Radio Broadcast (May 1928-Apr 1929)

One Method of Matching Coils and Condensers PRODUCTION TESTING WITH OSCILLATORS VARIOUS different types of apparatus have been devised and tried by the author in order to obtain accurate matching of coils and condensers with a minimum maintainence cost. These types have included aural and visual tests of all sorts, and the final system has been an adaptation of the beat-frequency audio oscillator. In a preceding article ("A Simple Unit for Measuring Impedances," September, 1928, Radio Broadcast) the importance of closely matching the mdividual coils and condensers in a radio receiver was stressed. In the beat-frequency audio oscillator we utilize two radio-frequency oscillators, one having a fixed frequency and the other a frequency varying from the fixed frequency up to any desired difference. For instance, one oscillator might be fixed at 500 kc. and the other vary from 500 to 510 kc. This produces a frequency difference between the two varying from zero to 10,000 cycles, and if these two oscillators are loosely coupled to a detector this beat will appear in the output as a variable audio frequency. This system has been used extensively to produce a compact audio oscillator with a continuous range and fairly constant output. The maximum frequency is obviously dependent only upon the frequencies of the two oscillators and the size of the variable tuning condenser. If both oscillators are tuned to very short waves and a large variable condenser is used, the beat can be made to go from low audio notes to radio frequencies far above the broadcast band. The lowest obtainable note depends entirely upon the degree of coupling between the two oscillators. If they are coupled closely they will "pull in" at comparatively high frequencies, and that will set the lowest obtainable note. To lessen this tendency it is customary to set one oscillator at a harmonic of the other. For instance, one might be fixed at 500 kc. and the other vary from 250 kc. to 255 kc. (the second harmonics being 500 to 510 kc), thus producing a 10,000-cycle maximum beat. To adapt this device to production testing in a manufacturing plant we see that the audio beat note will pass through zero when the two oscillators are identical in frequency. To match two condensers we merely need to make up two oscillators whose frequencies depend upon the capacities of the condensers under test, and if they match we will get a zero beat. This is the principle of this [system applied to condenser matching. The apparatus (Fig. 1) consists of two oscillators, a detecting system, and an audio amplifier. One unit of the condenser gang is connected to one oscillator and the other oscillator is connected alternately to each of the other units of the gang by means of switches. A small compensating condenser, C4, is connected across the tuned circuit of one oscillator and a variable midget condenser, C3, across the other. These are adjusted so that when the vernier, C3, is set at mid-scale and the two condensers under test are equal, a zero beat will result. If the condenser C2 is greater than Ci, the dial on the vernier C3 will have to be moved to one side of midscale to reestablish zero beat. The amount of motion necessary is a measure of the unbalance between Ci and C2. The vernier, O, should be about 35 mmfd. maximum capacity, and for such a condenser one division on a hundred-division scale corresponds to By RICHARD F. SHEA approximately three-tenths of one mmfd. Let us now take these parts and dispose them to make a satisfactory production tester. To fill this purpose the apparatus must be sturdy, reliable, and quick of operation. The whole should be mounted on a base of strong steel and should have a jig so that the condenser gang will mount quickly and securely on the base and always in the same position. This latter consideration is important as the position of the gang makes a great deal of Mr. Shea describes in this article an instrument suitable for use in the production testing of coils and condensers for use in manufactured receivers. The author has been connected with the radio division of American Bosch Magneto Company. Obviously the sensitivity and selectivity of a modern single-control receiver depends to a large degree on the accuracy with which the coils and tuning condensers are matched. — The Editor. difference in the stray capacities which are shunted across the gang. All the wiring should be of heavy bus to insure remaining in place and great care must be taken to make it uniform for all the units of the gang. The drawing of Fig. 2 suggests a successful layout. The coils Li and L2 must be identical, and to insure this fact they should be measured for inductance in the shields. The capacities of the tubes can be compensated by the adjustment of C4, so that the matching of Li and L2 and the uniformity of leads are the only strict requirements. To adjust this device place the condenser gang on the jig and set it at zero, i.e., minimum capacity. Set the vernier, C3, at 50° (on a 100° scale) and adjust C4 untd you get a zero beat. To check this and also the uniformity of the wiring, test about ten gangs and note the readings on all three switch positions. If it is found that the average in all cases is 50° then the adjustment is correct. If, however, all averages are off the same amount from 50° then C4 should be changed to bring them to 50°. If the average on any one switch position is off it means the wiring is non-uniform, and it can be fixed by pushing the bus wire nearer to or farther from the base. When the tester is correctly adjusted the average of a larger number of gangs should be very close to 50° on C3. We are now ready to set limits. If our gang has a capacity of 20 mmfd. at minimum, 200 mmfd. at mid scale, and 500 mmfd. at maximum, and we wish to hold units to within 1 per cent., then our allowance is 0.2 mmfd. at 0°, 2 mmfd. at 50° and 5 mmfd. at 100°. In setting these limits it must be borne in mind that the strays also add to the capacity, and the total minimum will be 40 to 50 mmfd., so that 0.2 mmfd. is holding them much closer than 1 per cent. In such a case 0.5 mmfd. would be the proper tolerance at zero, 2.25 mmfd. at 50°, and 5.25 mmfd. at 100°. If our vernier has 0.3 mmfd. per division this becomes approximately two divisions at zero, 7 divisions at 50°, and 17 divisions at 100°. Then in testing our gang we set it upon the jig and turn on switch S]. This matches condenser one with condenser two. We set the gang at zero and rotate C3 to get zero beat. Let us say it comes at 51°. Snap on switch S2 and readjust C3. This time we get 50°. Lastly turn on switch S3 and here we get 49°. This gives us 50°, 51°, 50°. and 49° for the four units, and as no two readings differ by more than 2 divisions the gang comes within the limits specified above. A similar process at 50° and 100° gives us the deviations at those positions. So much for condenser matching. Now let us turn to the application of such a system to coil matching. Here we have a much simpler problem and are able to obtain even greater accuracy for reasons which will soon become evident. 2 STAGE AMPLIFIER UX201-A UX201-A Note: C| goes directly to first Condenser, C2 is switched as indicated. C j — C 2 Condensers to be matched. C3 35 mmfd. Vernier. -C4 40 mmfd Neutralizing Cond. L!" Mf" L2-L2' 90 turns No.26 on 2"Coil, broken at 30 turns for Plate Coil. Coils in Copper Cans 3"diam. To Frame Fig. 1 — Circuit of beat-frequency oscillator for testing gang condensers • april, 1929 . . . page 387 • ' />