Date: Wed, 20 Jan 93 11:13:51 EST
From: jas@proteon.com (John A. Shriver)
Subject: Ultra-Linear output stages

The Ultra-Linear output stage appears to be an invention of David
Hafler and Herbert Keroes.  They claimed to have a patent application
on it, it would be interesting to do a patent search.  (It might also
be more revealing than some of their cloudy articles.)  At the time,
they were the Acrosound company, and were making the first
purpose-built Ultra-Linear output transformers, the TO-300 and the
TO-310.  (Hafler, of course, later left to start Dynaco, where he
started by making his own line of Ultra-Linear output transformers.)

They describe Ultra-Linear as a blend between triode and pentode
operation of beam power tubes.  In reality, it's just negative
feedback into the screen grids, which serves to linearize the output
stage.  It gives many of the advantages of pentode operation, without
losing all of the better tolerance of load impedance variation of
triodes.

The original articles were in Audio Engineering (now Audio), and
focused on their Acrosound TO-300 transformer:

1. Hafler & Keroes, "The Ultra-Linear Amplifier", Audio Engineering,
November 1951.

2. Hafler & Keroes, "Ultra-Linear Operation of the Williamson
Amplifier", Audio Engineering, June 1952.

These two articles were reprinted in the Audio Anthology Volume 2,
which has been in turn re-issued by Old Colony Sound Labs.  A third
article is also in this anthology:

3. Sarser and Sprinkle, "Gilding the Lily", Audio Engineering July
1952.

This article discusses an upgrade to the "Musician's Amplifier" (In
Audio Anthology Volume 1), which had used 807's in triode mode.  The
Peerless output transformer they were using happend to have 50% taps,
and they used these for Ultra-Linear operation.  Incidentally, they
switched to 5881 tubes.

Hafler and Keroes wrote other articles plugging their TO-300
transformer, including:

4. Hafler & Keroes, "Improving the Williamson Amplifier", Radio &
Television News, Feburary 1953.

The next article covers the smaller Acrosound TO-310 output
transformer:

5. Hafler, "Ultra-Linear Operation of 6V6 Tubes", Radio & Television
News, June 1954.

The best and most in-depth overall discussion of Ultra-Linear
operation (without quite such a product orientation) is in:

6. Ferguson, "Design for a 20-Watt High Quality Amplifier", Wireless
World, May & June 1955.  Reprinted in "High Quality Sound
Reproduction", Mullard Ltd.

This is probably hard to find, but is worth the effort.  This design
is known as the "Mullard 520" circuit.

Another English article that I suspect is useful (I've never seen it)
is:

7. Williamson & Walker, "Amplifiers and Superlatives", Wireless World,
September 1952.

I presume that this was their reaction to Hafler and Keroes.  (I think
that this is the Peter Walker of QUAD.)

An American follow-on to this Mullard stuff (which I haven't seen) is:

8. Porto, "High Fidelity Performance with Mullard's 520 Circuit",
Radio & Television News, April 1956.

Finally, for the Crowhurst touch, I can recommend:

9. Crowhurst, "Designing Your Own Power Amplifier, Part VIa: Special
Output Circuits", Audiocraft, February 1957.

That entire 8-part series is wonderful.  (If anyone has the April 1956
installment, I'm missing that issue and would greatly appreciate a
copy.)