[AMPS] Blown TL922A... What to do?

[Tom Rauch]

> Question: How relevant is the percentage delta when the Q is ~ 2.0?  In
> other words, is this percentage change at ~ 2.0 as important as a say a
> 46% change in Q from 50 to 73?
> 
> -Paul, W9AC
 
If we can maintain a "low-noise" conversation, you can decide for 
yourself.

First, we should understand this. The suppressor Wes measured 
was NOT the suppressor Rich sells. The suppressor Wes 
measured was a duplicate of the stock Ameritron suppressor with 
nichrome wire.

Wes did that to eliminate all variables and only consider the 
change in wire. The stock suppressor Rich sells actually produces 
high VHF Q in the anode system of the AL80A series of PA. It 
does NOT reduce VHF Q. It reduces ten meter Q, and 15 meter Q, 
but not VHF Q.

There is a good technical reason for that, and it's easy to see. But 
first more to your point...

Q of the suppressor, by itself, is not a direct factor in parasitic 
suppression. The suppressor is part of a distributed system of 
wires and components that have a complex impedance (resistance 
and reactance). The suppressor adds another impedance in series 
with this system (not a "Q" value), and the complex impedance 
(resistance and reactance) of the suppressor modifies the anode 
system impedance.

What you want to do is add enough resistance to stabilize the PA 
(if it needs it) at the frequency where the tube oscillates. In order to 
make s significant change the impedance of the suppressor has to 
be large compared to the impedance of the path it is in.

Consider the simple case where the suppressor a pure resistance. 
If the anode circuit presents a series impedance of 10 -20j ohms, 
the addition of 100 ohms of additional ESR (equiv series 
resistance) will greatly change anode system impedance and Q. Q 
of the anode system will drop from 2 to about 0.2.

If the added resistance was 1 ohm, there would hardly be an anode 
Q change!

In both cases, Q of the "suppressor" was zero.

You have to know what the impedance of the anode system is, and 
what you want to do, before you can decide if one suppressor is 
better than another.

There is a procedure for doing that. Most hams don't do that, but if 
thousands of units are manufactured it is a good idea to look at 
little things like this.

If you remove the stock suppressor in an AL80A or B, and replace 
it with a hairpin of nichrome, the VHF anode Q of the suppressor 
where the amplifier has a tendency to oscillate increase 
dramatically. That isn't normally a problem, because the amplifier 
barely oscillates up there anyway even when the suppressor is 
totally removed. Rich's suppressor doesn't really do any harm to 
stability, even though it increases VHF Q slightly. The only ill effect 
is it reduces tank Q, mostly on ten meters.

His suppressors are a partial solution in amps like the Clipperton, 
because they can be made to repeatedly oscillate near ten meters. 
In that case, the low ten meter Q helps stabilize the PA.

De-Q'ing the HF tank also helps reduce arcing in the tank, because 
it reduces peak voltages. That can help if you mistune the PA, or 
have a load fault.

So the Q of the suppressor, by itself, is a meaningless parameter. 
The actual important parameter is the impedance of the suppressor 
and how it affects the impedance of the tank.

That's why if you have an unstable amplifier with long thin anode 
leads, the suppressor must have more turns with a higher 
resistance in the suppressor. The shorter the anode leads and the 
wider, and the lower the tuning capacitor impedance at VHF, the 
lower the inductance and resistance in the suppressor can be for 
the same anode system dampening.

You want the parallel equivalent impedance of the suppressor to be 
large compared to the anode system impedance at the point where 
the suppressor is added.

Is that easy to follow, or not?

   
73, Tom W8JI
w8ji@contesting.com

--
FAQ on WWW:               http://www.contesting.com/FAQ/amps
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com