On Nov 18, 5:21*pm, Kendal Goodson wrote:
On Nov 18, 1:19*pm, exray wrote:





Kendal Goodson wrote:
Thanks Grumpy. *I'll check supply voltage next. It's not a thermal
problem. *The output is nice and stable until key down when it pulls.
It's also frequency stable during keydown, just on a different
frequency. *The oscillator isn't keyed, so I don't think it's an
oscillator startup issue.

73,
Kenn

I'm missing something. *You said that the 'driver' etc were disconnected
and the osc is not being keyed.

So what is actually being keyed?

-Bill

I see I've done a poor job of writing up the problem statement. *The
VFO at ~3.050 feeds a SA612 whose internal oscillator runs at 4.0 Mhz
and mixes the two to come up with ~7.050Mhz. *Output from the mixer
goes through a tuned bandpass filter to a class A driver and then to a
class C final. *When I found I had a problem with the VFO staying on
frequency during transmit, I disconnected the driver stage and
replaced it with a resistor termination. *The problem stayed
consistent. *The SA612 and the buffer/driver stage are both keyed.

Ahah! The SA612's input impedance will change by a lot when it is
keyed up vs keyed down.

Two ways to get around this:

1. Additional buffering between the VFO and SA612

2. Swamp the impedance change from power up to power down by adding
some parallel resistance. Should be DC-isolated from the SA612's input
because otherwise you will upset the balance.

Solution #1 requires some additional parts and current drain. I really
like the two-transistor W7EL buffer.

Solution #2 is just a capacitor and a resistor, but also assumes that
the VFO has enough drive that it won't object to being swamped.

Tim N3QE