KWS-1 - Weird Harmonics with 80m Tracking

I have had a low drive issue with my KWS-1 for several months. Alignment and neutralization always went well except the 80 (3.9MHz) trimmer C710 which never had any affect. I decided to replace C710 with a confirmed working NOS identical Erie trimmer and also replaced the C703 padder (390pF mica) and R208 (8200 ohm).

I started alignment by again minimizing the Mixer trap and 8Mc trap. Moving on to the 80m tracking I decided to use the alternate method of pulling RT101 and V104 and injecting a 3.1MHz signal into the 250kc Inj point (J205) and monitoring the Driver plate output at J201 while adjusting the 80m inductors and cap trimmers.

I set all the inductors screwed down about mid-way and the trimmers set to half-capacitance. I found an odd double signal on the scope ( see 1st picture) at the driver plate reading at 6.2 MHz. I switched to a spectrum analyzer and the plate shows the 3.1 MHz fundamental and 2nd – 4th harmonics. The harmonics are not coming from my Rigol sig gen, I checked that carefully.

I checked the signal at the driver grid J207 with the scope and it looks the same as the driver plate. On the spectrum analyzer, the 2nd harmonic down about 20 dB.

I tried peaking all the 3.1Mc inductors and 3.9Mc trimmers (with Rigol at 3.9MHz) hoping to get rid of the harmonics but tracking adjustments peaked the 3.1 and 2nd harmonic equally.

I can trace the 2nd harmonic 6Mhz signal all the way back to the input to the mechanical filter and across the modulator.

If I slowly sweep the signal generator on both sides of 3.1MHz… the harmonics disappear below 3.07 MHZ and above 3.13 MHz.

The big surprise is if I place a jumper across L701 (as the manual says for 80m tracking) – the harmonics disappear !! See last picture.

Can someone help me understand what’s going on?