More about the Tube Operational Amplifier

Dear Visitors and Members of Radiomuseum,

It should be mentioned that still today, in the "digital age" OpAmps (=Operational Amplifiers) are indispensable, because OpAmps are at all making it possible to digitize analog signals.
For this reason OpAmps today are produced, working with frequencies up to 10 GHz (example: OPA855 from Texas Instruments). These were used as Analog frontend (AFE) to feed analog to digital converters (ADCs) enabling, for example, Digital direct conversion receivers.
Also higher order Filters are still realized with opamp circuits. Most of today's mobile phone chips are operating with fast OpAmps or their modern counterparts, the current conveyors, realizing On-Chip Filter functions.
Everywhere where (continuous) fast calculations of nonlinear relationships, consuming only small power, are necessary, OpAmps could still outperform microprocessors.
The term "Operational Amplifier" appeared not until the early 1940's, although the concept of an high gain amplifier with feedback had already been discovered.
Originally the principle of the (high gain) Signal Amplifier with feedback has been discovered by Harold S. Black [1], who 1932 filed his Patent US2102671 and so established the basis for future developments.
Harold S. Black was the first who described all benefits of the new circuit technique and also presented all relating calculations therefore.
Following Black's principles, Harry Nyquist 1932 published his generalized rule for avoiding instability in a feedback amplifier [2], and Hendrick W. Bode developed systematic techniques of design whereby one could get the most out of a specified situation and still satisfy Nyquist’s criterion.
In Black's Patent file the major designated use has been signal amplification for telephone purposes where constancy of amplification, wide frequency range and low distortion is of importance.
For the same reason the amplifier has AC interstage coupling only, in contrast to the DC-Coupling used within the WW2 “Operational Amplifier“ (OpAmp) for Fire control.
At this time the OpAmp has been only a single-ended signal path DC amplifier (with a negative input). The differential OpAmp with a positive and a negative Input came up later, when differential Input stages and Feedback were combined. The differential amplifier first came into use as a medical amplifying device often called "biological amplifier", where common mode suppression was of higher importance.
The (single ended input) Feedback Amplifier became an Operational Amplifier when the american WW2 Antiaircraft (AA-) Directors like M9 or M10 [11, 12] came into use.

The Purpose of these AA-Directors (Computer itself named M3, M3A1, M3A2 or M4A1) was to predict the future course of an airplane from tracking data, delivered from radar and/or an optical tracker so that fired shells would hit the target at the correct position despite all disturbances like wind velocity and so on.

The Input values came from either SCR584 or  SCR784 Radar and from Optical Trackers.

The Computer was build up with (single ended input) patented Feedback Amplifiers, designed by K. D. Swartzel (US2401779A), shown at the right, where several resistors were connected to the summing point (negative input), so that a summing amplifier was established with a weighting of the particular input corresponding to its input resistor value.  It should be noted that these OpAmps have Nyquist's resistive Divider for level shifting the output signal of the preceding stage to the grid of the following stage.

With the aid of these summing amplifiers the different measured tracking data like slant range, elevation, azimuth, wind speed etc.could be processed at the same time to get as a result the predicted position of the airplane.
(see picture at the right).
To represents the present position and Tp represents the future position of the airplane which is to be calculated by the AA-Director. Shown below is the working principle of the AA-Director.

 

Installing and setting up these Directors had been a quite complex procedure, all amplifiers had to be adjusted for zero offset prior to its use by the personell.
This can be seen when having a look at the Adjustment panel. (right)

In addition to the Adjustment Panel there has been a Correction Panel for computing and rejecting disturbances like wind speed, ballistic displacements, drift and so on.

Up to that time differential amplifiers had been used for other purposes (mainly medical) and had a different development path.
The first who developed a differential amplifier with common mode rejection was Alan Blumlein, as he proposed with Patent GB482,740 with a common cathode resistor in a differential stage.

But at that time the common cathode resistor was rather small and only connected to signal ground.
1937 Franklin Offner then established a useful technique that appeared in his papers [3] presenting the use of common-mode (CM) feedback to increase CM rejection.
In 1938 J. F. Toennies discussed what might be the first form of what has subsequently come to be known as the long-tailed pair [4]. In this form of differential input amplifier, the common-cathodes are returned to a high negative voltage, through a high value common resistance.

 

Also in 1938 Otto H. Schmitt discussed a long-tailed pair form of amplifier [5]. The context of his discussion was not so much aimed towards optimizing CM rejection, but rather using such a stage as a phase inverter.
For this reason Schmitt's work hasn't been recognized for a long time.

Otto H. Schmitt extended his work in 1941 [6], where one can see the structure being more like today's differential amplifiers, especially regarding circuit symmetries.

 

This approach looked very similar to the often cited OpAmp, Loebe Julie could have designed (shown further below). Schmitt sent his contribution to the "Review of Scientific Instruments" already in July 1941.

He still described his circuit (right side) as
"Cathode phase inversion amplifier, a form of self-inverting push-pull amplifier which is now finding considerable application in scientific [5] and in commercial [16, 17] apparatus.
Essentially, the cathode phase inversion stage is a conventional push-pull amplifier stage in which the cathode resistor RK, common to both tubes, has been increased until the product of its resistance with the mutual conductance of either amplifier tube is large with respect to unity, and in which the cathode return (-) is biased to a suitably high negativepotential with respect to the control grid.
Under these circumstances, a signal applied to either control grid appears almost symmetrically amplified, but in opposite phase, in the two plate circuits.“

 

Left side:
Schmitt also presented in his article within the first stage a split cathode resistance with a variable resistance in between.


With this arrangement he was able to vary the amplification factor of his differential amplifier (curves at the right side).

 

Loebe Julie designed and built a simple,
compact (3-in. by 4-in. by 7-in.), modular operational amplifier using two twin triodes similar to the 6SL7GT (a high-mu triode similar to the 12AX7), which was much faster, having a corner frequency of 1000Hz, than the previously developed amplifiers and could replace the slower M9 Opamps used until this time.[15]

The first stage was differential with balanced plate loads. George A. Philbrick, having Prof. Ragazzini as line manager, used this design when building up a bombing simulation computer.

However, Ragazzini published a paper taking most of the credit for many of the ideas in the M-9 Gun Director and explaining how HE had designed the operational amplifiers. (as Bob Pease May 1992 wrote in his article, published in electronicdesign.com [18]).

In the paper, he took most of the credit for many of the ideas in the M-9 Gun Director. Mr. Julie's name was listed among the contributors, but no proper credit was given.

In 1940 as one of the first Harold Goldberg presented a complete multi-stage, DC-coupled differential amplifier [7], again a medical application, he 1944 added a pentode (as a current sink) [8] to supply the bias current of the first stage long-tailed pair.

Left side:
1946 G. Robert Mezger offered a differential amplifier design [9] with a new method of interstage level-shift coupling, using a 12J5 (here designated as T3) triode as the bottom level shift element (which acts as constant current source). The basic circuit is showing one half of the input circuit.

The complete circuit shows but a different behaviour of the bottom level shift element which would take to much place to explain here.

Previous designs had used either a resistive level shift like Nyquist, or the more recent glow-tube technique of  Miller (shown below), still being a single ended input amplifier [14].
Picture above: Miller's DC-Amplifier, showing compensation of heater voltage variation in the first stage and differential voltage output, with selectable gain by variable feedback.

Already at the End of the 1930's when George A. Philbrick had been a Foxboro employee and was dealing with the simulation of industrial processes, he described a one-tube circuit, performing some of the fuctions of an Opamp.
While this circuit (using floating batteries for power) can’t be termed a general-purpose op amp circuit, it nevertheless demonstrated some of the OpAmp working principles [19].

In WW2 George A. Philbrick has been a member of the NDRC (National Defence Research Committee) Division 7 [10], a unit of the OSRD (Office of Scientific Research and Development) like many other scientists, where he introduced his knowledge regarding process regulation, man machine interaction and simulation.
The NDRC Division 7 developed Antiarcraft Directors, Bomb-Sights, Torpedo Directors, Guided Bombs, Aircraft-Rocket aiming devices and simulation for development purposes.
Besides many other industrial vendors like General Electric, Westinghouse, Eastman-Kodak and much more, also the Foxboro Company had a contract with the War Department [13].
When Philbrick started his own business, he developed what today is known as the first commercial OpAmp. Below is a photo of the first models produced by Philbrick. The left side shows the very first series of OpAmps, the right side shows the better known K2-W series.

Looking back to AC-coupled amplifieres with high gain i could find one with very high gain, combining the amplifier stages with current sources:

Here another contribution, showing how much amplification could be realized within a single amplifier, this time again an AC-Amplifier, but working down to very low frequencies and having overall gain to the low impedance points X-X of approximately 7.5 x 10⁵ per side and to the high impedance points Y-Y approximately 5.5 x 10⁶.

G. Wylie, a Scientist of  National Standards Laboratory, Sydney, Australia described in [20] an amplifier where a double triode, such as the miniature 12AX7, could be used to construct a single amplifier stage in which one triode section provides a high dynamic load resistance for the plate circuit of the other.
Such stages, when connected in cascade to form an amplifier of high gain, provide a number of advantages not realized in the battery-operated single stage amplifiers of somewhat similar principle described by Meissner [21] and by Schmitt [22].
Each of the first three stages of the amplifier represented by the figure, which is given as an example, has the following properties:
(a) The dynamic resistance between the points A and B is app- roximately 10 Mohms (~ R3 * mu) and only the loading effect of the input circuit of the following stage prevents the gain very closely approaching the valve amplification factor, which is about 100. Since the valve amplification factor is almost independent of operating conditions, particularly for high plate supply voltage and small bias, the gain is very stable.
(b) Of a ripple voltage applied at A, only approximately 1% appears at B (~ ripple / mu), even for a frequency of 1 c/s. This is of particular significance for the first stage.
(c) The current fluctuations at A, representing the signal, are approximately mu (=100) times less than in a conventional amplifying stage and the disturbance set up in the voltage source connected to A is correspondingly small.
The loading effect on the double triode stage due to the succeeding input circuit can be greatly reduced by driving the following stage from the low impedance point C (2 kOhms impedance as compared to 100 kOhms at point B). Through the reduction in the Miller effect, the upper limit of the frequency range (with small phase shift) is extended from a few hundred to a few thousand cycles per second.
Inter-electrode capacitances still contribute to the admittance to ground from B. The gain of the first three stages more closely approaches the cube of the valve amplification factor and the amplifier is found to remain quite stable.
By extending the very high gain to high audio frequencies, a carefully designed layout is necessary to avoid oscillation due to unwanted capacitive coupling (recommendable: first stage should be built with low microphonic triodes).

At least coming back to the term "OpAmp" the following is showing the uprising of this term:
1947 the head of the wartime op amp development work research program Prof. Ragazzini of the Columbia University of New York stated in a key paper outlining the use of and the mathematical relationships of OpAmps:
"As an amplifier so connected can perform the mathematical operations of arithmetic and calculus on the voltages applied to its input, it is hereafter termed an ‘operational amplifier’."  

More evolution steps followed then but the above described are probably the most important in the history of Operational Amplifiers.
More about the evolution of Analog Computers could be find in :
“Moving Targets“ by Simon Lavington, Springer, 2011 (British developments)
“Between Human and Machine: Feedback, Control, and Computing before Cybernetics“ by David A. Mindell, Johns Hopkins University Press 2004, (American developments)
“IEEE Annals of the History of Computing“ by IEEE, Washington (System specific documentations)

Regards
Achim

References:

[1] "Feedback Amplifiers", H. S. Black, Bell Laboratories Record, Vol. XII No. 10, June 1934, pp. 290-296

[2] "Regeneration Theory", Harry Nyquist, Bell System Technical Journal, Vol. 11 No.3, July 1932, pp. 126-147
[3] "Push Pull Resistance Coupled Amplifiers", Review of Scientific Instruments, Vol. 8, January 1937, pp. 20, 21
[4] "Differential Amplifier", J. F. Toennies, Review of Scientific Instruments, Vol. 9, March 1938, pp. 95-97

[5] "Cathode Phase Inversion", O. H. Schmitt, Journal of Scientific Instruments, Vol XV, London 1938, pp. 100,101
[6] "Cathode Phase Inversion", O. H. Schmitt, Review of Scientific Instruments, Vol. 12 No. 11, November 1941, pp. 548
[7] "A High Gain DC Amplifier for Bioelectric Recording", Harold Goldberg, Transactions AIEE, Vol. 59, January 1940, pp. 60-64
[8] "Bioelectric Research Apparatus", Harold Goldberg, Proceedings of the I.R.E., Vol 32, June 1944, pp. 330-336
[9] "A Stable Direct-Coupled Amplifier", Robert Mezger, Electronics, July 1944, pp. 106-110, 352+353

[10] Summary technical report of Division 7, NDRC, Volume 3, Washington D.C., 1946

[11] Service of Antiaircraft Directors M9, M9A1, M9A2, and M10, FM 44-38, War Dptmt Field Manual, November 1944
[12] Service of the M9 and M10 Type Antiaircraft Directors, FM 44-38, Dptmt of the Army Field Manual, August 1952
[13] Contract No. OEMsr-1144 Supplement 3, "Development of steering mechanisms for torpedoes", Foxboro Company, Foxboro, Massachusetts, Project No. 7.3-69; from DTIC Document No. AD200795: Summary Technical Report of Division 7, NDRC, "Gunfire Control"
[14] "Sensitive DC-Amplifier with AC-Operation", Stewart E. Miller, Electronics, November 1941, pp. 27-31, 106-109
[15] "Unsung hero pioneered op amp" to find in the WWW (actually: w w w.tayloredge.com/museum/museum/opamp.pdf)
[16] "A new Bioelectric Application - Electroencephalography", W.E. Rahm, Electronics, October 1931, pp. 11
[17] "Electrical Control of Galvanometer Characteristics", O. H. Schmitt, Journal of Scientific Instruments, Vol XV, London 1938, pp. 234-237
[18] "What's All This Julie Stuff, Anyhow?", published in: w w w.electronicdesign.com/technologies/analog/article/21763955/whats-all-this-julie-stuff-anyhow
[19] "Designing Industrial Controllers by Analog", by George A. Philbrick, Electronics, June 1948, pp. 108-111
[20] "The use of double triodes in high-gain low-frequency amplifiers", by G. Wylie, Journal of scientific Instruments, Vol. 31, No. 10, October 1954, pp. 382-383
[21] Meissner, E. R., Electronics, 6, 1933, pp. 195

[22] Schmitt, O. H. A., Review of Scientific Instruments, 4, 1933, pp. 661

To thank the Author because you find the post helpful or well done.