Radio Broadcast (May 1928-Apr 1929)

'Z. RADIO BROADCAST ZJO o-cc So U)Q -5 >i Si 5l£ O 1000 500 200 100 50. SE1 >JSITIVI TY Ant. WOV-uf 20 ohm . Xnput io jst 1 100 200 300 400 500 600 700 WAVELENGTH Fig. 3 100 50 ! 20 ! io 5 b 1 SELECTIVITY ✓ ✓ -% s \ > K500 Me lers / / / / " \ k. N / / t / V \ \ 1 1 1, '/ /SOO A leters -\ ~ V \\ II II II If — \» — — H — il Ml 30 20 10 0 10 20 30 SHORTER WAVES LONGER WAVES KILOCYCLES OFF RESONANCE Fig. 4 typical design has been described by Rodwin and Smith. (Proc, I.R.E., Feb. 1928, p. 155). It is open to certain objections as to both convenience and accuracy. The current in the generator coil must be varied over a wide range, necessitating either a series of thermal meters or a radio-current transformer, and the mutual inductance between the coils may be modified by different amounts at different frequencies, by adjacent shields, or other conductors. On the other hand, the method has the advantage of impressing a field directly upon the loop designed for use with a receiver under test, instead of impressing a voltage in series with the loop, from which the equivalent field must be calculated. In accordance with the second method, the test voltage is developed across a small known resistance which terminates a resistance attenuation network fed by a measured radio current. This method presents the general advantage of allowing all current-carrying impedances to be buried in shields, exposing only a single terminal which is above the shield by the amount of the test voltage. It allows a rating which is directly expressible in field strengths for normal signal in the case of an antenna-operated receiver. An outfit which I think is interesting in that it is moderately portable and yields results comparable in accuracy with more bulky equipment, is the one which I am about to describe. A Standard Signal Generator 'T'HIS outfit was developed to fulfil four conditions: (1) A portable source equipped for use with external, unshielded batteries. (2) A range of output voltages from one microvolt up, with sufficient shielding to prevent the induction by stray fields of voltages in any adjacent tuned circuit comparable with the output voltage. (3) An accuracy well within the consistency of measurements with highly stable receivers. (4) The whole outfit to be reproducible by ordinary skilled shop labor. A diagram of the circuits employed is shown on Fig. 1. A single audio oscillator tube is provided within the apparatus, for modulation at a fixed frequency of about 400 cycles. This is the frequency normally used for the most common sensitivity and selectivity measurements. This oscillator includes the tube shown at the left of the drawing and the iron-core transformer tuned by a fixed condenser. This transformer feeds a modulation transformer through a resistance voltage divider marked "Modulation Control." The audio voltage is impressed by the modulation transformer through a one-to-one ratio upon 200 uj 2 £ ti a 100 g<? 80 08 60 o Z = ™ 40 => o N 20 =3 a. 1 TDE LITY Meters / N 500 Meter? \\ A 60 100 400 1000 5000 10,000 MODULATION FREQUENCY Fig. 5 -10,000 3 I 5000 c 2000 c : 1000 3 3 500 L 200 o > o o 5 100 80 1 SENSITIVITY ^ It V 4° A s — 100 200 300 400 500 WAVELENGTH Fig. 6 600 700 '-5)100 E« 80 j " 3(2*60 "o o 40 20 FIDI XITY Meters 51 — a\ 0 Meters ^\ — A 60 100 400 1000 5000 MODULATION FREQUENCY Fig. 7 the plate circuit of the radio oscillator tube, and is measured by a thermal voltmeter comprising a resistance, a thirty-ohm thermocouple, and a panel-mounting d.c. galvanometer shown at the lower part of the figure. A low pass permits the use of an unshielded external audio oscillator which may be positioned anywhere with respect to the signal generator and the receiver under test, and which may be connected to the signal generator through unshielded leads. The radio oscillator tube has a "parallel feed" plate circuit consisting of the secondary of the modulation transformer and a radio-frequency choke coil in series with the positive B battery terminal and the plate. The tuned circuit of the radio oscillator consists of a " vario-coupler " inductance which is connected by a metal belt to the variable timing condenser, both being • february, 1929 . . . page 231 « operated by a tuning dial on the front panel. A small variable condenser is provided in shunt with the main condenser for fine tuning adjustments. The tuned circuit is closed through an attenuator, which is bypassed to ground by a non-inductive variable resistance marked "Radio Control." This resistance thus furnishes a means for adjusting the modulated radio-frequency current flowing into the attenuator. The current which passes into the attenuator is measured on a four-ohm thermo-couple connected through a twin twosection filter into a panel-type d.c. galvanometer which is exposed on the front panel of the outfit. The output end of the attenuator terminates in a two-ohm non-inductive slidewire which is connected to the output terminals on the front panel. This slidewire consists of a short piece of No. 38 manganin wire stretched over a copper return path with an insulation strip 0.01 inch thick between them. Fig. 10, is an external view of the outfit. The various instruments and controls will be recognized from the description previously given. The external dimensions are 17 x 15 x 12 inches. Fig. 2 shows a conventional method of connecting the signal generator through a local or dummy antenna to a receiver under test. For the sake of completeness an external audio oscillator is shown. The receiver may be positioned at any convenient point near the source and twisted leads a foot or so in length do not introduce an appreciable error since the impedance at the generator end is never more a. 3 z o ,P CO _l < O Q > z o < a: h o <s> ^ OS 100 . 50 20 10 5 SENSITIVITY V V \ V *4 100 200 300 400 500 600 700 WAVELENGTH Fig. 8 1000 500 100 50 > > J c 10 > ! 5 1 SELECTIVITY j hoo Me ters 30 20 10 0 10 20 30 SHORTER WAVES LONGER WAVES KILOCYCLES OFF RESONANCE Fig. 9