Page 325 - Microsensors, MEMS and Smart Devices - Gardner Varadhan and Awadelkarim

SAW DEVICE DEVELOPMENT AND HISTORY 305

9. Easy to calibrate
10. Low cost
11. No moving parts that can suffer from wear mechanisms

However, in practice, sensors are less than ideal, and suffer from problems associated
with cross-sensitivity caused by poor selectivity.
To understand the fundamental sensing mechanism of a SAW-based acoustic sensor
better, a brief description follows (Morgan 1978): The sensing mechanism of a SAW
sensor relies on a change in the acoustic velocity of a bulk (or surface) wave within
(or on) the surface of the piezoelectric substrate. After transduction from acoustic to the
electrical domain, the frequency of the acoustic wave is measured. Another parameter that
can also be measured is the amplitude of the acoustic wave, or to be more precise, the
amplitude of the electrical oscillator frequency. This measurement is usually associated
with a decrease in acoustic wave amplitude, and this reduction is a measure of crystal
damping, thus giving an indication of viscous-damping effects.
In the development of acoustic wave sensors and their corresponding oscillators,
complementary instrumentation or systems must be developed in order to measure both
the dynamic and steady state sensor response. Specialised and expensive instrumenta-
tion, such as network analysers or vector voltmeters, can give a very detailed analysis of
acoustic transducers or their sensor configurations. However, this type of instrumentation
is far too sophisticated for general-purpose routine laboratory use. Moreover, it not only
requires highly qualified personnel for its operation but also is very limited in that only one
acoustic sensor can be monitored at a time. In addition, analysis of the data is extremely
time consuming, thus making it impractical to use such instrumentation for measuring
the signals from a microsensor array. Consequently, the vast majority of research papers
have based measurements of the frequency of acoustic wave sensor oscillators on the use
of frequency counters and chart recorders.
Other important design goals that concern advanced measurement systems are sensor
response and data analysis. In the present system design, this involves the transfer of data
directly to disk file, where postprocessing and analysis is carried out using commercial
software packages, such as MathCAD or Matlab. In the case of data from an oscilloscope,
the data are downloaded, and standard conversion software is invoked to convert the
resulting data from a waveform format to a form that can be used as suitable input for the
software packages mentioned earlier. As experimental data can be saved on disk, this will
contribute toward the ease of documentation. Future directions for the system are aimed
at processing and analysis being completed in real time. These can be achieved by the
inclusion of the necessary software classes into the existing system software program or
through dynamic data linked with other commercial software packages. In addition, system
calibration has to meet the requirements of the specific application. This can be achieved
by making use of the mathematical models of the sensor response. A calibrated system
will aid the user, as both the graphical and numerical display will present data in the
appropriate and more meaningful physical units rather than frequency (Campbell 1998).
Before any attempt is made to initiate a proposed system design for acoustic microsen-
sors, it is mandatory to thoroughly understand the sensing mechanisms and limitations of
acoustic devices for each particular sensing application. Using equivalent electronic circuit
models, a greater understanding of these transducers can be achieved. External factors that
cause changes to the circuit model also have to be considered. Some equivalent circuit