Is Ivor Catt’s work merely an obvious development from Classical Electromagnetism in the Digital Age?

Should Catt’s 20 page 1967 article have been published in a peer reviewed journal?

We can learn so much from Chris Spargo. I wish I could find his email where he said that classical electromagnetic theory must be correct because of the practical successes which have resulted.

In 1964 I was completely a creature of classical electromagnetism, and since graduating had had three years of experience in digital computer design, designing a small part of the first transistorised computer, the Ferranti Sirius , adding the “Divide” instruction to the instruction set, and also the aborted change from magnetostrictive delay line memory (torshional pulses down a wire) to the new magnetic core memory.

Motorola Phoenix developed some of the first integrated circuits, and made the fastest logic circuits, Emitter Coupled Logic (ECL). In 1964 they had developed a 1.35nsec logic gate with 2nsec output rise time. As they were only expert in semiconductor technology, they were concerned about the possibility that the increased speed might mean that problems like crosstalk between signals would have increased to an insurmountable level. As an expert in high speed signals (but with a mere 5 years of experience), they hired me to look into the matter.

IBM was developing a secret computer for the use of the National Security Agency (NSA). IBM told NSA that the necessary scratchpad memory, the Ballman Scratchpad Memory , 64 words, 8 bits per word, 20 nsec access time, could only be made to run at 35 nsec access time. This would slow down the whole million dollar computer. NSA decided to give the job to Texas Instruments (TI), giving a specification of 20 nsec., a partially populated model to be delivered within nine months. Because of complaints that TI always got such contracts, NSA gave a split contract to Motorola, and they competed.

At Motorola we beat TI and got the follow on contract, having delivered a partially populated memory with an access time of 14 nsec. compared with TI’s 17 nsec. We won the follow-on contract.

At Motorola, my team, headed by Emory Garth, designed a 13 layer printed circuit mother board, each signal plane being Sandwiched between two voltage planes, or on the surface. One of two problems was, what would be the crosstalk between two parallel signal lines  ? ( The other key problem was , what was the decoupling at a point between a 5v plane and a 0v plane? Not discussed here.) I read the literature, and found that peer reviewed literature, including Jarvis , told us that the crosstalk was related to the rise time of the signal. This meant that a zero rise time step would produce infinite crosstalk, and our ultra high speed logic would not be viable. Thus, peer reviewed classical electromagnetic theory as it had developed told us that our faster logic would not work. I also realised that we had a new kind of amplifier, the crosstalk signal being of larger amplitude than the signal which created it. I felt sure that this could not be true.

I had left the late Ken Johnson, then the best engineer in Ferranti R&D in Manchester, three years before, but now stumbled on him thousands of miles away, on a visit, walking down the corridor in Motorola Phoenix. I grabbed him, and told him about my problem. He told me there were two signal modes, and gave me the wrong ones. Later, I developed the correct two modes, the Even Mode and the Odd Mode . I discovered that they travelled at different velocities.

As I was committed to classical electromagnetism, I proved all this mathematically within the framework of classical electromagnetism (and so Chris Spargo would think it was another success for classical electromagnetism). For 50 years, I did not realise that in the diagrams , in the third trace in Figure 9.3, two electric currents must have been travelling in opposite directions down the same part of the same wire. I did not notice this for 50 years, because, wedded as I was for the first half of that time to classical electromagnetism, I could not have imagined such a thing. 50 years later, the belated realisation having been rejected by peer reviewed journals, I finally published this in an unrefereed journal. No professor or text book writer in the world will comment on this article, even though the editor put “Explosive” on the journal’s cover. See my friend the late Gordon Moran .

Should students be taught that in the diagrams in one of the top peer reviewed journals, in the third trace in Figure 9.3, two electric currents must have been travelling in opposite directions down the same part of the same wire?

For 50 years, a peer reviewed article had a photograph of two electric currents travelling in opposite directions down the same piece of wire, Figure 29 , and nobody, including myself, noticed. Was this part of classical electromagnetism? Does classical electromagnetism embrace today’s high speed logic, or is it admitted to be only about the sinusoidal radio signals of 1950, proven to be useful when working on radio?

Does all my theoretical work result from success in achieving practical results? Do my practical achievements validate the theory derived from them? It was necessary to understand the mechanism of crosstalk in order to achieve the practical result of the Ballman Scratchpad Memory. But the mechanism produced the (under classical theory) anomalous picture in the diagrams , which, once realised, demanded a change in theory. Here we see how theory and practice work hand in glove. But this includes new digital practice, not merely the old pre-digital practice which you, Chris Spargo, cite.

Ivor Catt   2 June 2015


I used the mathematics of classical electromagnetism and validated mathematically the two modes, EM and OM, and their two different velocities. I also showed these in oscilloscope photographs. This appeared to be a picture (in the photographs taken from the oscilloscope) again confirming classical electromagnetism. However, the mathematics did not point out, and I did not notice, that at the start, before they separated out, the two modes were superposed. So far so good. But the two modes had electric currents in opposite directions in/on the passive line.  So before they separated out, in/on the passive line there were two electric currents in opposite directions in a section of the same conductor. I only noticed this 50 years later.

So the mathematics and the pictures seemed to validate each other. But the theory, which had electric current in the conductors, must have been in some way wrong.

What a pity that no professor or text book writer in the world will bother to address, or contribute to, this difficult discussion. After 50 years of digital electronics, this difficult issue should be a central feature of today’s science.

Ivor Catt  2 June 2015