Radio Broadcast (May 1928-Apr 1929)

-z. RADIO BROADCAST Fig. 1 — Schematic circuit of thermionic milliatnmeter such cases it would be necessary to insert a condenser in the a.c. leads to the tube. The size of the condenser is determined by the impedance which it is permissible to insert in the circuit carrying the a.c. The resistance of the thermionic meter alone to the a.c. is approximately 50 ohms. Calibration by Calculation THE check between calculated and calibrated points agrees within 3 per cent, and this shows that it is not necessary to have 20 or 40-milliampere a.c. meters at hand to calibrate the thermionic meter. From the known data as to initial filament-current and emission characteristic it is perfectly feasible to get good calibrations by computation. The procedure followed for the 6-23 milliampere curve (curve 2 Fig. 3) was first to measure the bucking-out current through the 10,000-ohm resistor. This was 0.4 milliamperes. From the emission curve of Fig. 2, 0.4 milliamperes corresponds to 46 milliamperes, the initial filament current. This is I a.c. in the formula above. Then values of superimposed I a.c. were assumed — say 7, 10, 14, etc. milliamperes — and the resulting r.m.s. values figured. For instance, for 10 milliamperes of assumed a.c. I r.m.s. = j/(46)= + (10)2 ,/2116 47.1 inA. Going back to the emission curve, 47.1 milliamperes in the filament gives 0.52 milliamperes plate current. Although 0.4 milliamperes already flow in the plate circuit, the meter reads only .12 mA. because of the bucking-out current. Thus the meter reading of 0.12 milliamperes corresponds to 10 milliamperes of superimposed a.c. In the case of the 15-35 milliampere curve (curve 1, Fig. 3) there was no bucking-out current. The initial plate current of 0.02 milliamperes meant a filament current of 36 milliamperes. The combination with an assumed value of 33 milliamperes a.c. gives: I r-m.s. = V/C36) + (33)2 =,/i296 + 1089 = 48.8 mA. The plate current for 48.8 milliamperes . filament current is 0.75 milliamperes. Thus when the plate meter reads 0.75 mA., 33 mA., a.c. flows through the filament. Both the filament battery and the buckingout battery circuits on the meter set-up can be opened, and a.c. alone used to heat the filament— in which case currents from 35 milliamperes up can be measured. If lower current ranges are not necessary, the thermionic meter becomes simplicity itself. The raw materials required are only the tube, the onemilhampere meter, and a 22^-volt B battery. There are no restrictions as to d.c. path in the circuit in which current is measured. The wave shape of the current is immaterial — on any sort of wave the meter reads the r.m.s. value. Shunts may be used to extend the range indefinitely upward. The fact that a filament-current change of only 35 to 50 milliamperes gives full-scale change on the plate meter is both an advantage and a disadvantage. The limited a.c. range allows an open and easily read scale so that currents can be determined accurately. It also means, however, that an inconvenient number of shunts must be used to [give overlapping ranges. A possible alternative is to connect the grid of the tube to one side of the filament. By doing this the rate of increase of plate current . with filament current is cut down, and a range of approximately 35 to 60 milliamperes results. In other words, the minimum readable current is 58 per cent, of full-scale value as compared with 70 per cent, when the grid is tied to plate. However, this alternative method makes the plate current dependent in a greater measure on the plate voltage, so that changes in plate voltage damage the accuracy more than in the case of grid-plate connection. In Fig. 4 are given sample calibrations for the tube (grid connected to plate) when carrying a.c. alone, with and without shunts. These calibrations may be made with either d.c. or a.c. and then be used to measure any sort of a current. [Editors Note: Mr. Lampkin has indicated but briefly the uses to which such an instrument as he describes can be put. Anyone who has worked in the laboratory where small 40 10 E25 LU OS cc 3 20 lis LlI Hj < 10 nitial Curr€ nt= 0 .02 m 1. T2>Ii litial ( ^urren t = 0.4 mA. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 PLATE CURRENT-mA. Fig. 3 — Calibration curves of thermionic milliammeter 85 ao 75 < 70 E DC 3 60 <3 I 55 LU \— < 50 45 40 Shunt onFi ament =75 ohms 2>3hi nt = 150 o ims >NoE hunt "0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 PLATE CURRENT-mA. Fig. 4 — Calibration curves of thermionic milliammeter with two values of filament shunts. 1.6 1.4 12 1 1.0 | 0.1 ZD O S 0.6 3 D_ 0.4 34 36 38 40 42 44 46 48 50 52 54 FILAMENT CURRENT-mA. Fig. 2 — Filament-current, platecurrent characteristics of 199-type tube with 22\ volts on plate and grid a.c. potentials must be measured, either at low or high frequencies, will appreciate the advantages of this combination of tube and d.c. meter. As an example, let us try to measure the impedance offered to a 60-cycle current by a 30-henry choke coil. There are various methods, all of which are more or less complex. This impedance, however, is largely inductive reactance, and, if we knew the current through the coil at a given a.c. potential across it, this reactance could be calculated. From this calculation would come the value of inductance and impedance in which we are interested. At 30 henries, and with an a.c. potential of 110 volts, the current through the coil will be about 10 milliamperes. Now a thermo-couple that will measure currents of this value costs about $25 and requires a sensitive d.c. microammeter in order to read the rectified current. This meter will cost not less than $35 and probably will amount to $100. Therefore, in order to measure this small current of 5 to 10 milliamperes, equipment worth over $100 is required. The device described by Mr. Lampkin will measure this current easily and at much less cost than by the use of a thermo-couple and indicating meter. It is only necessary to put an initial current through the filament of the tube and then to add the current going through the choke. The differential of filament current will cause a differential in plate current which can be read on an inexpensive d.c. meter. After the tube and meter are calibrated, or when the values of plate current corresponding to certain values of filament current have been calculated, the meter is immediately useful. The change in plate current caused by the change in filament current can be obtained from a curve similar to those given on this page. Other uses for the device have been indicated in the box on the preceding j page. In all of the cases where a.c. and d.c. both flow through the device under measurement, care must be taken to prevent the d.c. current from flowing into the tube filament. This is a simple matter and requires only a large fixed condenser through which -the a.c. will pass but which offers a very high opposition to the flow of d.c. This milliammeter is one that can be built and operated by any home experimenter or any laboratory worker. The requirements are simple, a d.c. meter reading about one milliampere, a 60-milliampere filament tube, and a little patience at calculating what plate current will be read when a given a.c. cm/rent is added to the filament current. As the author points out the resistance of the voltmeter to the a.c. currents which it is designed to measure is of the order of 50 ohms. The effect of this resistance on the circuit in which this a.c. currents flow must be taken into account, but in general such an addition will not upset the circuit condition.] • march, 1929 page 326 •