Paraphase inverter question

[Dave H]

Can be introduced in simulation a build-out resistor to balance the output impedance as M Blencowe suggested, please?

http://deewm.com/Webcache/The_Valve_..._Inverter.html

I thinking 10k will be pretty much on spot

I tried it in sim on your circuit and 25k was the optimum value for the build-out resistor.

25k build-out resistor


no build-out resistor


Edit:
The above plots were with 2u2 coupling caps. The phase shift of the original 47n capacitors is significant at 50Hz. Below is a plot with 47n caps and no build-out resistor. The phase shift of the original 47u caps is almost enough to put the outputs back in phase (at 50Hz)

[Helmholtz]

This would indicate a mere 25K difference between output impedances.

[Dave H]

If you could tell me the output signal drops (from sim) at the PI plates with additional 225K AC loads, I volunteer to calculate the output impedances.

Yes, I could do that. Which schematic did you have in mind?

I calculated the output impedance of the 5D5 second triode in sim and it is 12k as Zollner stated.

[Helmholtz]

Which schematic did you have in mind?

Catalin's circuit.

[Helmholtz]

The build-out resistor will somewhat decrease the non-inverting PI output and require adjustment of the balancing resistors for perfect signal symmetry at the power tube grids.

[catalin gramada]

The build-out resistor will somewhat decrease the non-inverting PI output and require adjustment of the balancing resistors for perfect signal symmetry at the power tube grids.

Sure, I.ll need to do the amplification of paraphase more than unity by a factor. That meant less local feedback. In consequence the feedback resistor should be bigger, That is correct ? I used 470k + 33k in first instance as I found perfect balanced amplitude by measurements for the tube I used. Should be little bigger to compensate the attenuation by 22k/ 220k divider. Not sure if I interpreted correctly but I can do that. Thanks

[vintagekiki]

Perfect signal symmetry at the power tube grids, implies matched power tubes.

We have relatively not paired power tubes.
For the same idle current, bias voltage of one power tube is for example -40V and bias voltage of other power tube is for example -50V. Signal voltage 10VAC is provided on both grids power tubes.
What happens with signal when power tubes using a different bias voltage for the same idle current?

[pdf64]

It’s worth noting that catalin gramada’s circuit looks to be using a 12AT7.

[catalin gramada]

It’s worth noting that catalin gramada’s circuit looks to be using a 12AT7.

That is correct

[Helmholtz]

The phase shift of the original 47u caps is almost enough to put the outputs back in phase (at 50Hz)

You surely meant 47nF.

What do you get with 47nF coupling caps and a build-out resistor?
Still wondering if more than different output impedances is involved.

[Dave H]

If you could tell me the output signal drops (from sim) at the PI plates with additional 225K AC loads, I volunteer to calculate the output impedances.

220k AC load
Left triode 2.62V, Right triode 2.53V

110k AC load
Left triode 2.256V, Right triode 2.173V

Edit:
220k Left triode, 110k Right triode
Left triode 2.62V, Right triode 2.248V

Schematic

[Dave H]

You surely meant 47nF

I surely did

[Dave H]

What do you get with 47nF coupling caps and a build-out resistor?

47n, 25k build-out

[Dave H]

We have relatively not paired power tubes.
For the same idle current, bias voltage of one power tube is for example -40V and bias voltage of other power tube is for example -50V. Signal voltage 10VAC is provided on both grids power tubes.
What happens with signal when power tubes using a different bias voltage for the same idle current?

I'm not sure. Asymmetrical clipping, increased distortion before clipping, reduced output power?

[Helmholtz]

47n, 25k build-out
[ATTACH=CONFIG]55158[/ATTACH]

Great, that's what I was hoping to see. Thanks.
So large value coupling caps are not needed. It seems that all that is necessesary is equal load impedances as seen from the grids.