Radio Broadcast (May 1928-Apr 1929)

Data on Design and Operation A HOMEMADE THERMIONIC MILLI AM METER IT IS more than ordinarily difficult for the radio worker to measure alternating currents around or below 25 milliamperes. A necessity for determination of currents in this range usually leads to a contemplation of vacuum thermo-couples and d.c. microammeters. Full-scale ranges of 100 milliamperes can be had rather reasonably in selfcontained thermogalvanometers; but because of the current-squared crowding of the scale, readings below 25 milliamperes cannot be reliably taken. It is useful, then, to know that a vacuum tube can be fitted up rather simply to measure alternating currents from some 5 milliamperes upwards. A 199-t.ype tube, a onemilliampere d.c. meter, and a 30-henry choke are the chief accessories necessary to measure currents in the above mentioned range. In the range from 35 milliamperes to as high as desired only the tube, d.c. meter, and appropriate shunts are needed. Calibration of a home-made meter usually necessitates the use of a standard meter for the same kind and range of current. In this case, however, only a 0-100 d.c. milliammeter is required, and it can be improvised from the 0-1 meter if necessary. Design of Meter THE principle utilized in the thermionic meter is the change in emission current of a tube due to change in filament-heating current. Fig. 1 is the circuit diagram. The 0-100 milliampere meter is not a permanent part of the set-up, but is used only to obtain the initial filament-current, plate-current characteristic. Fig. 2 shows this curve for a typical 199-type tube. Appreciable emission is not had until some 35 milliamperes flow in the filament. It is evident 5 or 10 milliamperes of a.c. alone would have no effect whatsoever in producing emission current. The scheme, then, to measure alternating currents in this lower range is to pass an initial d.c. through the filament, and on this current superimpose the a.c. which is to be measured. The a.c. alternately adds to and subtracts from the steady filament current. If the temperature of the filament could instantaneously follow these current fluctuations, the plate current would swing up and down the emission-current curve of Fig. 2. However, because of heat capacity, thermal lag, frequency of fluctuations, and such, the temperature of the filament cannot follow the heatingcurrent fluctuations. What happens is that the temperature takes up a new value dependent on the new root-mean-square value of the heating current. The average value of the current, of course, does not change — but the average value plays no part in determining the filament temperature. The r.m.s. value of Parts required for construction of a 35-mA. thermionic millammeter By G. F. LAMPKIN the steady initial current is simply its d.c. value. The new r.m.s. value caused by the combination of the a.c. with the d.c. can be calculated by: Ir.m.s. = l/(Id.c.) + (la c ) This change in heating value of the filament current is evidenced by a change in the d.c. plate current. By suitable calibrations, and The author of this article, who is no stranger to the readers of Radio Broadcast, has used the meter described to measure the overall frequency characteristics of receivers, the currents into loud speakers, the a.c. in power-supply chokes whose inductance was being measured, and, as he says, once one has an instrument that will measure accurately small values of a.c, many other uses will be found for it. It is much less expensive than a combination thermo-couple and microammeter — and repairs are less costly, too. — The Editor fortunately by calculations, the magnitude of the superimposed a.c. can be determined from plate-current readings. In the set-up of Fig. 1 the 30-henry choke is necessary to insure that the a.c. takes its intended path through the tube's filament. The filament battery has a low a.c. resistance, and were it not for the choke it would bypass the larger part of the a.c. The usual d.c. resistance value of a 30-henry choke is approximately 300 ohms, which makes necessary a filament battery of 22| volts. The filament battery should be of the heavy-duty B type, for it must supply a current of 40 or 50 milliamperes. A 400-ohm rheostat or potentiometer is used for filament control. The filament battery could be made to serve also as a B battery, by returning the anode connection to the positive terminal. However, the comparatively heavy load which the battery must supply tends to cause a rather rapid drop in voltage; and, although the battery remains entirely suitable for A supply, the changed plate voltage leads to inaccurate results. For this reason, another battery is used to supply the plate voltage. One additional battery, a 4f-volt unit which supplies bucking-out current for the plate meter, is also used. When obtaining the emission characteristic of Fig. 2, readings are taken up to one milliampere plate current, the full-scale range of the meter. Then a 50-ohm rheostat, which is connected across the plate meter, is adjusted until the deflection is exactly half its original value. This will require a shunt of approximately 39 ohms on a Jewell meter. Beadings up to 2 milliamperes are taken, because the curve in this region is necessary for calculating calibration points. The sensitivity of the thermionic meter depends on what part of the tube's filament-cmrent, plate-current characteristic is worked. It may be seen that a given change in r.m.s. filament current at a low initial current, say 38 mA., will produce only a fraction of the plate-current change that would be caused at a higher initial current 60 mA., for example. The resistance in the circuit diagram of Fig. 1 through which bucking-out current is fed to the plate meter is fixed. The only control is the 400-ohm filament rheostat. The procedure in using the meter is to close the battery switch and adjust the filament rheostat until the plate meter reads zero. This will happen when the emission current equals the bucking-out current supplied through the 4j-volt battery and fixed resistor. Thus the size of the fixed resistor automatically determines what initial emission current must flow and accordingly what part of the filament-plate current characteristic is to be used. If the value of the fixed resistor is low, the bucking-out current will be large. To produce an equally large emission current a high initial filament current will be necessary, and the resultant sensitivity of the meter to a.c. will be good. However, extreme values of initial filament and emission current are detrimental to the tube's life and the accuracy of calibration and so should be avoided. The calibration curves of Fig. 3 were made by measuring the superimposed 60-cycle current on the tube's filament, and taking the corresponding plate-current readings. With a 2500-ohm resistor in the bucking-out circuit, the initial plate current, which was required to equal the bucking-out current, was 1.6 milliamperes; and the superimposed a.c, which gave a one-milliampere plate cmrent range, was 3 to 18 milliamperes. Greater sensitivities can be had by working higher on the emission curve, i.e., with smaller values of bucking-out resistance, but such is not advisable. A 10,000-ohm fixed resistor and an initial plate current of 0.4 milliamperes gave a range of 6 to 23 milliamperes a.c. (curve 2 of Fig. 3). With the bucking-out circuit opened, and the plate current adjusted to an initial current of 0.02 milliamperes, the a.c. maximum was 35 milliamperes (curve 1.). The calibration on the latter range departs much farther from the linear than does the one for 6-23 milliamperes. The lower range calibration works over a more restricted and a straighter part of the emission characteristic, so that it does not show the sharp bend present in the 15-35 milliampere curve. It is important to note that there must not exist a d.c. path in the circuit which carries the to-be-measured a.c. A d.c. path would draw current from the filament battery of the tube and disarrange the zero setting. In View of the author's thermionic miHiammeter set-up 0 march, 1929 page 325 •