Single-Ended Pentode Redux

[pre65]

If you need a proof reader at any time ?

[Cressy Snr]

If you need a proof reader at any time ?

Cheers Phil, I may take you up on that in a few weeks time.

[Paul Barker]

I don't understand then, but no problem.

Beautifully designed board arrived. Do I owe you something Nick?

[Nick]

All good. I got them as I wanted one myself, so getting 5 made was the minimum order size.

Your task now is finding the parts needed to populate the board.

[Paul Barker]

All good. I got them as I wanted one myself, so getting 5 made was the minimum order size.

Your task now is finding the parts needed to populate the board.

yep

[Paul Barker]

This link was on tubediy today, Crowhurst on feedback is there somewhere. Probably right up youre street Steve, warning it’s a massive pdf!

https://worldradiohistory.com/Archive-R ... 956-12.pdf

[Nick]

That's an interesting article. Good find. Summary: the varying Ra of a stage can lead to temporally exceeding the stability criterion and introduce periods of instability.

[Paul Barker]

Oh, Crowhurst was one of the greats I have a little booklet he wrote which certainly helped me in the beginning. I was swatting some distortion in Terman the other day and in that era transformers must have been much worse than the manufacturers claims today, because all Terman charts show linear 150hz to 5khz with 6db/octave either end. probably because in those days expectations were for telephony. Parafeed is also in early Terman writings but he calls it Shunt feed. He also explains away the bass phase issues as not being significant enough to sound in same way as the 6db roll off is important to sound. All them battles between Dave and Mike were about the phase angle, yet its not that significant.

Nice though to see these patristic insight’s from their perspectives.

I have some mumetal permalloy and also some 50% nickel e i lams. When Im retred ill wind them for parafeed and see where that gets me. Yes then theres a cap, but if transformer is much better, the cost of the cap to sound is probably a price worth paying. The phase thing probably overblown by Mike and Dave who were really both endlessly arguing to defent their corner of the market.

[Nick]

yet its not that significant.

Why, because a book written in the 1940's about radio engineering says its not?

[Paul Barker]

yes, your bigger phase distortion is drivers and rooms.

[Nick]

yes, your bigger phase distortion is drivers and rooms.

I will let Mark comment on that.

[Paul Barker]

There is a 1K/10n Zobel across the mains TX secondary to keep flyback spikes out of it, which from a Linear Audio article by MJ is supposed to work a treat.

The article arrived. But 10nf isnt in the book. Where did you get that number? its either misquoted or the article is misprinted, but the article repeats the figures in 2 places, I doubt they were both a misprint. Youre out by a factor of ten. is the misprint on youre side or whats youre source? The articla is mostly with ref to 1n4007 diodes, and points to the supurfluous benefit of schottky’s, something to the effect that its insignificant due to rhe low nature of the problem that tries to solve. He says its not a problem anyway effectively. BUT of import to us, he says the values youre meant to be using Steve are also fine for the EZ81. Maybe similar for me using 5R4gy masher version. But the snubber doesnt go across the diodes, it goes across secondary or in centre tapped like I use duplicate same snubber from each winding to the Centre tap. I now think his usual input choke snubber he’s always proposed is required, since its the other side of the diode. I suspect choke snubbing would not combine well with the use of first cap to alter b+. I think it would be better either or. Id personally take the voltage and design the amp from that voltage.

[Nick]

The cap is not that big a deal, its the resistor that's doing the work.

[Paul Barker]

The cap is not that big a deal, its the resistor that's doing the work.

yes I got that in the diy audio thread, oddly Morgan’s resistor value isnt acurately calculated either he says its inappropriate but it seems to fit all situations.

Obviously when I get my board populated I can get the empirical value, and we can for Steve too. Could also see what the 1k value does.

Morgan shows traces without peak with that value. The article is hard for me to follow though.

But input chokes at Vry’s flat morgans first words to me were to implement that. Hes very sure about that.

[Nick]

Which edition of LA is the article in? I am loosing track if we are talking about diode or transformer snubbers. Or for that matter choke ones. But personally I would worry about a fix if I had reason to believe there was a problem in the first place. But that's me wanting to measure stuff as well as listen.