Sensitive and accurate pressure transducers are useful in many areas of biology. Two common applications are the measurement of pressure changes within the respiratory system of an animal, and quantification of the small pressure fluctuations that accompany the humidification and warming of inspired air in whole-body closed-system or open-system flow-through plethysmography (see Malan, 1973; Bucher, 1981). The design of such transducers is often complicated by the requirement to measure very small pressure fluctuations in the presence of a substantial pressure offset. Moreover, such pressure offsets often vary slowly with time. Frequent adjustments are then required to keep the output of the pressure transducer from saturating in either direction.

Various transducers have been used to detect small pressure changes. Most commonly, a flexible diaphragm attached via a lever system to a strain gauge has been employed (see, for example, Bucher, 1981). Unfortunately, such transducers yield a low and sometimes non-linear output, require extensive supporting circuitry, drift with temperature, and have to be frequently recalibrated. In response to this problem, I have designed a simple, inexpensive, general-purpose pressure trans-ducer that is based on a commercially available, solid-state sensor. The sensor response is linear over the differential pressure range ±1243 Pa and consists of a pressure-sensitive integrated circuit in the form of a small diaphragm. Two gas ports connect to opposite sides of a diaphragm, permitting differential measurements. The diaphragm is housed in a compact plastic package, complete with electronics that provide a temperature-compensated, calibrated output of 497·3 Pa V-1 (typical accuracy ±1%), ranging from 1·0V (—1243 Pa) through 3·5 V (0Pa) to 6·0 V (+1243 Pa). The complete sensor is available from Omega Engineering Inc., PO Box 4047, Stamford, CT 06907-0047, USA (part number PX164005 BD 5V). Part number PX164010D5V can be substituted for measurement of differential pressures from 0 to 2486Pa. The addition of a simple circuit (Fig. 1), based on inexpensive and readily available components, allows the user to specify gain (1-50X) and offset, over the entire pressure range of the sensor. A description of the circuit follows.

A voltage regulator reduces the positive power supply voltage (12—18 V) to a well-regulated 8-00 V required by the sensor electronics. The output of the pressure sensor is buffered by Ul, and fed into an RC integrator that has a time constant of about 60s (see below, this paragraph), the output of which is buffered by U2. A further buffer (U3) provides a front-panel source of the unprocessed sensor output via connector JI. The output of Ul is connected to the non-inverting input of a variable-gain differential amplifier built around U4. Gain is controlled by the double-ganged wafer switch SI over the range 1-50X. The inverting input of the differential amplifier is switch-selectable (S2), and allows the user to select either a continuously variable offset (10-turn potentiometer VR2 and buffer U5) or a ‘floating’ offset derived from the RC integrator, that is close to the mean of the applied pressure over the last 60s. The latter setting is useful for measuring small, rapid fluctuations (e.g. avian respiration) superimposed on a large, slowly drifting pressure offset. The output of U4 is available on the front panel at J2. Because operation in floating offset mode produces an output from U4 that is symmetrical about 0 V, a summing circuit (U6—U8) adds a voltage adjustable by VR3 to the output of U4. This is necessary to prevent the input of negative voltages to the single polarity A/D converters often used in computerized data acquisition. The summed voltage is available on the front panel at J3. Circuit layout is not critical, but good wiring practice should be followed, such as the use of a single-point ground node.

The circuit is simple to set up. The only instrument required is an accurate digital voltmeter. After applying the supply voltage, the voltage at point A is adjusted by VR1 to 8·00 ± 0·01 V. The voltage at JI should now be 3·50 ± 0·05 V if no gas connections have been made to the sensor (1·00 ± 0·05 V in the case of sensor type PX164010 D 5V). Set switch SI to 1X gain, and switch S2 to fixed offset. The voltage at J2 should be adjustable to zero by the offset control VR2. The sensitivity of adjustment should increase markedly with increasing gain (the gains can be checked, if desired, by dividing the output voltage at J2 by the difference between the voltages at the inverting and non-inverting inputs of the differential amplifier). Set switch S2 to floating offset; the voltage at J2 should fluctuate about 0 V with pressure changes. Measure the voltage at J3, and with the gain on lx, adjust VR3 to give a suitable reading corresponding to no pressure differential (e.g. 1·250 V for an A/D converter or chart recorder with an input range of 0-2·500 V). Adjustment is now complete.

In critical applications the pressure sensor should be calibrated against a good manometer, and the slope of sensor output voltage plotted against differential pressure should be recorded (Fig. 2). Because the transducer is primarily employed to measure differences in pressure, offset pressures are first nulled by VR2, and the output voltage of the transducer corresponding to pressure changes is then divided by the gain set by SI and multiplied by the pressure/voltage slope to yield pressure changes in Pa. If calibrated in this way, and if precision resistors are used for gain determination, the absolute accuracy of the transducer as a whole should be better than 1 % over the entire pressure range. Resolution should in any event be better than 0·1 Pa.

Several transducers of this type have been made, and they have proved to be reliable, sensitive and accurate. Figs 3 and 4 show typical results in two very different applications. Both were plotted directly from data captured by a computerized data acquisition and analysis system (Lighton, 1985).

I thank Dr G. A. Bartholomew for financial assistance from US National Science Foundation grant BSR8400387, and Dr T. L. Bucher for providing the initial incentive for the design of the transducer.

Bucher
,
T. L.
(
1981
).
Oxygen consumption, ventilation and respirator) heat loss in a parrot, Bolborhynchus lineola, in relation to ambient temperature
.
J. comp. Physiol
.
142
,
479
488
.
Lighton
,
J. R. B.
(
1985
).
Minimum cost of transport and ventilatory patterns in three African beetles
.
Physiol. Zool
.
58
,
390
399
.
Malan
,
A.
(
1973
).
Ventilation measured by body plethysmography in hibernating mammals and in poikilotherms
.
Respir. Physiol
.
17
,
32
44
.