Before the war all Sylvania tubes used the best grade of India mica, but, now that this cannot be obtained in quantity, special procedures have been found that enable American and Brazilian micas to be used with equal satisfaction. For many tubes it is necessary to apply a roughening surface coat of an insulating spray in order to reduce the effect of metal vapors that may condense upon the mica surface. Getter tabs, hook wires, support rods, and connectors all have their own particular problems. The getter itself is an interesting chemical problem. In incandescent lamps, phosphorus is used as a getter because of its affinity for oxygen, but in tubes with coated emitters phosphorus is a poison, and besides, a more active material which will combine with gases other than oxygen is necessary. Magnesium was the first material used. It is well known as the material used for flash powder and photo flash lamps. A small piece of the metal is fastened on some part where it can be vaporized later. In recent years a better gettering action has been obtained by using a mixture of magnesium, aluminum and barium, the exact proportions varying with the requirements of the tube. The mixtures containing more barium give a darker color deposit and do not spread over the bulb as much as those with more magnesium. These getters of course must be kept stored in vacuum or they will be worthless and the tubes which have had the getter applied may only be allowed to stand a limited time before exhausting or the air will have destroyed the getter.
The exhaust process is much more than just sucking the air out of a bulb, for that is just to prepare the ground, as it were, for the necessary chemical changes which will make it into an operating radio tube. After most of the air has been pumped out, the heaters are connected to something more than normal voltage. The binders are first broken down chemically into gases which are then pumped out. A good pump is necessary because if these gases are not removed as fast as released they will cause undesirable interactions to occur and might also cool and condense on the other parts to cause trouble later. The barium and strontium carbonates on the cathode coating, now free of binder, give off carbon dioxide and are converted into the oxides required for emission. About this time a high frequency oscillator is used to make the metal plates red and sometimes even white hot. To most people a piece of metal is “all metal” but to a tube man it is as full of air as a porous brick. Heating these parts red hot drives out the air, but it comes out so slowly that for some high voltage transmitting tubes the exhaust process may last several hours. Careful control of this heat is important because too much will vaporize the metal which would be deposited on the insulation causing leak-age and possibly noise. Following this, the heater coil is moved closer to the getter which flashes and deposits active material on the bulb to react with any gas subsequently released.
When tubes are taken off the exhaust machine they are sometimes usable as they are, but the characteristics are not very stable. The ageing process consists of operating the tube at higher than normal current long enough to be sure that all the emitter coating is chemically changed and to permit gases and impurities still present to come out of hiding. An average time of about half an hour is required for this process in order to be sure that the whole surface of the cathode contains active electrons and that all traces of gas have been cleaned up by the still active getter. Tubes testing good after this are reasonably sure to give the required service, but regular tests are taken and put in operation at full rating for 1000 hours to assure high quality and to give warning in case the chemical or electrical properties of some material has changed.