DIY 450VA+ multi-tapped tube power trafo

[mnturner]

Not sure if this should be in the power supplies section, but I finally finished winding my first DIY transformer. I started this about a year ago but it got shelved until now.
It is intended to be the power supply for a stereo PP amp (probably using 6L6's). I still haven't decided on exactly what kind of amp I'll build with it but here are the details:
Scrapless EI, enough iron for at least 450VA
I designed for:
Two 6.3v filament windings (one with CT) able to put out about 8A each.
-150Vdc bias tap, ~250mA
-15Vdc tap
+15dc
+150Vdc, 270Vdc, 360Vdc, 400Vdc, 450Vdc, 500Vdc taps on the main winding able to put out upwards of 500mA continuous.
When finished the voltages were accurate to withing about +5% (unloaded). Filament voltages measure at 6.4Vac.

I'll post pics when the varnish has set and I've painted it so the damn thing isn't so ugly
I intend to use silicon rectifiers, so I didn't bother to CT the B+ windings.

Like I said earlier, it probably going to be going into a PP 6L6 type amp with russian 6P3S tubes, but I want to have the option of switching to other tubes such as real 6L6's, KT88, EL34, so does anyone have a suggestion as to output transformer impedances that could suit all these tubes?
I will wind the OPT's myself as well and I'd like to wind them with multiple taps with these different possible tube configurations in mind, does anyone know if doing this is likely to negatively effect the performance of the OPTs?
Any suggestions welcome.

[EssB]

What configuration, pentode, ultra-linear, triode strapped ?
To do all that means a sort of universal test OPT. If I was going to do that I would make it 50W with 5.5K primary / 4-8-16 secondary and add extra intermediate taps on the secondary. That way I could change output taps with a fixed 8R dummy load and effectively change the anode-anode load anywhere from 2.25K to 11K.

[mnturner]

Hey EssB,
I'll probably go with pentode mode possibly switchable to triode mode. I like the idea of ultra-linear but I gather that it doesn't go so well with 6L6 types or EL34's because of the low screen DC voltage rating.
I want to get somewhere around 40-50Wpc out of it; I already have bobbins for the OPT's that will hold enough iron for a good 60W, just need to buy the GOS lams and copper.
Using multiple secondary taps sounds like a pretty good idea though, that's the sort of thing I was thinking of.

[mnturner]

Here it is.

[EssB]

Originally Posted by mnturner

I like the idea of ultra-linear but I gather that it doesn't go so well with 6L6 types or EL34's because of the low screen DC voltage rating.

You'll have to look at the tube datasheets, some make an exception for the screen voltage rating in UL mode, e.g. the 1959 GE 6L6GC datasheet has a note on page 3 that allows 500VDC on the screens for UL.

Originally Posted by mnturner

I want to get somewhere around 40-50Wpc out of it; I already have bobbins for the OPT's that will hold enough iron for a good 60W, just need to buy the GOS lams and copper.
Using multiple secondary taps sounds like a pretty good idea though, that's the sort of thing I was thinking of.

KT88's will do that in UL with cathode bias and enough B+. The others may need fixed bias and/or pentode connection.
The 6P3S is lower rated, it may be 6L6-like but it's no 6L6GC. From the data I've found, 25W out seems to be possible.

I've just built up a test amp for a similar set of valves, that's the reason I answered your Q. At the moment, it's built up as a Fender circuit for testing with a musician friend: JJ 6L6GC, pentode, fixed bias, 50W 5K / 8R OPT. B+ is 490V idle biased to 40mA per side. Not optimum but I'm still getting 40W just below clipping and 50W if I push it to the guitar amp power amp rating with 5% distortion. No glowing anodes or bright screens. Mind you, the B+ sags about 50V at that power - not so much of a problem for you with your big transformer and silicon rectifiers.
When I get this musician test out of the way, I'll try some different configurations but I'll have to vary the load R to change Ra-a.
Simon.

[mnturner]

I was meaning to ask about that; i've even seen some commercial amp schematics that use 600V B+ in UL mode with 6L6GC's.
Is the important thing there a voltage breakdown of some sort or thermal damage caused by higher screen dissipation at those voltages?

With the russian tubes, I was actually thinking about AB2; I had previously designed a driver with low Zout and according to the datasheets you can get around 47W out of a pair in AB2.

Sounds good by the way, we use a Hotrod Deluxe, I love the tone of that output stage for guitar

[mnturner]

What do you think about this setup; the datasheet says in AB2 with a 3800R transformer primary, Vp of 360vdc and Vs of 270vdc you can get 47W. The maximum screen voltage is 300V, so if I wind the OPT with a separate UL winding I can set the Vp to 270Vdc and Vp at 360Vdc.
In simulation it seems to allow a little more than 30Wpc out and would allow me to disconnect the screen winding and operate in normal AB2 if I wanted to.
The 3k screen resistor is needed to prevent screen dissipation exceeding the maximum of about 2.75W on peaks.
The question is whether the screens, which are below their DC maximum voltage, would be damaged by the peaks caused by and input signal which almost reach 400V.
Obviously the plates, which are rated at 360Vdc, can survive AC peaks of at least 720V, but I don't know if screens are rated in the same way.
Any thoughts?

[Tony]

perhaps this can help you.....http://www.tubebooks.org/Books/lockhart.pdf

[mnturner]

Finally decided on a topology; I'm going to go UL with 6.6K load for EL34 or 6L6 and run the 6p3s' at lower voltage in pure pentode mode while I use them.
I have a couple in a little prototype setup right now, they glow up pretty good