684 
IV, CALIBRATION EQUIPMENT 
In the use of any system for recording transients, it is necessary to 
have an accurate calibration of its sensitivity and time resolution. It is 
also highly desirable to have a means of determining its transient response 
to an input signal of known form. In recording piezoelectric-gauge signals 
with long coupling cables, it is also necessary to determine the character- 
istics of these cables. 
It is common practice in much oscillograph recording to calibrate both 
voltage and time scales by means of a sine wave of known frequency and ampli- 
tude determined by an a=-c voltmeter of some sort. While this combination of 
functions is economical of time, it has the disadvantage that neither cali- 
bration is made in a fundamentally satisfactory manner for precision meas- 
urement. A sine wave is not ideally suited for interval measurements, as 
it is a smooth curve without discontinuities or other convenient reference 
points, and it is difficult to maintain accurate voltage calibrations of 
a=-c meterSe 
It has been standard procedure at this laboratory to attack the cali- 
bration problem in a different way. The timing signal is shaped to have a 
sharply peaked wave form easily and accurately measured. No attempt is made 
to use this signal for voltage calibration, but rather a step voltage of 
known amplitude is applied separately, This step voltage is not only easily 
measured if the recording system has a good response, but also serves as a 
calibration of the transient response characteristics of the system, This 
additional information could be obtained from steady-state signals only at 
the expense of using a wide range of test frequencies, a procedure not 
practical in field work. 
The concept of a routine calibration and transient-response test can be 
extended to include the impedance characteristics of any coupling network 
between the gauge and amplifiers. This is done, for example, in the "Q-step" 
12/ 
calibration principle,— 
Sy TEEPE UU PURER SSE 
—‘The Q-step principle, and its application to cable response testing, 
is described in NDRC Report A-306 (OSRD-4561). A brief description of the 
method is included for convenience in Sec. 6(b). 
in which, in effect, a known quantity of charge, 
=! hg = 
