Circlotron History

What the Heck is a Circlotron?

A Circlotron is a vacuum tube power amplifier with an unusual push-pull output stage. Rather than alternately driving halves of an output transformer primary whose center-tap is connected to B+, a Circlotron drives the entire primary from a floating bridge. Originally, Circlotron was an Electro-Voice (EV) trademark, these days it seems to refer to any amplifier with a bridged output stage and floating power supplies.

The Circlotron topology provides several advantages:

The output transformer primary needs only half as many turns as a conventional push-pull design; this reduces parasitic capacitance and improves output transformer performance.

The output impedance is lower than a conventional tube amplifier. For output impedance derivations see Pascal Sternis page and Tube Lovers Anonymous. Because of the low output impedance, several modern output transformerless vacuum tube amplifiers use the Circlotron topology.

To quote from the original EV A30 data sheet:

"No output tube DC in transformer... DC output current is removed from the output transformer through the use of a bridge circuit. All switching transients are eliminated through unity coupling of output tubes; class AB1 operation is effected with the resultant improvements in power output, efficiency and tube life. The primary impedance of the output transformer is one quarter of that found in conventional amplifier output circuits, affording a vast increase in power output at extremes of the frequency spectrum."

The name Circlotron is derived from the schematic representation of the amplifier output stage, where the bridge circuit is drawn in a circle.

History

The invention of the Circlotron amplifier has generally been generally attributed to Alpha M. Wiggins of Electro-Voice, Inc., Buchanan, Michigan. Wiggins' patent application for the amplifier was filed on March 1, 1954. However, several other inventors filed patents prior to Wiggins. Cecil T. Hall of Mount Lebanon, Pennsylvania filed a patent for the Circlotron output topology on June 7, 1951. Tapio Köykkä of Finland also filed a patent for the Circlotron on September 2, 1952. Mr. Shimada of Japan also invented a very similar amplifier (CSPP) in 1952. Electro-Voice eventually licensed the Hall patent.

To further complicate the inventorship issues, patent laws were not consistent around the world. In the United States, inventorship was based on date of conception while in most other countries inventorship was based on patent filing dates.

Philips licensed the Köykkä patent and built a transformerless Circlotron to drive an 800 Ohm loudspeaker.

Mr. Köykkä published the schematic of his amplifier in a German radio magazine Funk-Technik 7/1953.

An article about the Circlotron by Dan J. Tomcik and Alpha M. Wiggins, New Amplifier has Bridge Circuit Output, was published in the November 1954 issue of Audio magazine. This advertisement appeared in the 1954 November-December issue of Music at Home.

Another article about Mr. Köykkä's amplifier was published in a 1957 issue of Radio und Fernsehen.

U.S. Patent 2,705,265, C. T. Hall, "Parallel Opposed Power Amplifiers" was filed on June 7, 1951 and granted on March 29, 1955.
Finnish Patent 27332, Tapio Köykkä, (original amplifier) was filed on September 2, 1952 and granted on November 10, 1954.
U.S. Patent 2,828,369, A. M. Wiggins, "High Fidelity Audio Amplifier" was filed on March 1, 1954 and granted on March 25, 1958.
Finnish Patent 29642, Tapio Köykkä, (improvements to the amplifier) was filed on September 30, 1955 and granted on April 10, 1958.

Electro-Voice Models

Electro-Voice sold at least eight different Circlotron models. (See advertisement and price list from 1956.)The amplifiers were manufactured by Electro-Voice's RME division in Peoria, Illinois.
Click on the model links for the owners manuals which include schematics and parts lists. (Adobe Acrobat format)

Model	Output Power	Output Tubes	Output Transformer Primary Impedance	Photo
A15	15 Watt	(2) EL84	2000 Ohm CT	A15
A15CL - Lowboy Integrated Preamp	15 Watt	(2) EL84	2000 Ohm CT	A15CL
A20	20 Watt	(2) 6V6	2000 Ohm CT	A20
A20CL - Lowboy Integrated Preamp	20 Watt	(2) 6V6	2000 Ohm CT	A20CL
A20C - Integrated Preamp	20 Watt	(2) 6V6	2000 Ohm CT	A20C
A30 (early production)	30 Watt	(2) 6BG6	1000 Ohm CT	Early A30
A30 (late production)	30 Watt	(2) 1614	1600 Ohm CT	Late A30
A50	50 Watt	(2) 6550	1200 Ohm CT	A50
A100 (early production)	100 Watt	(4) 6550	600 Ohm CT	Early A100
A100 (late production)	100 Watt	(4) 6550	600 Ohm CT	Late A100

Some of these data sheets have added notations. I'm unsure of the accuracy of this additional information. I am sure that some of the printed values are incorrect!

VOIMA Model

VOIMA sold several models of Circlotron amplifiers.

Model	Output Power	Output Tubes	Photo
A2-20	20 Watt	(2) EL84	A2-20

Philips Models

Philips also sold Circlotron amplifiers. (See advertisement and price list.)

Model	Output Power	Output Tubes	Photo
AG 9006, AG 9006 Service	20 Watt	(2) EL81	AG9006
AG 9007	60 Watt	(4) EL36	AG9007

Carad Models

Carad sold Circlotron amplifiers as well.

Model	Output Power	Output Tubes	Photo
AAS26	12 Watt	(2) EL84 per channel	AAS26
AS16	?? Watt	(2) EL34	AS16
MPAS08	?? Watt	(2) EL34	MPAS08
PAS29	12 Watt	(2) EL84	PAS29

MEL Model

MEL also sold a Circlotron amplifier.

Model	Output Power	Output Tubes	Photo
PIC-35	?? Watt	(2) EL34 per channel	PIC-35

WARNINGS!

High voltages present in these amplifiers (300V to 500V) can be deadly! Do not work on these amplifiers (or any other electronic equipment for that matter) unless you have the training and experience to avoid killing yourself, damaging equipment and/or starting a fire.

EV Recommended (Factory) Modifications

The early A20, A30, A50 and A100 models have a tendency for thermal runaway due to excessive values of resistance in the output tube control grid (G1) circuit. Note that except for the A100 these changes were not documented in the owners manuals. Symptoms of thermal runaway include: red spots on the output tube plates, blown fuses and short output tube life.

The A20 and A30 originally had 470k resistors the output tube control grid circuit (R18 and R19 for the A20 and early A30, R17 and R18 for the late A30); these should be decreased to 100k. The A50 originally had 470k resistors in its control grid circuit (almost 10 times the maximum of 50k recommended for Class AB operation!). Sometime after the start of production the A50 grid resistors (R18 and R19) were lowered to 82k. The original A100 used 470k resistors in its control grid circuit (R18 and R19). In later production later production A100s this was changed to 82k.

Other Modifications

It appears that the values of the coupling capacitors from the 12BH7 driver plates to the output tube grids were not always changed when the output tube grid resistors were changed. This reduced the low frequency stability margins and increased the possibility of motorboating. EV did not increase the capacitor values due to leakage current concerns with the paper capacitors available at the time. When using modern plastic film capacitors, leakage currents should not be an issue. Larger values may be used to optimize stability.

For improved low frequency stability, my calculated driver coupling capacitor values for the reduced value grid resistors are: 0.47uF for the A20 and A30 (C7 and C8), 0.68uF for the A50 (C7 and C8) and 1.0uF for the A100 (C11 and C12 for the early A100, C14 and C15 for the late A100). Please note that I have only been able to try the A20 and A30 modifications. To my knowledge the A50 and A100 modifications are untried.

The later versions of the A30 and A100 reduced the values of the differential driver input coupling capacitors (C5 and C6 for the early A30, C9 and C10 for the early A100) from 0.047uF to 0.022uF. This enhances low frequency stability by separating the 3dB cutoff points of the differential driver input and output coupling networks. This change should be beneficial for any of the amplifiers in the series.

The reduced value output tube G1 resistors used in later production A50 and A100 Circlotrons still exceed the 50k per tube maximum specified on the 6550 datasheet. There could be benefits to further reductions in output tube G1 resistance. (Especially with current production tubes)

I would like to hear from other EV Circlotron owners to compare component values, as most of the original documentation of production changes was lost or destroyed. Please include the amplifier serial number in any correspondence so that we may compare the relative ages of our amplifiers.

Other Circlotrons

Russian Circlotron

Links

Time Capsule: History Of Electro-Voice (written in 1953)
Katsu-San's Room
Designing and Building Circlotron Audio Amplifiers
Nostalgia Air
Tube Classics
100 amplifiers - part 1

Acknowledgments

Electro-Voice Amplifiers
Thanks to John Overley, Frank Spain and Al Watson for providing historical knowledge, assistance and datasheets. Thanks to Doug for his research.

VOIMA Amplifier
Thanks to Letku for the Köykkä patents, VOIMA information and research.

Russian Amplifiers
Thanks to Konstantin for the Russian documents.

CSPP
Thanks to Katsu-san for the information on Mr. Shimada's CSPP amplifier.

Philips Amplifiers
Thanks to Jussi, Emil, Gert and Pirat for the Philips information.

Carad Amplifiers
Thanks to Jacques for the Carad information.

MEL Amplifier
Thanks to Ulrich for the MEL information.

General Assistance
Thanks also to Michael aka Tubie for the articles and his encouragement to start this page.

Questions? Comments? Corrections? Complaints?

Email Me!