0
Shopping cart
0
Cart is Empty

Oxide Cathode Life: Investigations into the Causes of Loss of Emission

A paper* read recently at the Institution of Electrical Engineers described the results of work by the Electronics Division of the Post Office Research Station (with its associated small valve factory) on the development of long-lived valves. The Post Office is interested in such valves mainly because of the need for them in submarine repeaters and other inaccessible apparatus; but their development is of considerable moment to others as well.

Apart from structural failures, with which the paper was not concerned, the life of a valve depends mainly upon the rate at which the emitting qualities of its cathode deteriorate with use. Loss of emission is the disease; if its causes are completely understood, we are more than half-way on the road towards finding a cure. There are two main accepted causes of oxide cathode deterioration: poisoning by attacks from residual gases, and the development and growth of inter-face resistance; on both of these the paper throws much new light.

Interface resistance was the subject of a recent article in Wireless World.** The authors of the paper agree with Eaglesfield in accepting Eisenstein’s view that interface resistance is due to the growth of a film between the cathode body and the oxide matrix, and that its production is due to deliberately introduced impurities, chiefly silicon.

They find, though, that the film does not increase in thickness as the valve ages. It appears to build rapidly up to its maximum thickness, after which a steady change in its nature sets in; it is to this change that the increase in resistance is due. They consider that it is caused, in part at any rate, by deactivation of the interface by gas poisoning, much as the matrix itself is deactivated; hence the complete elimination of residual gas would be likely to check the growth of interface resistance, in addition to ensuring longer life for the emitting surface of the cathode.

A method has been developed of making a triode or pentode measure its own residual gas pressure. As electrons flow from cathode to anode some collide with gas molecules. The positive ions, so formed, travel to the negatively biased control-grid and set up reverse grid current І_rg, from which the residual pressure can be derived. It has, however, been found more convenient to use a “vacuum factor”:

І_rg (µµA)
k = ————
І_a (mA)

Investigation of the behaviour of the residual gas under working conditions showed that when k was plotted against time there was always a sharp initial rise, followed by a slow, roughly exponential fall until a value k₀ (the “residual vacuum factor”) was reached; the value of k₀ is constant. The area enclosed by the k=f(t) and k₀ curves is a measure of the residual gas driven into the cathode and has been named the “gas integral.” It was found that valves with a high gas-integral were short-lived and that a low gas integral was an indication that long life might be expected.

Here, then, was one method of forecasting valve life after a test of comparatively brief duration; reliable tests of this kind are clearly needed for dealing with valves whose working lives may range up to 60,000 hours—say seven years of continuous running! Another of great value that has been evolved is the low-temperature total-emission test. Under working conditions emission is limited by the space charge; to measure the total emission there must be no such limiting factor. It is also desirable that the cathode temperature should be low enough for ionic equilibrium to be maintained within it. The method developed is to use the control grid as collector, making its potential +5V and earthing the other electrodes; for 6.3-V valves the heater is at about 2.6V.

In practice, valves are taken from the life-test rack, where they are running under working condition, and put through the total-emission test. The total emission, after a given number of working hours, can thus be plotted as a percentage of its original value.

The authors have no doubt that the amount of barium used for gettering the average valve is amply sufficient to absorb all residual gas if (and that is one of the big problems) physical association of gas and getter can be established. They have given much attention to the preparation of electrodes and supports and to pumping, gettering and ageing valves. It has been found that when residual gas is reduced (by methods suited so far to the laboratory rather than to the mass-production factory) to amounts far smaller than those in commercial valves, no deterioration in the emission occurs after thousands of hours of use. The authors’ estimates are always conservative; they see no reason why valves, with assured lives of 40,000 hours or more, should not be produced.

Poisoning of the cathode by occluded gases released from metal parts produces non-emitting patches on the surface of the cathode, which may be small and evenly distributed, or large and irregular. The authors in their investigations have found a new and unexpected source of cathode-poisoning gas. “A gas derived either directly or indirectly from the heated glass envelope,” they state, “is more destructive in action than any of the normal gas so far examined.” This gas is believed to be water vapour, which has been shown to have dire effects on a cathode at 1,000º K.

A second possibility is that the water vapour reacts with metallic carbides in the valve to produce unsaturated hydrocarbons of the acetylene type, which dissociate to form non-emitting carbon patches on the cathode.

R. W. H.
WIRELESS WORLD FEBRUARY 1952

* “The Life of Oxide Cathodes in Modern Receiving Valves.” G. H. Metson, S. Wagener, M. F. Holmes
and M. R. Child.
** “Valve Cathode Life.” C, C. Eaglesfield. Wireless World, December, 1951.

Shopping cart

Oxide Cathode Life: Investigations into the Causes of Loss of Emission