V. Matko et al., ON THE USE OF QUARTZ-CRYSTAL CAPACITIVE DEPENDENCE FOR MEASUREMENT OF0-1 ML VOLUMES, Sensors and actuators. A, Physical, 42(1-3), 1994, pp. 465-471
The use of a differential oscillator sensor structure in a capacitance
sensor is presented. Investigations are focused on the design and ope
rational analysis of an oscillator differential structure in which the
oscillation frequencies of the two oscillators are very close, and on
the application analysis of capacitive-dependent crystals. In additio
n, the excitation of the entire sensor with stochastic test signals ha
s been analysed by the correlation deconvolution method, also known as
the direct digital method (DDM). The compensation of temperature and
voltage influences, as well as disturbing noise signals, is included.
The area of operation and the uncertainty of the sensor with and witho
ut the test signal are given as well. When designing the capacitance s
ensor, the problems regarding the source of stable oscillation, temper
ature compensation, the influence of supply voltage, noise, and A/D an
d D/A conversion occur in the operation range under 1 pF. The pulse-wi
dth module, which forms pulse-width-modulated high-frequency current p
ulses, is the proposed solution. With these pulses the capacitor in th
e integration element is charged or discharged. In this way we benefit
from the fact that the capacitor's voltage increases linearly if it i
s charged by a constant cuff ent. As the charging is affected only by
the current pulses, which require an adequate current, the disturbing
noise signals do not affect the capacitor charging. Likewise, the puls
e-width module compensates the effects of temperature and voltage by m
eans of modulation. The correlation determination of the measuring val
ue is of prime importance for the determination of end values. Two com
puter-aided modes of operation are suggested: dynamic measurement cont
rol and the correlation determination of differences. Several experime
nts have been carried out to investigate the method's possible applica
tions. The experimental results of 0-1 ml volume measurements are show
n. The method is linear in the range of work and ensures an uncertaint
y below 0.01% in this range. The volume-measurement uncertainty (0-1 m
l) is less than 0.05% (T = 15-25-degrees-C).