[Amps] pandoras parasites

John T. M. Lyles jtml@lanl.gov
Wed, 6 Nov 2002 16:38:46 -0700


Yes I opened a box when I said the 8877 amplifier by RF Plasma Prod 
didn't use anode parasitic suppressors. I used the same unit, and 
modified it down to 5 MHz from 13.56, and didn't add them either. It 
didn't need them before and doesn't now. The 3XC1500A7/8877 is a very 
short, coaxial triode, excellent construction as far as stray lead 
inductance goes. The socket leaves something to be desired, but the 
version CPI/Eimac sells has four finger stock strips to press against 
the central grid ring on the tube. For VHF operation, we built our 
own socket around the original Johnson ceramic base, years ago at 100 
MHz in commercial transmitters. The grid is well earthed, and does 
shield the anode from cathode better than a structure such as 3-500Z. 
Cathode inductance is also quite low in this tube, enabling it to run 
high power at VHF and still have good gain.

In conversation with the tube company app engineer last week, they 
confirmed what a number of us have stated already, that you must use 
suppressors for big long structures such as in the glass triodes and 
tetrodes, but that it is purely optional for the 8877.  I believe 
that the argument of feedthru capacitance is only part of the 
problem, and lead out inductance plays a role as well.

In Bob Sutherlands article in Ham Radio, Jan, 1969 "VHF/UHF effects 
in gridded tubes" (also available as Eimac Amateur service newsletter 
AS-44) it is explained that the combination of electrode-lead 
inductance AND interelectrode capacitance may cause an internal 
resonance in UHF regions. These are entirely different from the anode 
strap/plate tuning capacitor LC resonance which is spoken about so 
often here.  Couldn't this also be a possible killer of tubes, i.e., 
source of the so-called melted  'gold balls' syndrome?

Carrying this though further, a tube like 8877 with low L, would move 
this resonance further up in the spectrum, to where it is impossible 
to support it with any gain back through the tube.


just ideas about it, I am open to various interpretations, but I 
don't buy the single parasitic cure all for all tubes.
73
K5PRO


--