178 



MEASUREMENT OF PRESSURES 



the circuit. For tourmaline gauge designs discussed in the preceding 

 section, Co has values of the order 10 to 200 /xyuf. Values of R greater 

 than 10^° ohms are realized only by very special precautions and design 

 of the complete circuit, and with R = 10^° ohms, Co = 2 X 10~^° farad, 

 the value of RCo is two seconds. This time can, of course, be increased 

 by using padding capacitance C at the expense of decreased terminal 

 voltage. In practical cases, leakage resistance and circuit sensitivity 

 requirements limit the time constants conveniently obtainable to times 

 measured in seconds rather than minutes. If the indicated voltage V 

 is to be a direct measure of the EMF Vo it is evident that the times re- 





t»R (C + C^) 

 (Q) SCHEMATIC ORCUIT (b)RESPONSE TO STEP PRESSURE 



Fig. 5.14 Piezoelectric gauge circuit and long time transient response. 



quired to establish the pressure and measure V must both be a small 

 fraction of the time constant, say one per cent; as a result, slow pump- 

 ing up of pressure and determination of F by a slow period meter are 

 both ruled out. 



The impractical requirements of static methods can be avoided by 

 use of what may be called a quasi-static method, in which the gauge is 

 allowed to come to equilibrium with an applied hydrostatic pressure; 

 this pressure is then rapidly released by manual opening of a valve or 

 by bursting of a diaphragm. ^^ The resulting voltage is then measured 

 by a high impedance voltmeter circuit, such as an electrometer type 

 circuit, or by a recording oscillograph. Methods of this type permit 

 pressure changes to be produced and their effects indicated in times of 

 a few seconds or less; the time constant requirements can be conven- 

 iently realized, and the pressure level can be determined by standard 

 methods. 



Although these quasi-static methods might appear to satisfy ade- 

 quately the primary requirement for a calibration, they suffer from two 

 sources of error which impair their usefulness. The first of these is the 

 phenomenon of cable signal (discussed in more detail in section 5.8), 

 which manifests itself here as an electrical signal developed by electrical 



" Sec, for example, the reports by Borden (9), and by Arons and Cole (3). 



