[Amps] Non-linear linear

[Manfred Mornhinweg]

Paul,

your problem sounds, looks, smells, tastes and feels like a 
self-oscillation happening only when the bandswitch is in the 15/17m 
positions. So I guess it actually IS a self-oscillation! ;-)

> Shared PI input coil for 10/12 and 15/17 but with padder caps 
> switched in for the lower freq.

So basically you have a tuned output, tuned input oscillator. Just a 
tiny amount of feedback is enough to make such an amplifier oscillate.

> One anomaly I notice is that at the lower power input on 15 the plate 
> current is about 390mA, where as on 80/40/20/12/10 its about 300mA.

That too indicates that the beats is oscillating.

Some possible cures for that kind of problem:

- Better shielding between input and output tanks, if it isn't very good 
already. If you use a combined bandswitch for both, then you need 
shielding between the wafers.

- Improve the grid grounding, if possible.

- Watch where the tank circuits inject the ground return currents into 
the chassis. Don't forget that even thick sheet metal isn't a perfect, 
inductance-free conductor. The tank circuits can couple through the ground.

- Watch where the supply, filament and bias bypass caps connect to ground.

- If the circuit is just at the limit of oscillating, often you can 
stabilize it by adding resistive loading to the input circuit, without 
eating up too much of your drive power.

- Check your RF switching (relay) setup, and make sure that there isn't 
a lot of feedback getting through it.

Manfred


========================
Visit my hobby homepage!
http://ludens.cl
========================