it's not THAT much smaller.

Also, have you just measured this just connecting it straight to your PicoScope? If so, the input impedance is very high - 1 MΩ. Have you taken any measurements with a lower impedance load on the output? A quick look at a Chinese datasheet would suggest trying around 10 k. MPX is on pin 5.

Thats a good question Mr pranker!
I havent done that lol but the picoscope does have a built in function generator that kicks out some nice waves!
Il give that a go and post the result here!

Here is a 1khz test on the picoscope! direct connection between generator and scope input squarewave!

Here is a 1khz test on the picoscope direct connection between generator and scope input sinewave!

Yes Mr pranker!
it is as you are saying unloaded (scope impedance) i should imagine the encoder would perform better in a loaded state? but TBH the circuit lacks the filtering networks that ive seen on other encoders, good expectations may be a bit premature? lol