[Amps] RF Chokes

[w1itt at att.net]

Eimac, in their book "Care and Feeding of Power Grid Tubes", available as a free download from their site, offers as a rule of thumb that to limit current in the plate RF choke to one percent in a Pi-net amp, the choke inductance should be 100 times that of the Pi-net coil.  Few amplifiers  meet this spec, because series resonances arise in typical solenoid wound chokes long before the required inductance is achieved.  The use of permeability tuned chokes is not new, having been implemented in the commercial and amateur  worlds.  Pundits are quick to point out that while ferrite loaded chokes look good on the test bench,they will go into core saturation when DC is passed through them. Through judicious choice of core materials and attention to physical dimensions, it is possible to build chokes that provide high inductances without series resonances through the HF range.
I have a network analyzer (HP/Agilent 8714C) with which I make my living, and I've been able to develop a 965 microhenry choke, without resonances between 1 and 50 mhz, which will handle 3 amps of DC without showing any change in inductance. ( I ran out of bench supply at 3 amps, but that should satisfy most amateur amplifiers.)  The 965  microhenry choke does not strictly satisfy the one-percent requirement in some 80 and 160 meter amps, depending on plate load impedance, but it's   an order of magnitude better than what most of us use now, and it minimizes the need to wory about current handling capability of the bypass capacitor at the bottom.  In retrofits, it may be necessary to tweak the coil taps to uncompensate for the reactance swings that occur just above and below the bad old choke resonances.
I manufacture 965 microhenry chokes about 6 inches long, as well as an 11 inch 2150 microhenry choke for the 160 meter purist, both without resonances from 1 to 50 mhz. Anyone interested may contact me off-list at w1itt at att.net.

73
Norm W1ITT