men FREQUENCY AMPl.iriERS 203 



tioii it may 1)0 iiotfd tliat an exiflli'tU test for the pri'soiicf of iiruU-sir- 

 al>lf pick-up is the closing of a switch {S) placed at the input of the 

 anipiitier. With this switch closed there should he no appreciable 

 input to the detector. 



Direct high frequency' amplification measurements require input 

 units made up of very carefully constructed attenuation boxes or 

 potentiometers and well shielded oscillators. Such units have been 

 developed in connection with field measurements and are described 

 in a pai)er on "Radio Transmission Measurements," by Messrs. 

 Bown, Knglund and Friis and "Note on the Measurements of Radio 

 Signals." ' by Knglimd. 



On the right in Fig. 3 is shown, as mentioned before, the circuit 

 diagram of a "tube voltmeter" such as is used in many high frequency 

 measurements. The tube voltmeter is essentially a plate current 

 curvature detector. The grid is made negative by means of the grid 

 battery £c, so that the normal plate current of the tube is very small 

 (of the order of 50 microamperes or so), and this plate current is 

 further balanced out by means of the potentiometer arrangement 

 Rt. Ri. so that the plate current meter reads zero when the input to 

 the tube voltmeter is short-circuited. This arrangement has the 

 advantage of making it p)ossible to utilize the entire scale of the 

 meter and to obtain the measured voltage from a single reading 

 instead of the difference of two readings. Such a tube voltmeter 

 built with an "N" tube will give a deflection of 1 microampere for an 

 input voltage of about 1/5 of a volt, and the calibration will stay 

 remarkably constant for several months and is independent of the 

 frequency at which it is calibrated. The values of the resistances in 

 the resistance boxes used at high frequencies may, therefore, be 

 checked by using the boxes for calibrating a tube voltmeter first 

 at 60 cycles and afterwards at, for instance, 1,200 kilocycles. If the 

 two calibration curves obtained are exactly identical, then the resist- 

 ance has not changed appreciably within this frequency range. 



In measuring the amount of "feed-back" amplification in a receiving 

 set, it is not possible to use a method as direct as described above. 

 The "feed-back" or regeneration in a set is, as already mentioned, 

 due to the coupling between the grid circuit and the plate circuit of 

 the tubes through the grid-plate capacity, and will depend upon 

 both the load in the plate circuit and the nature of the input circuits. 

 If. for instance, it is desirable to measure the amount of "feed-back" 

 amplification due to the coupling between the loop circuit and the 



• Proc. Inst. R. E., Vol. 11, No. 1, Februarv-, 1923. Proc. Inst. R. E.. Vol. 11. 

 No. 2. .April, 1923. 



