Home » , » Chopper Stabilization of the Vacuum Tube Op Amp

Chopper Stabilization of the Vacuum Tube Op Amp

But, even with the use of balanced dual triode input stages, drift was still a continuing problem of early vacuum tube op amps. Many users sought means to hold the input referred offset to a sub-mV level, as opposed to the tens to hundreds of mV typically encountered. The drift had two components, warm-up related, and random or longer term, both of which necessitated frequent re-zeroing of amplifiers. This problem was at least partially solved in 1949, with Edwin A. Goldberg’s invention of the chopper-stabilized op amp (Reference 48: E. A. Goldberg, "Stabilization of Wide-Band Direct-Current Amplifiers for Zero and Gain," RCA Review, June 1950, pp. 296-300. See also: "Stabilized Direct Current Amplifier, US Patent 2,684,999, filed April 28, 1949, issued July 27, 1954. (A system for lowering vacuum tube op amp offset voltage, drift, and gain errors, by means of an AC-coupled parallel-path, with synchronous rectification, and DC signal reinsertion— in short, the chopper-stabilized op amp)).
The chopper-stabilized op amp employs a second, high gain, AC-coupled amplifier. It is arranged as a side-path to the main amplifier. The chopper channel is arranged with the input signal path AC-coupled to the inverting input of the main DC-coupled amplifier, and a 60 or 400Hz switch periodically commutating to ground. The switching action chops the small DC input signal to AC, which is amplified greatly (1000 or more). The AC output of the chopper path is synchronously rectified, filtered, and applied to the main amplifier second input. In the resulting composite amplifier, main amplifier drift is reduced by a factor roughly equal to the chopper gain.
With chopper stabilization, op amps could have offset voltages stable to a few μV, and long terms drift sufficiently low that manual zeroing wasn’t required. Another key benefit was that the DC and low frequency gain was also boosted, by an amount equal to the additional gain factor provided by the chopper channel. By this means, the DC open loop gain of a chopper amplifier could easily exceed 100,000 times (100 dB). Goldberg’s amplifier of Ref. 48 for example, had a DC gain of 150,000,000, or 163dB.
But, there were some serious downsides to these early chopper amplifiers. The basic chopper architecture described above essentially "uses up" the non-inverting second op amp input of a dual triode pair, to apply the DC-offset correction signal. Thus all of the early chopper op amps operated in an inverting-only mode. In time, improved chopper architectures were developed to overcome this limitation, and the very high DC precision was made available for all modes of use.
A second limitation was the fact that the first chopping devices used were mechanical switches (vibrators). As such, they were failure-prone, often before the tubes used alongside. In time all solid-state chopping devices were to be developed, but this didn't impact vacuum tube chopper amps.
Frank Bradley and Rawley McCoy of the Reeves Instrument Corporation discussed yet another variation on the M9 op amp design, in 1952 (Reference 49: Frank R. Bradley, Rawley McCoy, "Driftless DC Amplifier," Electronics, April 1952, pp. 144-148. (A description of the Reeves A-105 Analog Computer, which used as the computing amplifier a circuit partially similar to the M9 op amp. Adds chopper stabilization for low drift)). In the Bradley-McCoy circuit, a circuit similar to the M9 topology (but with a dual triode front end) is augmented by the addition of a chopper side path. The resulting amplifier had a DC gain of 30,000,00 (150dB), and very low drift and offset voltage.
By the mid to late fifties many companies began offering solutions to analog computing using chopper-stabilized op amps. However, this wasn't true right after the chopper amplifier became available in the early 1950's— it came about later on.
Shortly after 1950 Granino and Theresa Korn published the first of their textbooks on analog computing, Electronic Analog Computers (Reference 50: Granino Korn and Theresa Korn, Electronic Analog Computers, McGraw-Hill, 1952. (A classic early work on the uses and methodology of analog computing. In Chapter 5, an op amp circuit attributed to Bell Labs and the M9 project is described, along with many other examples, and design details)). This book, along with the 2nd Edition in 1956, became the early op amp user's standard reference work. The op amp example fifth chapter of the 1st edition shows a few chopper-stabilized examples, and Goldberg's work is mentioned. By the time the 2nd edition came out in 1956, chopper amplifiers dominated the examples.
Among the circuits presented in the Korn and Korn Electronic Analog Computers 1st edition was a later version of the Bell Labs designed op amp for the M9. A distinct evolutionary path can be noted in this schematic, shown in Figure 4 below.
This version changes the input stage from a single pentode to a dual triode, with Miller-compensation added for improved stability (adjustable by the 2kΩ potentiometer). A close comparison finds that the Frost amplifier (Reference 47, again) is topologically almost identical with this M9 op amp version. And, vis-à-vis the original Swartzel design of Fig. 3, further subtle changes to be noted are different AC compensation networks.
Schematic diagram of late M9 system op amp designed at Bell Telephone Laboratories
Figure 4: Schematic diagram of late M9 system op amp designed at Bell Telephone Laboratories
next Use of the Non-Inverting Op Amp Input
Share this article :


Post a Comment

Please wait for approval of your comment .......