Single-Ended Pentode Redux

[Cressy Snr]

Oh bugger, It’s 1nF and 1K. That’s what’s in my cct. I’ve written it down wrong. The Zobel is from AC to AC across the bridge rectifier, which means it’s across the transformer secondary.
There is no snubber of any description either across the choke itself or from each end to ground. Maybe I’m being thick (perfectly possible) but to me spikes is spikes, so where you stop them makes no difference as long as you stop them.

Or are we saying that the flyback spikes need stopping at source ie the choke itself, to avoid punching holes in the winding insulation of said choke?

[Paul Barker]

ill have to add edits after as by time i finish editing Nick has answered long before Ive edited. Morgan says “ a single damped snubber across each secondary is all we need to suppress the swept resonance” his inappropriate values work in both high and low voltage scenarios.

[Paul Barker]

Oh bugger, It’s 1nF and 1K. That’s what’s in my cct. I’ve written it down wrong. The Zobel is from AC to AC across the bridge rectifier, which means it’s across the transformer secondary.
There is no snubber of any description either across the choke itself or from each end to ground. Maybe I’m being thick (perfectly possible) but to me spikes is spikes, so where you stop them makes no difference as long as you stop them.

Or are we saying that the flyback spikes need stopping at source ie the choke itself, to avoid punching holes in the winding insulation of said choke?

cap is either side of choke to ground, resistance is dcr.

bridge rectifiers otherwise need four zobells, but transformer secondary as youre doing is only one per secondary or two for centre tapped. And youve got the right “inappropriate” values!

[Paul Barker]

I dont think the choke snubbing need is because of the transformer leakage L diode capacitance issue. The article also shows charts of leakage inductance calculated and measured the traditional shorted secondary method. He shows the shorted secondary test is accurate enough. But he is using a Hameg lcr. My hameg is only a desk multimeter and a low distortion sign wave generator from 5hz to 50khz. fine for me though.

[Paul Barker]

I very much doubt the revenue from Morgan’s books and articles pays for his equipment. He sported a real nice new scope at Vr’s. I dont recall which it is, but he was like the cat who got the cream about it.

[Nick]

Oh bugger, It’s 1nF and 1K. That’s what’s in my cct. I’ve written it down wrong. The Zobel is from AC to AC across the bridge rectifier, which means it’s across the transformer secondary.
There is no snubber of any description either across the choke itself or from each end to ground. Maybe I’m being thick (perfectly possible) but to me spikes is spikes, so where you stop them makes no difference as long as you stop them.

Or are we saying that the flyback spikes need stopping at source ie the choke itself, to avoid punching holes in the winding insulation of said choke?

cap is either side of choke to ground, resistance is dcr.

bridge rectifiers otherwise need four zobells, but transformer secondary as youre doing is only one per secondary or two for centre tapped. And youve got the right “inappropriate” values!

So sounds like you are talking about a entirely different thing to what I thought we were talking about and what that board I sent is designed to help with.

[Paul Barker]

well that board is for the leakage inductance to diode capacitance isnt it? I wasnt intending it could work out the choke snubber.

[Paul Barker]

just a quick other quote now from Frits on phase distortion at audio transformer level. “if the frequency characteristic is practically flat ‘i.e. 20 : 20k’ the phase shift may normally be neglected.” obviously now were talking better transformers like Sowter et al. Termon was speaking in a period when 150 to 5k was the equipment he referred to.

[Nick]

Its to find the value of the resistor needed to make the Q of the transformer secondary less than 1, so to prevent ringing caused by switching. I don't think the diode capacitance comes into it directly other than it can be one mechanism where switching can cause the transformer to ring. I imagine it would also work to find the self resonant frequency of a choke.

[Nick]

just a quick other quote now from Frits on phase distortion at audio transformer level. “if the frequency characteristic is practically flat ‘i.e. 20 : 20k’ the phase shift may normally be neglected.” obviously now were talking better transformers like Sowter et al. Termon was speaking in a period when 150 to 5k was the equipment he referred to.

Ok, but you were talking about Termon.

TBH, the quote "if the frequency characteristic is practically flat ‘i.e. 20 : 20k’ the phase shift may normally be neglected.” is a bit of a tautology similar to "if a is large, 1/a will be small.

[Paul Barker]

Its to find the value of the resistor needed to make the Q of the transformer secondary less than 1, so to prevent ringing caused by switching.

Ah, I might not bother then, as Ill be valve rectified and no mosfets. otoh it would serve to clean switch mode power supplies in filament.

In my gm70 though Ive pretty much decided to go Stephie hum cancellation on AC.

But the board should find a use for some occasions. The snubber Morgan proposes for the valve amps without switching power supplies, should be fine, just add the choke snubber if its LC. Ill see if I hit a brick wall sourcing the whole BM.

[Nick]

Its to find the value of the resistor needed to make the Q of the transformer secondary less than 1, so to prevent ringing caused by switching.

Ah, I might not bother then, as Ill be valve rectified and no mosfets. otoh it would serve to clean switch mode power supplies in filament.

In my gm70 though Ive pretty much decided to go Stephie hum cancellation on AC.

But the board should find a use for some occasions. The snubber Morgan proposes for the valve amps without switching power supplies, should be fine, just add the choke snubber if its LC. Ill see if I hit a brick wall sourcing the whole BM.

Well, a valve rectifier still switches, so the snubber has a use. Its just there is no reverse conduction period. Not sure what Mosfets have to do with anything.

Whatever, if you are not going to use it find someone who will.

[Cressy Snr]

I've finally got the shit tip of a spare room redecorated and repurposed as a writing environment (God it was a mess!) so here is the schematic of the PSU.

Power Supply:

E-LinPSU.png (72.47 KiB) Viewed 1987 times

And the best simulation I could do with PSUD, given that it won't do branching power supplies:

E-Lin LCLC.png (111.66 KiB) Viewed 1987 times

[Nick]

And the best simulation I could do with PSUD, given that it won't do branching power supplies:

You could add a current tap to mimic the other channel.

[Cressy Snr]

Cheers Nick.