Topband: Neutralizing your 833As

Mike Waters mikewate at
Sun Jan 17 20:02:09 EST 2016

Hi Dave,

Thanks for the advice! What you describe below is an improved version of
what I had planned to do.

I'm likely over-thinking this. Much ado about nothing, as the old saying
goes. I've been accused of doing that before on this amp project (and he
was right. :-)  Let me elaborate a little on what my concern is.

I don't think I ever drew a schematic of the PA grid and plate circuits,
but I'll try and describe the important details, leaving out things like
the grid biasing arrangement, RF chokes, DC blocking capacitors, the TO-220
swamping resistor string, circuit protection, etc.).

1. There is no tuned grid circuit. (If there was, it would be simple to
2. There's a 200 ohm 100w non-inductive swamping resistor between the grid
and ground.
3. From the grid to ground is the 200 ohm secondary of the input
4. The input is the 50 ohm primary of that transformer.

Here's the idea I had in mind, good or bad: Add a tertiary (third) winding
to that input matching transformer --grounded at one end-- to obtain the
voltage for neutralizing. The other end would be in series with an air
variable (already have that)* and a fixed capacitor connected directly to
the anodes.

Maybe I'm over-thinking this, but what I expect is too much or too little
feedback voltage

*Photos of the neut cap is at;O=D
neut-cap_5445.JPG and neut-cap_5444.JPG

The reason for the fixed capacitor between the anodes and the variable
capacitor I mentioned was to reduce the chances of the air variable arcing
over, which would be catastrophic. That concept is

73, Mike

On Sun, Jan 17, 2016 at 12:20 PM, Dave Olean <k1whs at> wrote:

> Hi Mike
>    You need a signal generator and a crystal detector to sniff in that
> amp. Feed a low level signal in with the filaments on but HV off and see
> how much RF leaks thru to the output connector, then mess around with
> neutralization schemes until the RF drops way down. I do that all the time
> and it works great. I have neutralized a six meter KW with a 4CX1000A, a
> bunch of 4CX 250B amps and driver amps, etc. It never failed. Shoot for 10
> dB or more attenuation than the tube can produce and it should work FB.
> The nice part is no high voltage to worry about. You can get the
> neutralizing wire or the cap set for a perfect null at your frequency
> pretty fast. I used to use a signal generator wiyth an HP 415E VSWR
> indicator. Put 1000 Hz audio modulation on the signal generator and you can
> read attenuation right off the meter on the 415.. Slick!  Now I have two
> port a vector network analyzer, so the HP 415 sits on the shelf, but either
> method is quick.
> [snip]
> Maybe you would let me pick your brain sometime about my swamped-grid
>> dual-833C legal limit amp. It works fine on 160 and 80, but it needs
>> neutralizing if it's going to be stable on 40 (and work at all on 20). I
>> have a plan to neutralize it using a tertiary winding on the input
>> transformer, but I foresee a lot of trial and error. But I don't have time
>> to work on that right now.
>> I wasn't even sure if two of those triodes in parallel would even work,
>> so I didn't spend a lot of time making it look pretty.

More information about the Topband mailing list