Paraphase inverter question

[vintagekiki]

Again stupid question.
Is there a ripple when the amplifier is off?
If a ripple exists when the amplifier is off, this indicates a (general) problem of grounding.

Second question is whether what you hear in your room comes from speaker or from amplifier.
At many amplifiers, very quiet buzz comes from the amplifier (not speaker), either due to the poorly mechanically fixed power transformer for the chassis, or from swirling, whirlwind (Foucault's) currents transformer core through the chassis.

[Helmholtz]

(the earth clips are not used)

Do you think this matters? I prefer to connect earth clips for audio measurements, as long as a little additional ground capacitance and ground lead inductance doesn't matter.


BTW, in his book "The Physics of the Electric Guitar" Vol. 2, Prof. M.Zollner explains the paraphase PI with and without feedback (self-balancing) and mentions different plate output impedances, low frequency phase shift and additional HF phase lag by the cascaded triodes as (potential) drawbacks.

[Dave H]

Do you think this matters? I prefer to connect earth clips for audio measurements, as long as a little additional ground capacitance doesn't matter.

I'm not sure I understand. Connect the earth clips where, to chassis? If neither point being measured is ground that's all you can do.

[catalin gramada]

This is what You,re looking for, please?

[Helmholtz]

Connect the earth clips where, to chassis?

Sure. Floating differential measurements are better done with active differential probes.

[catalin gramada]

Again stupid question.
Is there a ripple when the amplifier is off?
If a ripple exists when the amplifier is off, this indicates a (general) problem of grounding.

Second question is whether what you hear in your room comes from speaker or from amplifier.
At many amplifiers, very quiet buzz comes from the amplifier (not speaker), either due to the poorly mechanically fixed power transformer for the chassis, or from swirling, whirlwind (Foucault's) currents transformer core through the chassis.

1.No ripple with power off. 2. The noise is coming from speaker, I can feel touching how is working when power on.

[Helmholtz]

Post #79: Looks perfect.

Now please measure between grids and study the influence of increasing the 47nF coupling cap.

[catalin gramada]

Post #79: Looks perfect.

Now please measure between grids and study the influence of increasing the 47nF coupling cap.

Are You talking about interstage between first stage and paraphase ? Couplings between inverter and power grids ? Or both ? Thanks

The traces in post 79 was done without Pi coupled to the grids. Are You suggesting to connect the PI to the power grids and to measure in differential mode the influence of capacitor value in respect with the amplitude of ripple in differential mode ?

[Helmholtz]

Are You talking about interstage between first stage and paraphase ?

This. See Dave's sim results.

[Dave H]

Sure. Floating differential measurements are better done with active differential probes.

I no longer have access the those sorts of things. I have to make do with minimum test equipment and a small table in the spare bedroom (and LTSpice of course)

[Dave H]

I think now I understand why the long tail pair PI has much better bias supply ripple rejection than the floating paraphase. The floating paraphase by its very nature converts the common mode signal at the bottom of the power tube grid resistors to a differential signal at the power tube grids. This is because of the low output impedance of the virtual earth triode and the fact that its output is the inverse of the other triode. The bias supply ripple should therefore be much lower when a floating paraphase is used. This common mode to differential conversion is masked by the phase shift at 50Hz caused by the capacitor values used. If the capacitor values had been larger or the frequency higher the ripple on the output would have been much worse.

[Helmholtz]

I no longer have access the those sorts of things.

Same with me. I just wondered why you recommended to not connect earth/ground clips.

[Dave H]

Same with me. I just wondered why you recommended to not connect earth/ground clips.

I didn't want to risk one of them being connected to a PI output.

[vintagekiki]

[ATTACH=CONFIG]55093[/ATTACH] [ATTACH=CONFIG]55094[/ATTACH]

Ripple has nothing to do with the (para) phase inverter.
Ripple is about 100 mV / 50Hz, and the amplifier is without any signal from the generator.
Don't you think about ground loop?

EDIT 190909
Just remember the experience from the studio when the amplifier connects to the processor and puts it all together on the mixer.
All devices are earthed via the power cord.

Silence in space only when all earthing is at one point.

[catalin gramada]

Ok, I tied the inverter back to power grids and this is what I get in differential mode:

This is with 47nF interstage



This is with 470nF interstage



To be honest I don't see any difference