Solar Powered 80m cw QRPp Tx

The Tx which was designed by Jonny Apell, SM7UCZ gives about 20mW.

A plot from my Spectrum Analyser showing the first few harmonics from the transmitter when it is correctly terminated with a 50R load. As can be seen, the 2nd harmonic is only 22dB down from the primary frequency, this is not really good enough despite the low output power of the transmitter (The peak at the left edge of the display is the zero reference marker).

I've since talked with Jonny by email and he said*Maybe the output impedance and filter don't match?
Impedance is about V*^{2}/(output power × 2)

So 3.6V × 3.6V / (0.02 × 2) = 324 ohms !! We need a transformer there!!

The same plot with the Tx unterminated. The 2nd harmonic only 18dB down.

A wider plot of the correctly terminated Tx, I don't know why the 5th harmonic is lower than expected.

A plot from my Spectrum Analyser showing the first few harmonics from the transmitter when it is correctly terminated with a 50R load. As can be seen, the 2nd harmonic is only 22dB down from the primary frequency, this is not really good enough despite the low output power of the transmitter (The peak at the left edge of the display is the zero reference marker).

I've since talked with Jonny by email and he said

So 3.6V × 3.6V / (0.02 × 2) = 324 ohms !! We need a transformer there!!

The same plot with the Tx unterminated. The 2nd harmonic only 18dB down.

A wider plot of the correctly terminated Tx, I don't know why the 5th harmonic is lower than expected.