[Amps] RF screen current

[John T. M. Lyles]

The question was:

>Perhaps some of the gurus can enlighten me. This is something which I've
>never seen referenced in any tube manufacturer's publications, but I've
>always wondered if it would ever be a potential problem and possibly
>limiting factor in VHF/UHF amplifier design.

....

>At VHF or UHF, these capacitances are such that the output tank circuit will
>invariably have a high Q. This Q will result in a high RF current
>circulating within the tank circuit with most of this circulating current
>flowing through the tube itself. Within a tetrode, it is the screen grid
>which forms part of this circuit.

>Whereas the screen grid has certain D.C. dissipation limits as per the tube
>manufacturers datasheet, I have never seen any mention of the maximum A.C.
>(RF) current ratings for the grids.

The short answer is:

Absolutely, the limitation at high power/frequency with tubes becomes 
the screen dissipation. If the transit time effects are not a limit 
first. These days this is correct assumption

No RF current ratings are given, for the reasons which Rich alluded 
to (the # parallel bars) and the fact that it would be given for only 
one cavity/circuit Q and frequency (locating the peak current 
somewhere in the screen element), whereas a J/S or Watt rating is 
universal, and can be found via analysis for whatever circuit and 
frequency that is used. One can calculate the current in the 
elements, of course, but also x voltage and get power(peak). Times 
duty factor to get average power (usually CW for ham stuff).  Power 
dissipation infers peak current somewhere in the grid or screen. 
Also, there is probably some leveling or averaging which occurs due 
to the thermal mass of the elements and how they are mounted, i.e., 
via water cooled supports, wires, bars, etc.

The reason that really big gridded tubes at VHF are made double-ended 
is to improve utilization of the limited screen grid area to make big 
RF power. Double ended circuits put the tube in the center of a half 
wave (or higher integral half waves) standing wave voltage peak, so 
that the E field interaction between elements is maxium while the 
magnetic (and current) component is located outside or near the 
outside of the tube connections. In a regular jug with handles, the 
field peak is in the top dead head space of the tube, and at the 
seals the RF voltage may be reduced, with a 1/4 (hopefully less) 
standing wave "up the tube" interaction region. If the tube gets any 
longer, then you pass 1/4 wave at the operating design freq, and 
waste a lot of the elements, not to mention putting the highest 
current right in the waist of the screen grid. If it gets bigger in 
diameter (a reasonable way to get higher power) then higher order 
circumferential modes become easy to excite, causing UHF or L band 
oscillations which our normal parasitic suppressors cannot fix. So 
the trick is double ending the ciruit with special tetrodes and 
triodes, such as THALES TH628 (formerly Thomson) or BURLE 7835.

Going to pyrolytic graphite is another trick to allow increased 
dissipation and get higher power. Beyond the scope of this 
discussion, but quite interesting technologies seem to be in new 
tubes.
73
John
K5PRO
--