Tom said:
"This requires a network close to the tube with a low impedance
presented to the tube at high frequencies (a "T " won't do the job,
and neither will a network some distance from the amplifier)."
In a new design, the pi-net tuned input can be right next to the cathode
circuit. But, when working with amplifiers that were built years ago, one
must live with the existing limitations.
In the Swan Mark II, it was easy to add a new wafer and mount the relays and
pi-network under deck - right next to the cathode where it should be.
But, when I modified my Alpha 76 - which has no tuned input to use 3CX800A7
tubes, I studied the mechanical layout for hours trying to include a tuned
input switched by the existing band switch. I could not figure out how to
do it, since it is very tight inside and the bandswitch uses a different
degree of rotation between contacts than any others I had. The Alpha 76
normally uses a toroid input to match inpedances, but that would not have
any of the benefits of a tuned input. After much internal conflict, I
finally decided to abandon the tuned input on the 76. It works fine and is
clean enough.
Using a remotely switch pi-network at the rear input connection of the
amplifier is not ideal but it is much better than none at all. In fact, the
overall length of the coaxial line between the tuned circuit and the cathode
may not be much different than the length of some commercial amps. As I
recall, my old HT-41 has just a few inches between the input connector and
the cathode, so a remotely switch pi-network at the input connector would
add significant improvement and would be the only reasonable approach to try
to improve the old battle-axe.
73, Colin K7FM
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|