Radio Broadcast (May 1928-Apr 1929)

162 use special means of cooling. Water and air are generally used cooling agents, sometimes without special means to secure motion, as in the ordinary small vacuum tube which radiates heat like an electric lamp, sometimes with forced air cooling, as when a fan is directed on a hot bearing, and sometimes, as in the water-cooled vacuum tube, in the form of a stream of water constantly pumped past the surface which requires cooling. The last means is the most effective and also the most elaborate and inconvenient. It has the added disadvantage that, once the apparatus is designed to rely on it, the flow of cooling liquid must be maintained whenever the power is on, or destruction of the unit will result. Fig. 2 is a sketch of a water-cooled vacuum tube. Looked at from the outside such a tube is about half glass and half metal. The cathode or filament end is glass. At this end the filament leads are brought out, and generally the grid lead issues along the side of the glass cylinder a few inches below the filament terminals. The anode or plate is the portion of a tube which gets hottest from the electron bombardment, as may be seen in a radiation — or air-cooled — tube when the metal of the plate becomes red hot. The obvious device is then to make the anode a part of the vacuum-enclosing wall of the tube and to cool it on the outside by running water over it. The lower half of the tube shown is, accordingly, the plate itself in the form of a hollow cylinder, with the grid and filament inside. This construction entails an air-tight seal between the glass and metal, and until the development of such seals to commercial practicability, water-cooled vacuum tubes could not be built. A tube is conceivable in which the entireouter container would act as the anode and the filament and grid leads would be brought out through hermetically sealed insulating bushings, but to make such outlets air-tight and at the same time capable of standing up under thousands of volts potential between the leads and the metal of the platecontainer would be an excessively difficult job. The practicable form of the water-cooled tube is that shown, in which a glass cylinder is used to support the low tension elements and to bring out conductors from them safely, and the water surrounds the metal anode only. In the water-cooled tube all three methods of transfer of heat mentioned in physics courses — conduction, convection, and radiation — are active. Heat is radiated from the outside of the tube and water-jacket, and within the waterjacket convection takes place, as well as direct conduction from the hot metal of the anode to the water. The last mode of transfer predominates and the tube cannot survive unless it is kept "up. Water is not the only possible cooling liquid for vacuum tubes. Other non-conducting or very poorly conducting liquids may be used; oil, in fact, is superior in some instances, as at very high radio frequencies. The weakest portions of a vacuum tube are naturally the seals, and, in a water-cooled unit, especially the anode seal, because of its special construction and large size. The principle used is to make the metal at the seal very thin (a few thousandths of an inch) and soft, so that it will adhere to the hot glass in cooling, during the process of manufacture. This leaves a highly vulnerable belt at this point, where a break or puncture may readily be caused. Usually one side of the filament in these large tubes is at ground potential, and the plate is maintained at a positive potential of several thousand volts. The tube plates must accordingly be insulated from the ground, but as the water supply system, containing tanks and mains, is normally grounded, it becomes necessary to use RADIO BROADCAST Water Inlet FIG. 2 some special device to insulate the anodes while supplying water to them. This is accomplished by the use of such long columns of water in hose made of insulating material that the resistance is of the order of a megohm. Usually it is convenient to wind the length of hose required around a cylindrical form; such water coils, as they are called, are one of the principal outward features of large vacuum tube transmitters. The length of hose needed depends, of course, on the specific resistance of the water; fifteen feet for each tube is a common length. Fig. i illustrates schematically the layout of such a cooling system, showing a vacuum tube, its water jacket surrounding the anode, the water coil, a pump for maintaining the water flow, and a radiator for cooling the water before it returns to the anode. Any of the methods of water cooling used in steam plants, such as cooling towers in which the water falls successively into vertically arranged troughs, or cooling ponds with or without sprays, are applicable to radio transmitters requiring special cooling measures. When the cooling system is closed the same water may be used over and over. Such circulatory systems are of course essential when the cooling fluid is too expensive to be wasted, as in the case of distilled water. If a plentiful supply of suitable water under sufficient pressure is available an open system is less expensive. For example, a radio station might take water from a river sufficiently far upstream to get the required head and then discharge it through a tail race like that of a hydro-electric power plant. In some instances radio stations get cooling water from city water supply systems. But water from the usual sources, even if it is clean, apparently free from suspended matter, and drinkable, may not be useful for cooling vacuum tubes. Impurities in JULY, 1928 the water may cause the formation of scale on the outside of the hot anode, as in boiler tubes. But in the case of a vacuum tube, this deposit is much more serious, because it interferes with the rapid transfer of heat from the plate to the water stream and results in gassing and destruction of the tubes. The rate of deposition of scale depends on the chemicals in the water, the temperature of the anode, and the rate of flow of the water; for a given quality of water more scale is deposited in a given time as the temperature increases. Hence, in modern water-jacket design rapid water flow is the objective; a relatively thin stream is shot past the plates. Soft or distilled water is much to be preferred to hard water for radio cooling purposes. The suitability of the water may be determined by analysis. The table below shows the chemical content held in solution in the cooling water of four different radio stations: GRAINS PER GALLON SUBSTANCE 1 2 3 4 Calcium carbonate 4.73 0.26 3.85 None Calcium sulphate None 2.76 Trace 0.45 Sodium carbonate 3.91 None None None Sodium sulphate 3.09 0.28 0.24 None Sodium chloride 1.02 1.02 0.85 0.67 Sodium and potassium nitrates 0.41 None None None Magnesium carbonate 1.37 0.13 1.68 None Magnesium sulphate None 0.76 0. 19 0.47 Aluminum and iron oxides 0.36 Trace Trace Trace Silica 1.22 0.17 0.81 0.15 Total 16. 11 5.38 7.62 1.74 Specimen No 1 came from the water supply of a mid-western town, and the station using it had a poor tube record as might be expected. The second and third samples were from wells, and the stations using them were only a few miles apart in the eastern part of the country. Although the total grains per gallon figures are not so far apart for the two, it will be noticed that the distribution among the substances listed is quite different. These two samples are probably average, and each deposited considerable scale under normal operating conditions. Specimen No. 4 came from a spring near by. As far as these salts go it is almost as good as distilled water. However, organic matter is also found and may play a considerable part in the deposition, as was shown in one case where scale formed rapidly during the summer and more moderately during the winter, although the difference in operating temperature was not enough to explain the effect. When all is said and done the best course, with tube prices what they are, is to use a closed circulatory system employing distilled water. The temperature of the water is usually measured at two points on each frame before and after it has passed the anodes. Tube manufacturers generally specify an outlet temperature of 70 degrees Centigrade as the limit. As a matter of fact this is much too high for the best conditions of operation, especially by broadcast standards. A figure around 30 degrees Centigrade is more to be desired. If the flow is adequate the inlet and outlet temperatures will not differ by more than five degrees Centigrade. When the temperature and composition of the cooling water are such that scale deposition cannot be avoided, the only remedy is to institute a regular cleaning schedule for the anodes. Dilute hydrochloric acid in a four-to-one solution may be used. Each tube must be removed from the transmitter periodically and the anode immersed in this solution, which may be held in a stone or crockery container. If the scale forms rapidly each tube must be cleaned once a week or even oftener. Such frequent removal of the tubes from the transmitter entails added danger of breakage. The acid may be run through the water jackets to obviate this, but of course it must be thoroughly washed out before power is again applied to the set.