Harold Black and the
Negative-Feedback Amplifier
Ronald Kline
0
n August 2 , 1927, Harold
Black, a young Bell Labs
engineer just six years out of
college, invented the negative
feedback amplifier in a “flash
of insight” while riding the
Lackawanna Ferry across the
Hudson River on his way to
work. (Bell Labs was then located on West Street in Manhattan.) Black recalled, “I felt
an urge to write but had nothing to write on so picked up my
moming paper. By sheer coincidence one page was blank.
Here was a perfect set-up, lots
of room and fully dated. With
this to implement my job, I
started the first written record
pertaining to the stabilized negative feedback amplifier. Years of
study and many failures preceded this sudden conception of
stabilized feedback. Despite immediate recognition of its importance, years of additional
work were required before it
found substantial commercial
use” [ 1, p. 7231.
Negative feedback became
widespread [2]. It allowed the
Bell system to reduce overcrowding of lines and extend
its long-distance network by
with network theory from the
beginning. Like Maxwell’s circuit equations, which were dev e l o p e d in t h e 1860s t o
understand the induction coil,
negative feedback was not tied
to the technology of its origins
(vacuum tubes) and became a
fundamental principle of electrical engineering with innumerable
applications
independent of the original
hardware used for its invention.
Invention
The joumey to the Lackawanna Ferry began in 192 1 ,
when Harold Black (18981983) g r a d u a t e d f r o m
Worcester Polytechnic Institute with a Bachelor’s degree
in electrical engineering and
took a job with the Engineering Department at Western
Electric. (The research branch
of this department formed the
nucleus of Bell Labs when it
was established in 1925.) A
pressing problem in the Bell
system at the time was that the
One of the newspaper pages used by Black to jot down his
distortion and instability of
early ideas on feedbad (Photo: AT&T Archives)
vacuum-tube amplifiers were
compounded when they were
connected in tandem over a
means of carrier telephony. It enabled the long-distance system based on carrier tedesign of accurate fire-control systems in lephony. (Carrier telephony became feasiThe author is with the College of Engi- World War 11, and it formed the basis of ble in 1917 with the invention of a
neering and the Science and Technolog? early operational amplifiers, as well as practical wave filter by George Campbell,
Studies Department, Coimdl University. precise, variable-frequency audio oscilla- an AT&T researcher who had invented the
Ithaca, NY 148.50. This work ”as support- tors. With the transition from vacuum loading coil independent of Michael
ed by the National Science Foundation. tubes to microelectronics after World War Pupin around 1900.)
Engineering Education Coalition, Synthe- 11, negative feedback retained its status as
The gain of the amplifiers varied with
sisproject, headquartered at Cornell Unr- an integral part of communications and plate voltage, temperature, aging of the
wrsity Summer. 1991.
control systems because it was associated tubes, etc., while the nonlinearity of the
82
IEEE Control Systems
high-gain device. Correctly designed, negative feedback reduced distortion and noise,
while stabilizing gain by making
it dependent on the passive fkedback network, instead of the
troublesome active elements
(vacuum tubes). Two other
AT&T researchers, H.T. Friis
and A.G. Jensen, had investigated the feedback caused by the
plate-to-grid capacitance in a
vacuum tube and its influence on
amplification in 1924. They
noted this feedback could increaSe or decrease the amplification (i.e., be positiveor negative).
But they regarded plate-to-grid
nuroiu 3 . D L U C K 111 i Y 4 1 W W L J U ~ M UJ rrrc dmplibing equipfeedback as an unwanted phenumamDlifiers for a
,.
*
‘17g distortion by i.eiwsincg
nomenon in amplifiers, to be
the umplijier’s output and feeding it back into the inneutralized by adding a capaciioto: AT&T Arc,hiws)
tor, and did not see the benefits
of negative feedback [9].
thinking” [4, p. 63) and agreed
Black probably knew about their work
to let him pursue the topic, provided it did sought a way to reduce it at the output of
not interfere with his other work. Black the amplifier. This was a critical step and because it was published in the Bell SysJournal, but it is unclear if
approached the problem by trying to make put him on the track to the eventual solu- tem Technic~~l
it influenced his research. Black recalled
vacuum tubes operate in a more linear fash- tion.
ion, i.e., he wanted to make the active eleAlthough resurrected in the 1970s for that he was working on the feedforward
ments of the amplifier produce less single-sideband microwave radio, the and other related topics when he had the
distortion. Mervin Kelly, well-known later feedforward amplifier did not work well insight about negative feedback. Mervin
as the head of Bell Labs when his researchers for carrier telephony in the 1920s. Black Kelly, however, said, “Finally, a matheinvented the transistor in 1947, was then in applied for a patent on the invention in matical analysis convinced him that by
charge of electron tube research and coop- February 1925, which was issued in Octo- merely inserting a part of the output power
erated with Black on the endeavor, but to no ber 1928 [6], and then, as a member of a into the input in negative phase, he could
avail [SI.
development group at Bell Labs, tried to obtain any desired reduction in distortion
Black then had an important insight make it work in a systems environment. products at the expense of a sacrifice in
that lielped him “reframe” the problem. He built an experimental amplifier to re- amplification. His final mathematical
He recalled that in 1923, “I attended a duce the invention to practice, but precise at?ulwis was conceived while he was
lecture by C. P. Steinmetz [chiefengineer balances and subtractions of signals were crossing the Hudson River on the Lackawanna Ferryboat en route from home to
at General Electric] at an A E E meeting hard to achieve and maintain in practice.
[the American Institute of Electrical EngiThis was the state of Black’s work on the laboratories” [5,p. 722, my emphasis].
Black had the page of The New Yurk
neers, a forerunner of the IEEE] and was telephone amplifiers in 1927, the year of
impressed by the Steinmetz way of getting the famous flash of insight on the Lack- Times witnessed by a co-worker the momdown to the fundamentals of a problem. As awanna Ferry. What occurred between the ing of his insight, then set out to build his
a result I restated my assignment as being feedforward experiments and that fateful amplifier and prepare a patent application.
that of removing distortion products from moming on the Ferry is not clear. Negative He submitted an extremely long application
the amplifier output. I immediately ob- feedback was a concept diametrically op- ( 5 2 pages. 126 claims) in 1928, but the
served that by reducing the output to the posed to feedforward and was not an ob- patent office objected to many of the claims,
same amplitude as the input and subtracting vious direction to explore. Black probably apparently because his concept of negative
one from the other, the distortion products knew about the two main types of feed- feedback flew in the face of accepted theory.
only would remain which could then be art- back at the time: I ) the comparison of The examiners finally awarded the patent
plified in a separate amplifier and used to output and input signals to generate an nine years later, in December 1937 [lo],
cancel out the distortion products in the origi- error signal to control the output, which after Black and others at AT&T developed
nal amplifier output ...Thus, the Feedforward had been widely used in mechanical and both a practical amplifier and a theory of
Amplifier came into being” [4, pp. 64-65].
electrical control systems [ 71; and 2) posi- negative feedback.
Although not based on negative feed- tive feedback, which had been used for
Development
back, this amplifier was an essential step to oscillation and increased amplification
As Black recalled, the road from the
its invention because Black had reformu- (regeneration) in radio equipment since
lated the problem. He was no longer trying about 1913 [8]. But Black took a much Lackawanna Ferry to a practical amplifier
to prevent vacuum tubes from causing dis- different approach. He used negative feed- was long and rocky. He had a difficult time
tortion. He accepted that distortion and back to reduce the amplification of a very with the amplifier “singing” (breaking
1 . .
August 1993
. 1
83
tance-capacitance tuned variinto oscillations) and devised a
able frequency audio oscilladesign rule to guard against
tor described in this paper was
this instability. In May 1928,
William Hewlett’s contribuHany Nyquist (1889-1976) and
tion to the paper, and was furother communication engineers
thermore the foundation on
at AT&T conferred with Black
which the Hewlett-Packard
about using his amplifier for a
Company was built” [ 181.
n e w cable carrier system.
T h e h i s t o r y of H - P ’ s
Nyquist, who received his Ph.D.
founding has elements of the
in physics from Yale in 1917,
stereotypical story of indepenthought Black‘s design rule was
dent inventors working alone
too stringent and did an analysis
in a garage. David Packard
of negative feedback. This work
and Hewlett did start their
led to what later came to be
company in Packard’s garage
known as the “Nyquist criterion”
(Hewlett lived in a cottage on
for determining when an amplithe grounds). But the influfier with negative feedback was
ence of Stanford and Terman
stable. He published the paper
was considerable. Hewlett and
containing the criterion in 1932
Packard retumed to the area
during the patent office deliberaafter graduating from Stanford
tions on Black’s patent and
in 1934. Packard took a leave
joined Bell Labs in 1934 11 11.
of absence from GE to accept
Black recalled, “Although this
a fellowship in 1938, while
criterion is simple in expression
Hewlett retumed to work unand application, Nyquist’s derider Terman after finishing a
vation of it required a matheMaster’s degree at M.I.T. in
matical-physical intuition given
1936. Terman encouraged the
to few men” [ 1, p. 7231. Black’s
pair to form a company to marclassic paper on the negative A page jrom Harold Black’s notebook, pmhabh lute 1920.
ket Hewlett’s variable-frefeedback amplifier, published in (Photo: AT&T ArchivrJ)
quency oscillator, lent them
1934, referred to Nyquist’s pa$538, helped them get a bank
per and his stability criterion
vanced training in mathematics and phys- loan of $1000, and helped them work out
[121.
In that same year, during the develop- ics - a common aspect of the history of a deal with IT&T who bought their international patent rights in exchange for unment of a coaxial-cable camer system electronics in this period.
derwriting their U S . patent application.
with a passband of 1 MHz and the possiAn early customer for the audio-frebility of several hundred amplifiers, anHewlett-Packard and the
quency oscillators was Disney Studios,
other Bell Labs theorist, Hendrik Bode
Negative Feedback Amplifier
who ordered eight for the film “Fantasia.”
(1905-1982), led a group of mathematiThe application of the negative feedcians in the development of design tech- back principle was taken up fairly rapidly By 1940, the fledgling company had nine
niques that took full advantage of Black’s at Stanford University by electrical engi- employees and had moved out of the gainvention [ 131. Bode, an applied mathe- neering professor Frederick Terman and rage to develop a full line of products matician who received an M.A. from Ohio his students. In 1939, Terman, William based initially on the negative-feedback
State in 1926 and a Ph.D. in physics from Hewlett, Robert Buss, and Francis Cahill amplifier 1 191.
Columbia University in 1935 1141, pub- wrote a paper that described many uses for
lished his paper, “Relations Between At- negative feedback: in a laboratory audioCommon Themes
t e n u a t i o n a n d P h a s e in F e e d b a c k frequency amplifier, an audio-frequency
The story of this amplifier illustrates
Amplifier Design,” in I940 [ 151. The pa- voltmeter, a tuned-radio receiver, high-Q many themes in the history of technology.
per, and a resulting book in 1945 (which circuits, and laboratory oscillators. Nega- Although the invention can be traced to a
included what engineers now call “Bode tive feedback resulted in low distortion, “flash of insight” by a single person, the
plots”), used the powerful tools of net- stable gain, and a small phase angle, while inventor was well trained in mathematics,
work theory to show how to design feed- protecting the circuits from the harmful engineering, and science, and had been
back amplifiers with the desired gain and effects of vacuum-tube aging, variable working with others in this area. The develfrequency response in a precise manner supply voltages, and so forth. These char- opment of the amplifier was truly a team
[16]. Thus, although Black is usually rec- acteristics were of high value in precision effort at Bell Labs; theorists with excellent
ognized as the inventor of the negative voltmeters, oscilloscopes, and oscillators. mathematical skills developed a theory that
feedback amplifier, its development and Terman, who often is called the founder of helped engineers understand the original inthe recognition of its possibilities in com- what came to be known as “Silicon Valley” vention and develop it further. The applicamunications, measurements, and control because of his promotion of commercial ties tion of the invention by Hewlett and Packard
systems were the result of a group effort between Stanford and local electronics firms not only started a new business, but reinbetween engineers and theorists with ad- 1171, recalled in the 1970s that the “resis- forced the connections between the univerI
84
IEEE Control Systems
sity and the electronics industry that b e
came a characteristic of what people later
called the Silicon Valley style of invention. Thus, the history of the negativef e e d b a c k a m p l i f i e r is an excellent
example of the complex interplay between
theory, experiment, and practice in the
institutional settings of established industrial research labs, booming businesses,
and expanding universities that became
common in the U S . electronics industry
after World War I1 [20].
[SI Mervin J. Kelly. "Career of the [I957 AIEE
Lamme] Medalist." Eke. En,?.. vol. 77. pp. 720722. Aup. 1958.
[ 14) Mac E. Van Valkenberg. "In Memoriam:
Henrik W. Bode (1905-1982),"IEEETrans.Aut~.
Conri-ol, vol. 29, pp. 193-194, 1984.
[6] Harold S. Black. U S . Patent, 1 686 792, filed
Feb. 3. 1925. issued Oct. 9. 1928.
1151 Hendrik W. Bode, "Relations between attenuation and phase in feedback amplifier design,''
B e l l S j r t . Tdi..I.,vol. 19, pp.421-454,July 1940;
reprinted in J.E. Brittain, Tuf.riirigPoints iri Ameri('at1E/cc.tric.crl Histor\, pp. 359-361.
171 Otto Mayr. Thc Ori,yiris of Fert/huc,k Control.
Cambridge. MA: M.I.T. Press. 1970.
[ 161Henrik W. Bode. NehwrkAnalysis undFeedhod Amplifier De.ri,qn. Princeton, NJ: D. Van
Nostrand. 1945.
[ X I D.G. Tucker. "The history of positive feedhack: The oscillating audion. the regenerative receiver. and other applications up to around 1923.''
Rutlro K. Elri . E r t , q . vol. 42. pp. 69-80. 1972.
Acknowledgment
[9]H.T. Fnis and A.G. Jensen,"High kquency amplifier~."BcllS\..\t TCYh ./..vol.3,pp. 181-205.Apr. 1924.
The author thanks Richard Compton
and Sheldon Hochheiser for reading an
earlier draft of this paper.
[ I O ] HaroldS. Black,U.S. Patent,? 10267I.filed
Apr. 22. 1932. i\sued Dec. 21. 1937.
References
[ I ] Harold Black,"Invention in engineering,"Elec..
Eng., vol. 77, pp. 722-723, Aug. 1958.
[2] Stuart Bennett, A History ofControl Eri,qirieering, 1800-1930. London: Peter Peregrinus. 1979.
[3] Harold S. Black, "Inventing the negative feedback amplifier," IEEE Spe(./rum,vol. 14. pp. 5460, Dec. 1977.
[4] E.F. O'Neill, Ed., A History of Erigirzeer-irz,q
and Science in the Bell System: Trunsmissiori
Technolog! (1925-1975). Bell Labs. 1985.
[ 171 James C. Williams. "The rise of Silicon Valley." Anio.. Heritage I m w t . Techno/., vol.6, pp.
18-24, Spr./Suin. 1990: and Stuart W. Leslie and
Bruce Hevly. "Steeple building at Stanford: Electrical engineering. physics, and microwave research,"
Proc. LEEE, vol. 73. pp. 1169-1180, July 1985.
[I81 F.E. Terman. W.R. Hewlett et a/., "Some
applications of negative feedback with particular
reference to laboratory equipment," Proc,. IRE,
vol. 27, pp. 649-655, Oct. 1939; reprinted in J.E.
Brittain, Tirrnirz,g Points in American Electrical
History, pp. 351.357: quotation on p. 350.
[ I I I Harry Nyquist. "Regeneration theory," Bell
S\sr. T ~ l i./..
. vol. 1 1 , pp. 126-147, Jan. 1932.
121 Harold S. Black, "Stabilized feedback amplifiera,"Ekv Eri,q.. vol. 53. pp. 114-120, Jan. 1934;
Bell S\..sr. T K ~ , I. . vol. 13. pp. 1-18, Jan. 1934:
reprinted in James E. Brittain. Tirrriing Points ir7
Amrric~uri Elec.tr.ic.cr/ Histor-y. New York: IEEE
Press. 1976. pp. 343-349.
[ I Y ] Tekla Perry. "When the car was out, the
business was 'in'." IEEE Spectrum, vol. 25, pp.
44-45. Apr. 1988.
[20] Ronald Kline. "An overview of twenty-five
years of electrical and electronics engineering in
the Proc.eedi~7,y.\of'rhe IEEE, 1963-1987," Proc.
/EEE. vol. 78, pp. 469-485, Mar. 1990.
[ 131 Hennk W. Bode. "Feedback: The history of an
idea." in Selec.tecl P u l ~ r r so r i Mutheniuric,al T r e d
iri Coritrol T/ieoi:\.. Richard Bellman and Robert
Kalaba. Eds. Neu York: Dover. 196.1, pp. 107-124.
BuildSim'"1.0
i
CORRECTION
i
The new approach to omulatfon o n the MarJntorh R PC
The footnote to the article "Sliding
Mode and Classical Controllers in Magnetic Levitation Systems," by Dan Cho,
Yoshifumi Kato, and Darin Spilman
(IEEE Contrd Systems Mugu5ne. Feb.
1993, pp. 42) incorrectly identified part
of Y. Kato's company affliation. Kato
was on leave from "Nippondenso Co..
Ltd." (not "NipponDenso Co.," a4
stated in the footnote).
Point & Click
Modeling
Automated
Project
Management
6 lnte ration
Mehods
1
~
Reader Service Number 4