neutralization

[conisat]

Any one know any links to Neutralization method/techniques.
Is it just another form of negative feedback? or something quite different.
The only info I have been able to dig up relates to pre 60s where everything was transformer coupled, any guidance or help on on theory or application to modern cap coupled designs welcome..
Thanks

[Enzo]

What do you mean by neutralization?

[tboy]

Neutralization is a technique used in radio-frequency power amplifiers to prevent oscillation. Unless you're designing an R.F. power amp, you're probably chasing a wild goose.

[Enzo]

Oh geez, that never occurred to me. I learned my electronics on short wave gear in the 1950s. Didn;t come to mind here in this context.

[Bruce / Mission Amps]

Man the last time I neutralized an RF deck had to be over 18-20 years ago in my old Yaseu FT101E before I sold it to some other would be Ham.
Those 6JS6C sweep tubes, (big inter-electrode capacitance) did have issues... especially on the top of the 10M band.... 28-29.7MHz and would oscillate and draw current with no drive all by themselves... especially if the tubes weren't identical, not well neutralized and the tank circuit wasn't just right.
Then it was very difficult to set bias and idle current plus get resonance by dipping the plates... ha ha.. there's a term from a long long way back in another lifetime.


Here's the circuit.
Note the little 10pF ceramic (TC27) trim cap in series with C125, a 100pF 1000v ceramic cap from the plate to the grid.
That is negative feedback but in MHz not audio.

http://www.foxtango.org/ft101/graphi...%20Circuit.gif

[jrfrond]

Neutralization??? In audio gear???

This is for RF gear. The closest we get to neutralization is snubber caps and networks.

[mooreamps]

Neutralization??? In audio gear???

This is for RF gear. The closest we get to neutralization is snubber caps and networks.

Is it ???? You mean, you don't do plate neutralization on your triode gain stages ? Perhaps one could say the gain stage of a triode set at about 20 or so is not condusive to allow the amplifier to break into self-oscillations. However, I feel I must given running with the higher plate voltages and very wide bandwidths of some of these gain stages. I have only seen this just one other place, assuming these prints are correct, inside the dumble preamp.

-g

[jrfrond]

Is it ???? You mean, you don't do plate neutralization on your triode gain stages ? Perhaps one could say the gain stage of a triode set at about 20 or so is not condusive to allow the amplifier to break into self-oscillations. However, I feel I must given running with the higher plate voltages and very wide bandwidths of some of these gain stages. I have only seen this just one other place, assuming these prints are correct, inside the dumble preamp.

-g

Not unless it's absolutely necessary. I find issues like this can usually be tackled with proper wire dress and/or shielded cable. But hey, we all keep out 47 and 100pF caps handy for these reasons, right?

[Steve Conner]

Anyone designing an audio amp that needed actual neutralization (as opposed to slugging) is probably in need of medication, if not a padded cell.

Neutralization is intimately connected with Miller capacitance. In audio circuits, we just add a grid stopper resistor, and the Miller capacitance forms a low-pass filter with it. Using the usual 68k Fender grid stopper, this limits the response of a preamp triode stage to only a little outside the audio band.

If we want to amplify RF at 28MHz or whatever, using tuned tank circuits, this loss of bandwidth is obviously unacceptable. But if we remove the grid stopper resistor, feedback through the Miller capacitance from plate tank to grid tank makes the stage into an oscillator.

Neutralization cures this by making up a second RF signal, equal in magnitude to the RF voltage at the plate of the tube, but 180 degrees out of phase with it. This is fed back to the grid through a tiny capacitor equal to the tube's plate-to-grid capacitance. This capacitor is usually made variable, and needs adjusted whenever the tube is replaced: this is what Bruce would have been doing when he "neutralized his RF deck".

The result is to cancel the effect of the Miller capacitance with a kind of "anti-Miller capacitance", allowing the tube to amplify clear up to very high frequencies without instability. Which as I said to start with, is completely pointless for audio work.

Pentodes and tetrodes have much lower Miller capacitance, but above a certain frequency, even they need neutralized. We're talking hundreds of MHz by now, though.

Transistors and FETs in radio circuits need neutralized even more than triodes.

[mooreamps]

Not unless it's absolutely necessary. I find issues like this can usually be tackled with proper wire dress and/or shielded cable. But hey, we all keep out 47 and 100pF caps handy for these reasons, right?

Yes. Actually I use 100 pF for a pentode front end. For a triode front end, it's actually around th 470 pF range.