Discuss This Article

Accentuating the negative

By Michael W. Dorsey

Black at Bell Labs in 1941 with amplification equipment based on his negative feedback principle. (Reprinted with permission of Lucent Technologies Inc./Bell Labs.)

Harold Black’s 1927 invention, called the negative feedback amplifier, paved the way for routine long-distance telephone service by eliminating distortions that could reduce to gibberish the sound of a phone call amplified over and over again on its way across the country.

Among the numerous papers that Harold S. Black ’21 left behind at his death in 1983, many of which were recently donated to the archives in WPI’s Gordon Library, was a hardbound notebook in which he recorded tidbits of information he wanted to remember: checks he’d received, conversations he’d had, interesting articles he’d read. Here and there he also jotted down his thoughts. On Aug. 10, 1982, he wrote: “HSB—There is always a better way.”

It was a fitting motto for an engineer who, nearly six decades earlier, had solved the most vexing problem facing the young telecommunications industry, and in the process discovered one of the most important principles to emerge in electrical engineering during the twentieth century. Black’s 1927 invention, called the negative feedback amplifier, paved the way for routine long-distance telephone service by eliminating distortions that could reduce to gibberish the sound of a phone call amplified over and over again on its way across the country.

Black discovered that if a portion of an amplifier’s output were fed back into the amplifier in negative phase, the inverted distortions would cancel out the distortions introduced by the amplifier. The negative feedback principle would find wide-ranging applications in industrial controls, computing, cybernetics, guidance systems, information theory, and many other fields of engineering. In 1952, Black received the Scientific Award from the Research Corporation, with a citation that read, “Very few men have had the fortune to influence an entire field of human industry as profoundly as Harold S. Black.”

Over the years, the notion of negative feedback has become imbedded in disciplines well beyond engineering, including psychology, sociology, physiology, and ecology. In fact, negative feedback loops, where a product or action acts on an earlier stage of a process to reverse the direction of change, are found frequently in nature, where they help maintain equilibrium in biological systems.

A career of fits and starts

Black’s epochal discovery occurred just six years after he graduated from WPI with a bachelor’s degree in electrical engineering. A New Jersey native, he took a job with the engineering department at Western Electric Company in New York City, a unit that would combine with AT&T’s research department a few years later to form Bell Labs.

Black’s distinguished career got off to a rough start. “On my very first working day,” he recalled years later, “I lost three sets of tools, then had the bottoms of my trousers burned off when I brushed against a battery. Not long after that, I made a wrong connection to a rack of transmission equipment and blew out dozens of vacuum tubes.”

As an apprentice engineer, Black was assigned tasks that he considered menial. He asked his supervisor if he could focus instead on high-quality amplifiers for long-distance telephone communications. “Of course,” his boss told him, “provided you don’t let it interfere with your work.”

Black’s interest was understandable. One of AT&T’s most pressing technical challenges was finding a way to transmit thousands of channels of phone calls coast to coast, with minimal distortion. The company demonstrated the first transcontinental phone line in 1915, but even with loading coils (which decreased signal loss) every eight miles and banks of vacuum tube repeaters, the quality of the sound was poor and the cost of a phone call was astronomical.

To achieve economical long-distance phone service, AT&T would need to develop amplifiers of far greater quality than any then in use. Despite his boss’s reluctance, Black was able to apply his keen mind to the problem. At first, working with Mervin Kelly, who would later head Bell Labs, he tried to build better vacuum tubes to reduce the distortions and instabilities they introduced.

In 1923, when that approach had shown little promise, he began to think about the problem in a new way. Instead of reducing the distortion, he would try to remove it from the amplifier’s output, leaving behind just the original, pure signal. His first solution, which won him the first of his 357 patents, was the feed-forward amplifier.

In this device, the input signal is subtracted from the output, leaving just the distortion. The distortion is then amplified separately and used to cancel out the distortion in the original, amplified signal.

The technique worked, but the equipment was touchy and difficult to maintain. Black continued to refine the feed-forward amplifier for three more years, all the while searching for a more elegant approach. The search ended on Aug. 1, 1927.

The page from the Aug. 1, 1927, New York Times on which Black jotted down equations and diagrams describing the negative feedback amplifier, an idea that came to him in a flash while riding the ferry to work.

A famous ferry ride

The day started like any other for Black. He ate breakfast in his rented room in East Orange, N.J., caught the Lackawana train to Hoboken, then transferred to the Christopher Street ferry, which would take him across the Hudson River to lower Manhattan. It was a clear morning and the sun glinted off the Statue of Liberty.

Other engineers headed to Western Electric clustered on the foredeck to chat and enjoy the view, but Black kept to himself, as he usually did, thinking. At exactly 8:15, the solution to the problem he’d been wrestling with for six years came to him in a flash. Search-ing his pockets and finding nothing to write on, he ran to the ferry’s newsstand and bought a copy of the New York Times.

He leafed through the paper and found a page that was printed very faintly, but which clearly showed the date. On it, he sketched a canonical diagram of a negative feedback amplifier and a few basic equations to describe it, then he signed the page. When the ferry docked, he hurried to the lab, where he asked another engineer to review what he’d recorded and testify that he understood it. The framed page is now a treasured artifact in the Bell Labs archives.

The story of Black’s flash of insight has become one of the great myths in engineering history. But Black always maintained that the notion of negative feedback came to him all at once, Mervin Kelly wrote that he believed Black had worked out the basic details of negative feedback before that famous ferry ride, and only needed to arrive at the final mathematical solution en route to the office that morning.

It has also become clear over the years that while Black deserves the credit for the invention of the negative feedback principle, other Bell Labs theorists played pivotal roles in making it a success, including Harry Nyquist, who devised a widely hailed criterion for determining when a negative feedback amplifier would be stable and free of distortion, or “singing,” and Hendrik Bode, who developed design techniques that helped other engineers widely adapt Black’s idea.

From skepticism to acceptance

Though it is now a fundamental tool in electrical engineering, Black’s discovery was initially met with skepticism within Bell Labs, in part because the idea was radical, but also because Black, as an engineer with only a bachelor’s degree and a few years of experience, rankled the well-educated researchers and theoreticians who tended to drive the lab’s technical development. Black’s idea also met stiff resistance from the U.S. Patent Office, which took nearly 10 years to approve his application.

The delay in the patent was partly due to the large number of claims that Black included (the final patent filled 52 pages of text and 35 sheets of drawings) and his stubborn refusal to part with any of them. Alton C. Dickieson, Black’s lab assistant at the time, later remembered that Black early on saw a wealth of applications for negative feedback and “talked endlessly on this topic to his colleagues, and to anyone who would listen.” Black’s long-range view of the potential of negative feedback, Dickieson said, explains why he wrote his patent claims so broadly and defended them so fiercely.

Black worked diligently to build working prototypes of a negative feedback amplifier. In 1928, AT&T decided to put his discovery to a major test. The company simulated a transcontinental phone line in a laboratory in Morristown, N.J., with miles of cable, folded back on itself, and a negative feedback amplifier every 25 miles. The trial was deemed a success, and Black’s invention went on to become a vital component of the nation’s long-distance telephone infrastructure.

Black remained at Bell Labs for 40 years, continuing to innovate in many areas of electronics, including signal modulation, earning patents and recognition, though never again reaching the heights he hit so early in his career. His discovery of negative feedback brought him an almost endless series of awards and honors, including WPI’s Goddard Award for Professional Achievement, the Institute of Electrical and Electronics Engineers Lamme Medal, and induction as the 25th member of the Inventors Hall of Fame. At WPI, he is memorialized with the Harold S. Black Scholarship, established by his wife, Meta, in 1992, and the Harold S. Black Award, given by the Electrical and Computer Engineering Department each spring to exemplary seniors.

Black’s notes in the WPI archives make clear that he delighted in these honors and took great pride in his achievements. One sentence that he included in a handwritten autobiography neatly sums up his opinion of his place in engineering history, an opinion that seems now to be widely shared. “It is no exaggeration to say,” he wrote, “that without Black’s invention of the negative feedback amplifier, combined with his continuous and ever-expanding research and development of new feedback concepts, our exotic communications of today would not be possible.”

Maintained by: webmaster@wpi.edu
Last modified: Aug 26, 2005, 12:00 EDT
[WPI] [Alumni] [Home]