The diode as an energy-controlled, not a charge-controlled device.

The traditional theory of operation of the diode is, for me, one of the many casualities of advances in electromagnetic theory during the last 25 years[1].

Whilst at Motorola, Phoenix, in 1964, work on the problem of how to interconnect high-speed (one nanosecond) logic gates led me to the same general conclusion as had been reached (unknown by me until 1972) by Oliver Heaviside a century before when he tackled the problem of how to improve undersea telegraphy from Newcastle to Denmark[2].

"... [The electric and magnetic fields] are supposed to be set up by the current in the wire. We reverse this; the current in the wire is set up by the energy current through the medium around it. The sum of the electric and magnetic energies is the energy....

".... A line of energy-current is perpendicular to the electric and magnetic force...."

Our conclusion was that what he called "energy current" travelling down between the two conductors[3] guided by them as a train is guided by two rails, was the important feature of signalling, and not the electric charge and electric current in or on the wires. Twenty years later my view hardened when I came across the Catt Anomaly.

Let us deliver a 1ns-wide pulse down a long transmission line terminated by a diode (Fig.67). When the pulse reaches the diode, it does not carry any charge with it. Catt's Anomaly shows that charge could not have travelled fast enough to keep up with the pulse, which travels at the speed of light. If we are agreed that the diode will respond (for instance 'start to conduct') after a delay which is small (say 100ps) compared with the time delay down the transmission line delivering the pulse, then it must be responding to the energy current, that is, the TEM wave or pulse approaching it in between the two conductors. This TEM pulse enters directly into the side of the crucial interface or surface between the p-region and the n-region which together make up the diode.

Note the phrase on page 30 col.2; "Nothing ever traverses a capacitor from one plate to the other". Applied to the diode, this seems to say that nothing travels across the junction from the p-region to the n-region, or vice versa. The only travel is along the surface between the two regions, in a direction at right angles to the generally supposed direction of movement.

When the leading edge of the pulse reaches the near edge of the diode, it finds a change in characteristic impedance. As a result, most of it is reflected, but a small portion continues forward to the right, down the very narrow transmission line made by the surface between the p and n regions. It is possible that the effective dielectric constant is large so that the velocity of propagation, , from left to right along the p-n interface is very slow. At the speed of light in a vacuum, the round trip across the p-n interface of a diode a tenth of an inch wide would be 20 picoseconds, but since the effective ε  will be bigger than for a vacuum, the delay will be greater.

When the step reaches the right-hand edge of the diode, it sees an open circuit and reflects back toward the left, so that the total voltage across the junction doubles. When it gets back to the front (left-hand) end, it reflects toward the right again (except for the very small portion which escapes across the Zo mismatch back into the transmission line leading away to the left). By this mechanism of zig-zag repeated reflections across[4] the diode, the amount of energy (current) in the p-n surface increases in a series of diminishing steps which approximates to an exponential (Fig.66). When the energy density builds up beyond some critical level (0.7v), there is a 'snap', and the later advancing energy current sees a short circuit, and reflects with inversion[5].

Since no charge has been introduced into the p-n interface, it is totally inappropriate to explain the mechanism of the diode in terms of extra electrons. The explanation must be novel, in terms of the amount of electromagnetic energy present; that a level in excess of some critical value (0.7v) causes the TEM wave travelling down the p-n interface to see a change in what is ahead of it, from open circuit to short circuit. That is, beyond that critical amplitude the p-n interface cannot accept more energy and rejects it.

This developing analysis of the behaviour of a diode is totally at odds with the traditional view, based on electrons, holes, energy barriers across the p-n interface that charges are trying to climb up. Why does this earlier theory succeed in correlating at all with experimental results?

"... if in conversation you insisted that your elder daughter was identical to your younger daughter, whereas in fact their "equality" only related to their parentage, every conclusion that followed this absurd assertion would not necessarily be absurd. For instance, if you knew the address of one daughter you might therefore know the address of the other. In the same way, it is possible for 'valid' results to come from absurd postulates (like the absurd postulate that a diode is full of particles called electrons buzzing around trying to climb hills [in the wrong direction at the wrong speed]).

"... the two non-identical daughters might have the same address. It is these 'echoes of truth' which masquerade as scientific truth today" (ref. ll vol 1, p15).

False theories, like the theory that the diode is a device controlled by charge, exist in the real world, and so are influenced, or somewhat directed, by the imperatives of the real world in which they find themselves, at least when it comes to the moment of truth: the checking of theory against experimental result.

The Catt Question. (Was Catt's Anomaly.)

Traditionally, when a TEM step (i.e. a logic transition from low to high) travels through a vacuum from left to right (Fig.1), guided by two conductors (the signal line and the 0v line), there are four factors which make up the wave: (1) electric current in the conductors, (2) magnetic field, or flux, surrounding the conductors, (3) electric charge on the surface of the conductors, (4) electric field, or flux, in the vacuum terminating on the charge.

The key to grasping the anomaly is to concentrate on the electric charge on the bottom conductor. During the next 1 nanosecond, the step advances one foot to the right. During this time, extra negative charge appears on the surface of the bottom conductor in the next one foot length, to terminate the lines (tubes) of flux which now exist between the top (signal) conductor and the bottom conductor.

Where does this new charge come from? Not from the upper conductor, because by definition, displacement current is not the flow of real charge. Not from somewhere to the left, because such charge would have to travel at the speed of light in a vacuum. (This last sentence is what those "disciplined in the art" cannot grasp, although paradoxicallyt it is obvious to the untutored mind.) A central feature of conventional theory is that the drift velocity of electric current is slower than the speed of light.

For further information on the Catt Anomaly, see letters in the following issues of Wireless World; aug81, aug82,oct82,dec82, jan83.

Heaviside and the Catt Anomaly.

Oliver Heaviside did his work on Energy Current too early to discern the Catt Anomaly. The idea that electric current comprised electrons was still only an "ingenious .... theory [by] J. J. Thomson" in the 1905 Harmsworth Encyclopaedia, p2184. So the firm conviction that the electrons which comprised electric current had mass came too late for Heaviside.


[1]This section was first published in Electronics and Wireless World, September 1987, p903.

[2]ref.5c.

[3]i.e. the Poynting Vector.

[4]These insights will meet the same indifference as was discussed in Footnote 24, p13.

[5]I agree with L. Turin that this is simplistic, since the forward voltage drop of a diode is not sharp, and follows a law which includes electric current and temperature. The purpose of this section is to take the first step away from the conventional theory, which is obviously balderdash. Since it was first published in 1987 it has excited no comment, not even a riposte.