The AM Forum

Hello guys.
I'm building an amplifer with 2 GU81M russian pentodes,
and I'm planning the tank circuit. The GU81M's has very high output capacitance, thus, I would like to use a roller coil in conjuction with a vacuum cap with low min capacitance  to be able to tune the higher bands. Can get hold of very heavy duty russian 3.7uH roller coils, and also 5-250pf vacuum caps.
On my last amp (2x813) I had some issues with the plate choke, so this time, I want to try another approach, that being to feed DC at the 50ohm reference point in the tank circuit instead of at the anodes.

I have seen schemes of it being done succesfully, by inserting a teflon insulated magnet wire, through the tank coil, keeping the dc blocking cap at the normal place,  and getting by using a small choke at the feedpoint.

If I had used other tubes, I would definately have done just that, and wound the tank coils using brake line for the low band coil, and something larger for the hiband coil, with a wire inserted before winding.

Now I'm thinking about still using the coaxial feed for the lowband coil using brake line, and putting the 3.7 uh roller coil and vacuum cap at DC potential, with a blocking cap between the roller coil and the low band coil, see attached picture.

I will be housing this in a large compartment, capacitors and rollers will be motor driven.

Will the dc component on the vvc be of any trouble?


Here it is described by john fielding, scroll down to 'A better anode feed method'

https://ogrik2.ru/b/john-fielding/valve-amplifiers-explained/30339/8-designing-an-anode-tank-circuit/9

Also included a sketch of how SM6APQ have done on multiple amps,using copper pipe.

Answer to myself.
Dc offest is added to the cap and tank, so its voltage rating needs to be higher.
Should be safe with a 10kv rated C1.

Brake tubing here in the US is plated steel, not advised for RF work.

DOT regs specifically say not to use copper as to the low burst rating  in automotive service.

Those GU-81's are marvelous tubes for the $.  The sockets are more than the tubes!

73DG

The TMC GPT-10 has a hot PI coil. As a result the RF choke is simple. The diagram of that QRO RF output stage is in alignment with the cited Fielding article.

(the attacment shows a triode and 4500V plate supply.
The original uses a 4CX5000 and 7500VDC, and a different cathode input tank.
All the plate output parts are original values and voltage ratings.)

I don't have a problem with hot tank coils. They were good enough for Grandpa.

Refrigeration tubing is copper and available in flexible type, in suitable cross-sectional
diameters.

Very good idea to NOT have DC on the antenna line...

The T-368 also has a hot Pi-L network in its design. (does not cover up to 10m, etc...)

                  _-_-

Use refrigeration copper line. Sold in 25 and 50 ft lengths. Don't use brake line. No ferrous material should be used in RF, not even stainless. ST

Yes, I know, have replaced all lines on my 58 plymouth with US correct steel line. Nice to work with,and fairly easy to get good looking bends.
Here in Sweden, we do have CuNi brake lines available.
Have used it in link tuners and such with good results.

Works fine on my homebrew amp!
The Collins tank circuit runs @ B+.

Thanks for the encouragement.
Will float the tank at dc, with a coupling cap between tank coil and C2 to be able to use a smaller spaced variable. C1 will be a 10kv russian vacuum cap. Taps on the tank will be managed by vacuum relays (once again russian stuff).