Computers, vacuum tubes, semiconductors and printed circuit boards all have an older history than most people think, according to Clive “Max” Maxfield (right), keynote speaker at the recent Allegro and OrCAD 16.3 Virtual Conference event. Max is a prolific author and commentator on electronics and computers, and he recently launched TechBites.com, a “collaborative community” for scientists and engineers.
With a speech entitled “Electronics miniaturization through the ages,” I suspected Max might find something from the Middle Ages to talk about. He actually started his talk with the evolution of tetrapods (four-limbed creatures with fingers and toes) 350 million years ago. The first tetrapods, he noted, had eight fingers on each hand. Had this persisted until the evolution of humans, we’d be using a base-16 numbering system, very handy for computers, Max noted.
Computers, however, would have to wait until the 1820’s until Charles Babbage designed his Difference Engine and Analytical Engine. While that story is well known, I have to admit I was not familiar with the Z1, the first true mechanical computer. It was built by German inventor Konrad Zuse in his parents’ living room in Berlin in 1938. It used binary numbers, floating-point arithmetic, processing units, and a 64-word memory.
Mechanical computers have low clock frequencies – the Z1, for instance, ran at 1 Hz. The Harvard Mark 1, built in the U.S. during World War II, used electromechanical relays. Max noted that electromechanical relays go back to the 1820’s and were subsequently used in early telegraph and telephone equipment.
Subsequent computers used vacuum tubes, and these too have a long history. In 1883 an associate of Thomas Edison discovered the Edison effect, which describes thermal electron emission. It was later employed in vacuum tubes, but not by Edison. Sir John Ambrose Fleming invented the first vacuum tube diode in 1904. If Edison had followed up on the Edison effect, Max noted, the electronics industry might have started 21 years earlier.
Many people think that the 1947 invention of the transistor by William Shockley, John Bardeen and Walter Battain was the single most important electronics innovation of the 20th century. But the transistor goes further back. Physicist Julius Lilienfeld filed the first patent for a transistor in Canada in 1925, but failed to publish any research articles about it. Here again, as Max noted, the electronics era could potentially have been launched 25 years earlier.
Since the Virtual Conference was about modern-day PCB design, it’s only fitting that Max would also talk about the history of printed circuit boards. He noted that Edison wrote a letter to a friend about the possibility of printing additive traces and using conductive inks, but didn’t pursue the idea further. In 1903 Albert Hanson filed a patent for a “printed wire” that produced conductive metal patterns on a dielectric by stamping copper or brass foil patterns. This helped pave the way for double-sided, through-hole PCBs. Of course, true PCBs didn’t take off until there were ICs to place on them.
Max also looked into the future, speaking of such topics as embedded optical interconnect with optical waveguides, silicon circuit boards, stacked bare die on diamond substrates, carbon nanotubes, and quantum computing. Exciting stuff, but here I cite a word of caution. Silicon Valley old-timers will remember waferscale integration, a promising future technology that attracted hundreds of millions of dollars and some very well-known people in the 1970s and 1980s. Unfortunately, it didn’t work.
I do agree with Max, however, that “we are on a constantly accelerating development curve,” and that we may have just reached “the knee of the curve.” Max said that “it is almost impossible to conceive what the world of electronics will be like in 20, 50 or 100 years.”
The Virtual Conference also featured a keynote by Keith Felton, group director of product management at Cadence, who spoke about the challenges PCB designers are facing (see my earlier blog) and about the advances in the Allegro and OrCAD 16.3 releases. At last count there were over 600 conference attendees.
The two keynotes are available on an on-demand basis. A quick registration is required. Go into the Conference Hall and you’ll find the keynotes.