[Amps] Reducing grid currents and improving gain

[John Lyles]

A little tube history here:
RCA designed and manufactured beam power tetrodes such as the 6L6, 807, 
and 829B to increase power output, allowing the plate voltage swing to 
go lower without a large increase in screen current during the minimum.
Without using a suppressor grid (as in a pentode), the beam was 
'squeezed' (converged by focusing) where it passed through the gaps 
between screen grid wires, and so intercept was reduced. In addition, a 
space charge region was established out front of the screen between it 
and plate, to minimize the return of secondary electrons from the plate. 
This was called a virtual cathode, and it returned secondaries to the 
plate. A beam power tube would have lower screen current than a pentode. 
They did this by adjusting the pitch between grid wires, the alignment 
and spacing between grids and the other two electrodes, and the voltages 
applied.

Beam power tubes were a revolutionary step from earlier tetrodes with 
electrons from cathode that sprayed out through the grids, with no 
particular means of alignment. In 1955, RCA extended this concept to 
high power with tubes like the 6448 and later the 4616 tetrodes. In 
these, unit tetrodes were made with a filament rod or bar, enclosed on 
three sides by a copper frame. The edges of this U-shaped frame were the 
control grid. Carefully aligned was screen grid on the outer side 
following the same principles of the smaller glass tubes. Similarly, in 
triodes RCA built electron-optics using beam forming walls around each 
cathode bar, and shielded the control grid wires or bars from direct 
electrons. Carrying it further, they build shielded-tee structures that 
overlapped the grid bar, and reduced capacitance from it to plate. The 
shield acted like a zero-voltage screen. Big tubes like the 6949 made 
this way could produce 1/2 megawatt of CW power above HF.

RCA also developed radial cathodes that had alternating dead and active 
emitter bands, to emit beams in strips. By placing the grid wires in the 
regions without beam, it would greatly reduce grid current and emission 
due to bombardment from the cathode beam.

So what did Eimac do to improve their product? They developed a line of 
focused triodes and tetrodes. Using similar striped cathodes like RCA, 
they focused beams to avoid interception, This not only lowered grid 
current but also reduced IMD. Examples include the 4CX600J and 4CW800B 
beam tetrodes, and a series of focused triodes including the 8874, 
3CX800A7 and U7, 3CX1500A7/8877, 3CX4000U7, 8938, 3CX1500U7, 8962, 8963, 
3CX5000A7 and U7, and so forth up to the 3CX15,000B7. One of the big 
problems with oxide cathode tubes was the deposition of barium on the 
close-by grid, leading to loss of beam control as the grid became an 
emitter. Coatings were applied to the grids, permitting higher 
dissipation without excess grid emission. Sputtered carbon-titanium 
coatings called "Pyrogrid" were one example. Gold was a more expensive way.

Both companies, and I'm sure others as well, followed similar lines of 
development to produce excellent tubes that allowed power levels to 
rapidly increase to hundreds of kilowatts without grid destruction.

73
John K5PRO