SE Output Tranny - Dumb Question

[sampleaccurate]

Can someone explain why this wouldn't work?

If a non-gapped transformer intended for single ended use were equipped with a third (tertiary) winding in addition to the normal primary and secondary, why couldn't a DC current in this winding be used to cancel out the magnetic field created by the bias current in the primary, thereby virtually eliminating core saturation due to DC current flow and increasing the output capability of the transformer. The only disadvantage I can see is some heat generation and the need for another DC power source.

What am I missing here? I just need somebody who knows more about this to point out where I'm going wrong. It seems like a much less expensive solution than increasing the core size and gapping and I'm sure it would be done if it worked.

Thanks.

[Yvesm]

Because the DC source will short the audio signal apearing on the tertiary winding.
Unless you make the DC source such that it has an hi impedance for audio frequencies.
For exemple applying it thru a large inductor with plenty of iron inside

Yves.

[hemgjord]

It will work.
There are as you say a few drawbacks. you need the sam number of ampere turns as in the primary which will devote approx the same core widow area for compensation winding as for the primary.
You will loose primary inductance, increase leakage inductance and waste some extra power for a equal size core.

/Olof

Edit:
OK, thanks Yves I missed the the most severe point

[sampleaccurate]

Originally Posted by Yvesm

Because the DC source will short the audio signal apearing on the tertiary winding.
Unless you make the DC source such that it has an hi impedance for audio frequencies.
For exemple applying it thru a large inductor with plenty of iron inside

Yves.

The tube plate would see a low impedance reflected through the tranny by the DC coil - that makes sense.

Great answer. It seems obvious now. Thanks.

I guess there's no free lunch.

[Arnulf]

Alas, no free lunch. But there is an area (one I have yet to explore): parafeed connection. Solid state CCS isn't prohibitively expensive and neither is coupling capacitor, at least compared to the price of new, true SE, OPT.

[Steerpike]

Exactly what you describe was used in some Telefunken table radios. The compensation winding was used as a DC filter choke with a large electrolytic cap to make up an LC filter, from which the overall B+ and screen supply for the amplifier was derived.

When one such transformer failed, I suceeded in using a PP/CT transformer in its stead. Ultra high fidelity isn't a design feature in a shortwave table radio.

With a single-ended ungapped OPT, there is also the option of using the Philips' "series balanced output" configuration:
http://people.cs.uu.nl/gerard/FotoAl.../SerBalOut.htm

[DougL]

Originally Posted by Arnulf

Alas, no free lunch. But there is an area (one I have yet to explore): parafeed connection. Solid state CCS isn't prohibitively expensive and neither is coupling capacitor, at least compared to the price of new, true SE, OPT.

I disagree. Since the CSS needs twice as much voltage for the same output on a transformer, you are basically pushing the cost into the power supply.
Not saying it isn't valid, but its probably not a big cost saving.

YMMV

Doug

[Yvesm]

As Olof stated, it's DC amperes per turn that "premagnetize" the iron.

So you must use as few turns as possible on a large core with minimal gap to maintain a decent primary inductance
Incidently, this improves hi frequency response by reducing the leakage inductance wich increases as the square of the number of turns.