952 



Popular Science Monthly 



the connections in first diagrams. The 

 ammeter will read correctly, but the volt- 

 meter records the voltage-drop across both. 



This arrangement is very much in error 

 because the drop in the ammeter is greater 

 than in the resistance. In problem 2 the 

 voltmeter is connected across the resistance 

 only, as in second diagram. The current 

 flowing through the voltmeter will be very 

 small compared to that flowing in the 

 resistance. The per cent of error in the 

 ammeter is small. The voltmeter in the 

 connections shown will read correctly. 



A simple problem may serve to make 

 clear the described method. What is the 

 correct resistance of a 40- watt lamp? The 

 ammeter is a milliammeter, resistance of 

 .045 ohm. The voltmeter has a resistance 

 of 15,000 ohms. The voltmeter reads 

 no volts and the ammeter .3636 amp. 



SOLUTION: 



Problem i. Voltmeter Across Resis- 

 tance. 



The current through voltmeter, by 

 Ohm's Law, equals 

 no 



= .00734 ampere. 



15000 



The current through resistance equals 

 •3636 — .00734 = .35626 ampere. 



Per cent of error, 

 .0073 



= 2.43 per cent. 



.35626 



The voltmeter records correctly. 



Problem 2. Voltmeter Across Resistance 

 AND Ammeter. 

 Voltage across ammeter, 

 .3636— .045 = .1638 volt. 



Voltage across resistance, 



iio.oooo — .1638 = 109.8362 volts. 



Per cent of error, 

 .1638 



.149 per cent. 

 109.8362 



The ammeter records correctly. 



The ammeter in problem i reads too high 

 by 2.43%, or about 1.51 ohms. From the 

 arrangement in problem 2 the voltage reads 

 too high by only .149%, or an equivalent 

 of .0595 ohms. 



From the foregoing method a rule may 

 be summed up as follows : When measuring 

 a high resistance, connect voltmeter across 

 resistance and ammeter both ; when meas- 

 uring a low resistance, use voltmeter across 

 resistance only. — T, E. Martin. 



Increasing the Efficiency of the 

 Pouisen Arc Generator 



IN THE past the Pouisen arc generator of 

 undamped or sustained waves has 

 usually been used directly in series with the 

 antenna in which it operates. The effi- 

 ciency of this arrangement is variously 

 stated as from 15 per cent to 50 per cent, 

 depending largely upon the constants of the 

 circuits and the details of design of the arc 

 generator itself. The wide use to which 

 generators of this type have been put in the 

 past few years, in the United States Navy 



and elsewhere, 

 has caused a 

 good deal of 

 study to be put 

 upon the sys- 

 tem with a view 

 toward main- 

 taining the 

 highest effi- 

 ciency. By the 

 addition of 

 condenser and 

 inductance cir- 

 cuits it has 

 been found 

 possible to in- 

 crease materi- 

 ally the output 

 of a single arc oscillator, and some results 

 have been secured which would appear 

 most surprising to those who are familiar 

 with only the older forms of the apparatus. 

 United States patent number 1,179,353, 

 issued in 1916 to L. F. Fuller, shows one 

 method of increasing the radiation or 

 antenna current produced by an arc 

 generator. As indicated in the figure, the 

 only material change is that of adding the 

 two condensers i and 2. The first of these 

 is placed directly in series with the arc 3 

 (which is supplied with power through the 

 leads 4 and 5 in the usual way), and the 

 second is shunted about the arc and the 

 first condenser. Above the arc and shunt 

 circuit is connected the loading inductance 

 6, with a small portion arranged to be short- 

 circuited by means of the signaling key; 

 the high potential end of this coil is led 

 directly to the aerial 7. The capacity of 

 the two condensers is adjusted by trial, 

 and depends upon the size of the antenna 

 and the power of the arc. When the best 

 values are used, the antenna current 

 is often more than doubled without using 

 any more power than normally. 



New arc circuit which 

 increases the radiation 



