LTP NFB question

[catalin gramada]

Hello. How can inject some global nfb into ltp please ? The circuit is excellent balanced using -100v negative into the tail. The problem is injected even some small nfb amounts like 6 db or so into the grid of non inverted section over 10 ohm shunt makes the outputs severe unbalanced. May you suggest another solution please ? Thx.

[Helmholtz]

Schematic?

[catalin gramada]

Schematic?

This is what I have now

[pdf64]

Are you sure that the feedback was negative, rather than positive?

How about adding grid stoppers, eg 10k or 22k on each?

[catalin gramada]

Is negative . Outputs of PI without nfb and with it.yell-inverted, blue -non inv

[Chuck H]

Try this
Grid loads are of equal value (I used 330k)

[Helmholtz]

Try this
Grid loads are of equal value (I used 330k)

I like the idea of equal grid loads.

But with your proposal the NFB signal would be applied equally to both grids, so both NFB and PFB. Means no feedback in effect.
The left grid leak should go to ground instead.

[Helmholtz]

Does the plate DCV change when you connect the 220R FB resistor?

[catalin gramada]

Does the plate DCV change when you connect the 220R FB resistor?

Yes it did 1.5-2Vdc around measured over the plate of the section I applied nfb

[Chuck H]

I like the idea of equal grid loads.

But with your proposal the NFB signal would be applied equally to both grids, so both NFB and PFB. Means no feedback in effect.
The left grid leak should go to ground instead.

Interestingly when I arrange it with the grid loads to ground I get no significant NFB. The way I arranged it in the redraw works perfectly in ltSPICE. 1V in = 10V out. Whereas without NFB there's clipping at over 18V.

[Helmholtz]

What happens to symmetry if you insert a grid stopper at the right side of the same value as the left side grid leak (e.g.1M).
I know this will change/increase HF response, but it's meant to test if the problem is caused by grid load asymmetry.

[Helmholtz]

Interestingly when I arrange it with the grid loads to ground I get no significant NFB.

Only the left side grid leak should go to ground.

[Chuck H]

Yes. This works well also. There was a tiny shift in phase but I didn't check it for an absolute figure. The output for both redraw circuits is identical.

[catalin gramada]

Like this ?

[Chuck H]

Like this ?

I tried that. Didn't work.

Both of my redrawn circuits work. Helmholtz has more technical prowess by a stretch so if he thinks the input grid leak should be grounded, do it that way. So, the second redraw. ta2