[Amps] tetrode push-pull amp schematics

[John Lyles]

I think this thread is old, but have been busy commissioning a new HPA, 
another pair of Diacrodes (tetrodes) at 200 MHz, class B. Its working 
now, has consumed me since February, the installation and turn on.

In 1998 my associate and I designed and built a push-pull tetrode 
amplifier for 2.8 MHz fixed frequency. It used a pair of TH555A tetrodes 
from France, each rated for about 200 kW. It runs pulsed, 20 pulses per 
second, each pulse about a millisecond. It is in cutoff the rest of the 
time with approx -700 volts of grid bias, and during pulses, it is class 
A. Reason for this becomes evident below. Screen and plate voltage stay 
on all the time, 1300 on the screen and 15 kV on the plate. Each tube 
conducts beam 360 deg of the RF cycle when pulsed on. A push-pull tank 
circuit was not necessary as this is the IPA for a more powerful pair of 
triodes, EEV BW1643J2. They are rated 450 kW each, and run as cathode 
followers to have extremely low output Z. The grid to cathode space is 
tickled by the IPA drive. I had separate pi networks on each tetrode 
output, as they drive the triodes through independent 500 ohm coaxial 
feeders that I designed using a small brass rod for center in a 3 1/8 
inch outer conductor. The pi net was only needed to provide some phasing 
control without impedance transformation, to account for various strays 
and component mechanical variations. Push pull output of the cathodes 
from triodes was combined into a single cavity with a stack of of 0.5 
meter diameter ferrite rings that were water cooled, and provide the L. 
A series of parallel 100 pf vacuum caps resonates this along with having 
a DC bias coil to tune the 'cavity'. In the center of this contraption, 
lies a proton beam line, with a ceramic gap. Its purpose is to capture 
the fast protons as they go around a storage ring and bunch them in 
clusters.

Why class A? When the protons bunch up, they accumulate a very large 
charge and the current across the cavity is in tens of amps. This 
current (the image of it in the pipe) must bridge through the push pull 
amplifier, and pass without getting cutoff. Class A throughout ensured 
that the tubes conduct 100% through the RF cycle, despite what flywheel 
effect of the cavity exists. (pretty low Q actually with all that 
ferrite). And since it is pulsed, the power consumption isn't large, a 
few amps of DC plate current per tube with about 40 amps peak plate 
current. Running big shortwave tubes in class A was fun, had to pay 
attention to parasites and has a lot of ferrite around them for stoppers.

This thing is still running well, although the physics is getting old 
and maybe we gotta do a new design for different bunches, meaning low 
VHF power. Another challenge, and there you are. Nothing wrong with 
class A except it isn't needed in most cases of radio...

73
John
K5PRO