Using the UTC LS-33 as a Push-Pull output for 2A3's

[danFrank]

Hello Everybody,
I have a pair of UTC LS-33 transformers (125ohm,125ohm:125ohm,125ohm:voice coil) and was wanting to use them for a 2A3 push pull amp. Basically what I would be doing is puting all of the 125 ohm windings in series to get a 2000 ohm primary with center tap. Then I would wire the voice coil taps for 5 ohms and use an 8 ohm speaker load which would give me an effective primary impedance of 3000 ohms, just right for a pair of 2A3's in fixed bias.
Basically, my question is: How should I arrange the four 125 ohm windings to get the best fidelity, least leakage inductance, least capacitance, etc.? The terminal arrangements with winding resistances are as follows:

Winding # Terminals: DC Resistance:
1 1,2,4 21.6 ohms
2 3,5,6 27.6 ohms
3 7,8,10 22.9 ohms
4 9,11,12 26.0 ohms

I don't think that this transformer is bifilar wound, like the old Mcintosh unity coupled designs, but I may be wrong. But it seems to me windings # 1 and 3 are very similar, closely coupled and likewise for windings # 2 and 4.
Originally, the LS-33 was a line to line (500ohm to 500ohm) matching transformer with additional windings for voice coil outputs.
Does anybody know how the Ls-33's are wound and could offer me some guidance?
Thanks in advance... Daniel

[SY]

Honestly, I'd spend an hour or two trying the windings in their various combinations with a scope and generator to see what the optimum is, paying particular attention to phasing. If I had to guess, going outward from the core, the winding order would be 1 3 4 2. It is almost certainly not bifilar, nor would you want it to be for your use.

Do check the specs to make sure that it will handle your proposed levels. The LS series are quite beautiful and well made.

[Tom Bavis]

The 1960 UTC catalog is on my website if anybody needs it - http://www.audiophool.cjb.net/Techno.htm LS-22 is 20W, 10-40K bandwidth.

I would do as Sy suggested - try different combinations and see what happens, especially at the high frequency end. The higher any resonances occur, the better. I would also carefully measure the turns ratios in case they "adjusted" for losses - probably not, given the low winding resistance. (After all you can count on a 117 : 117V isolation transfomer to be about 1 : 1.05.)

[danFrank]

Winding #........... Terminals: ..........DC Resistance:
1 ...........................1,2,4 ...................21.6 ohms
2........................... 3,5,6................... 27.6 ohms
3 ..........................7,8,10 ..................22.9 ohms
4 ..........................9,11,12 ................26.0 ohms

Maybe this will make my original post clearer...
Daniel

[SY]

Tom, great resource, very useful. Thanks!