Best way to do grid stopper on cathodyne

[Tonegeek]

See attached. I have a cathodyne PI and have experimented with limiting drive by using Merlin's (Valve Wizard) idea. Should I put the grid stopper in before the grid leak, or after it. I notice Rob Robinette puts it after. I have done it both ways and it sounds the same to me, but does it affect the bias of the PI? Not worried about Miller capacitance. Which way is best in this case? Thanks.

[martin manning]

I have a cathodyne PI and have experimented with limiting drive by using Merlin's (Valve Wizard) idea. Should I put the grid stopper in before the grid leak, or after it?

If you put it before you will get a voltage divider effect, therefore lower gain. BW is essentially the same.

[roberto]

I would avoid the voltage divider configuration.
That's it.

[pdf64]

With the bootstrapping of the grid leak, is there really much of a potential divider?

[Tonegeek]

So should think of this as a voltage divider vs current divider? Merlin shows this mod as a voltage divider. Rob Robinette implements it as a current divider (grid stopper after the grid leak. I am leaning toward the current divider version, but do I need to be concerned at all about the effect either way has on the bias of the PI? Just trying to wrap my head around the purpose of the grid leak (besides bleeding off charge from the coupler).

[pdf64]

In the linear range, there's no current flow at the grid. So no effect on bias, no current divider effect.
The purpose of the grid leak is to provide a Vdc reference for the control grid, with respect to its cathode, and hence it enables the valves operating point to be created / defined. Without it, the Vdc at the control grid will float off, we know not where.
In case you didn't get the implication of the bootstrap reference I made, it will make the effective value of the 1M grid leak much greater, hence the signal AC voltage drop across the 470k is likely to be minimal.

[Tonegeek]

I am still a bit confused. I got on this issue after reading this : http://www.valvewizard.co.uk/cathodyne.html at the paragraph "Avoiding Unpleasant Overdrive". If I understand the trend of replies though, I should avoid putting the grid stopper before the grid leak. But, as df64 explains, what ai am doing may not make much difference anyway as there will be minimal voltage drop. BTW, Valve Wiz offers this as a solution in case the PI is over driven, so that implies things getting non-linear.

[pdf64]

Why confused? Your empirical experience is that you couldn’t hear a difference, and the theory indicates that there would only be a minimal, probably an inaudible <1dB, difference.
Whichever option you choose will almost certainly be fine

[Tonegeek]

I suppose I am over-thinking it. Thanks.