Lucky me, I have a pair of original Z565 OPTs -- But I have 4 ohm speakers...?

[baudouin0]

You can run out of drive with a lot of plate-to-plate feedback. Maybe 220k would be better. It also gives you an option pentode O/P as UL also does something similar. I think it may need a bit of simulation to verify. 6ej7 great but make R5, R12 a trim pot. I prefer a non fet/mosfet CCS as you can get quite a spread from one device to the next. Pentodes don't have perfect AC balance as some of the signal ends up in the screen. I found separate screen resistors + a cap between seemed to work best.

 

[baudouin0]

 

[rongon]

Thanks!
The trimpot for R5. R21 is a good idea. A 100R trimpot should do nicely, and I have some.
I like the idea of separate screen grid resistor for each 6EJ7, with a cap from screen to screen, and thanks for calculating the new R values. Yours drop right in.

Re: feedback resistors R20 and R21:

• Open loop, the circuit has about 65x gain, or 36dB.
• with R20, 21 = 220k ohms, the gain is reduced to 14.85x or 23dB, so 13dB of NFB.

Gotta punch some holes...

 

[Zung]

Interesting read

 

[rongon]

That is interesting. I guess if I'm going to use these OPTs with EL84s, fixed bias will be the way to go. I'll have to factor that in.
Or...

Go back to 6P43P (EL86) and use those with cathode bias? The concern will then become whether the 6EJ7 (actually 6J51P) pentodes can fully drive a UL 6P43P output stage with only 300V B+ to work from...

 

[martyh]

Or try back bias with el84
https://www.aikenamps.com/index.php/what-is-back-biasing

 

[ErikdeBest]

6F12P will work. This schematic will do the trick - just run it from 300V instead and perhaps reduce 10k to 2k (3 watt)
Here's my implementation of 6F12P and 6P43P as in the Little Miracle amp.
The 6P43P cathodes are around 30V and I used 1uF coupling caps instead of 0.22u from 6F12P to 6P43P.
View attachment 1101206

@kodabmx a question regarding your compound OPT wiring. If I see this right, if one connects 460VAC (2x230) across the primary, I obtain 24VAC at the secondary. That makes for a 19.2 transformation ratio, or about 2k2 primary for 6R secondary. Is that right?

 

[kodabmx]

Yes.