832 
32 AS EB 
application resolves itself into a search for a suitable 
compromise among these factors, and sometimes the 
final design cannot satisfy all the requirements to the 
extent desired. It has been established that tourmaline 
is satisfactory with respect to linearity, freedom from 
hysteresis, and stability of its pressure sensitivity, but 
there remains the problem of arriving at a design ade- 
quate from the points of view of sensitivity, frequency 
response, and ruggedness—all factors which are to some 
extent under the control of the designer. 
Piezoelectric crystals develop polarization charge 
when subjected to pressure changes. This charge dis- 
tributes itself over the capacitance which is in parallel 
with the gauge element, and consequently the amplitude 
of the voltage appearing at the amplifier input is not 
only directly proportional to the applied pressure change 
but also inversely proportional to the total capacitance 
parallel to the gauge element in the input circuit. 
The necessary gauge sensitivity must therefore be 
evaluated in terms of the following controlling factors: 
(1) magnitude of pressure variations to be recorded, 
(2) total capacitance of the gauge circuit (this depends 
principally upon the length of cable required to connect 
the gauge to the amplifier or preamplifier), and (3) 
amplifier sensitivity. 
A lower limit is sometimes placed upon the required 
sensitivity by the presence of spurious signal arising in 
the cable or the mounting of the gauge. In many appli- 
cations, a piezoelectric gauge is connected to recording 
equipment by means of shielded cable, a certain length 
of which may be unavoidably exposed to the pressure 
wave as the latter advances from its source. Spurious 
signal arising in such cases is an integrated effect which 
generally introduces little error into the determination 
of initially discontinuous or rapidly changing pressures, 
but adds progressively more signal as the wave en- 
counters more cable. It is evident that measurements of 
small pressures behind an initial high pressure region 
may thus become highly inaccurate and an effort to 
obtain reliable values of impulse by integration of the 
pressure-time curves may be entirely vitiated. 
Experience has demonstrated that most ordinary 
commercial rubber microphone cables and polyethylene 
cables are completely inadequate since the signals they 
develop when subjected to pressure variations are very 
large and may even be, under some conditions, larger 
than those which would be produced by the gauge ele- 
ments themselves. 
A number of cables were developed which proved to 
be adequate, and one type—a copper tubing cable, 
originated at the David Taylor Model Basin—was used 
successfully for sevetal years in under water studies. 
More recent developments have led to the production of 
a polyethylene cable (designated F.O. 5879 by the 
Simplex Wire and Cable Company) which appears to be 
superior to any other cables previously available and 
which is equally satisfactory for measurements in both 
gaseous and liquid media. However, even in this cable, 
ARONS ANDY R- He 
COLE 
the signal has not been completely eliminated (although 
it has been very greatly reduced) and one must always 
make sure that the gauge signal is large relative to the 
residual cable signal in each application. 
Although much work in the past has been done with 
single-ended systems using coaxial cables,’ it is well 
known that push-pull or balanced systems effect ma- 
terial reduction in residual cable signal by canceling out 
components which are symmetrical about ground. It 
seems unlikely, however, that such a system would 
eliminate the difficulty entirely, since there is no reason 
to expect the cable signal to be a perfectly symmetrical 
phenomenon. 
The sensitivity of a piezoelectric gauge is directly 
proportional to the available sensitive area of crystal. In 
cases where spurious signal due to cable or gauge mount- 
ing is a significant factor, the permissible lower limit of 
sensitivity imposed by such signal places a lower limit 
upon the size of the gauge. On the other hand, the dis- 
cussion of high frequency response and pressure-field 
distortion given in Section V shows that higher fidelity 
is obtained with smaller linear dimensions of the sensi- 
tive unit. Specific requirements as to accuracy with 
respect to high frequency response consequently put an 
upper limit on the linear dimensions of the gauge, par- 
ticularly the dimension along the direction of propaga- 
tion of the wave, and this upper limit must be reconciled 
with the lower limit indicated by the necessity of 
swamping spurious signal. In some cases the two re- 
quirements cannot be reconciled, and one or the other 
must be compromised in designing the gauge. This is 
particularly true where it is desired to measure very 
rapid pressure variations in liquid media. In gaseous 
media where the high frequency response requirements 
are usually less stringent, large gauges can be used and 
the above requirements can generally be reconciled. 
In the actual construction of gauges, a given sensi- 
tivity can be obtained and the linear dimensions 
minimized by stacking two or more crystal plates in 
parallel. This feature has been utilized in all UERL 
gauges, but, except for a few special cases, it has not 
proved practicable to use gauges having more than four 
plates in the pile. 
B. Construction of Gauges 
The object of the UERL group was to develop a unit 
suitable for measurement of explosion-produced shock 
waves in air and water. Many of the basic principles, 
however, apply to measurement of transient pressure 
waves in general, and properly modified UERL-type 
gauges should be useful under a wide variety of cir- 
cumstances. 
The principal difficulties with the early gauge designs 
proved to be lack of stability of pressure sensitivity and 
prevalence of large systematic discrepancies between 
3R. H. Cole, “The use of electrical cables with piezoelectric 
gauges.” OSRD Report No. 4561. 
