ABSTRACT We have designed and built a piezoelectric sensor, from the commercially available Kynar piezo film (Pennwalt Corporation), which can be used for various pressure
measurement applications. The sensor is 42 x 19 x 2 mm,70710
with the piezo film thickness alone being 52 qm. We tested the sensor and found good linearity. We plan to use thcsc sensors to carry-out pressure measurements under the bony 9foiJlifiences of tbe foot. We discuss the circuitry and
problems involved in such measurements, and suggest
solutions to these problems.
INTRODUCTION
Several investigators have used piezoelectric sensors.
Hennig et at. [13 developed a shoe insole (for the measurement
Of foot contact stress), consisting of an array of 499 lead
il' Oflate titanate sensors, each being 4.78 x 4.78 ›‹ 1.2 mm. These were then embedded in a 3--4 mm layer of highly resilient silicone rubber. All the Sensors were laid out on a thin copper gauze sheec Thin copper wires wer• attached to the upper electnxle, and thesc ran to a small backpack, containing
Karr ct at. [2] deVélopcd a sensor system for the measurement of arterial pressure waveforms. Starita «t at. Ms j developed a matrix of sensing sites consisting of 128 10-mm diameter circuldf sensors for measuring pressure betwccn the bare foot and the floor.
SEATSOR CONSTRUCTION
Figure 1 shows the cross section of the sensor. The piezo ffmis 52)imUJck.Rezo film has a very high impedance, ana therefore we use 5hieldcd Wire to minimize 60-Hz interference from different SOurGes. A shielded wire was soldered to
then the clip was firmly attached to the piezo film. This
avoided melting the film. We glued insulating Mylar films (42
side of the piezo Um with epoxy. 1“1us was then giued with epoxy and taped to metal (42 ›‹ 17
1. 2 lllfD), which provided solid backing to the otherwise soft
insole.
CIRCUIT DESCRIPTION
This sensor feeds the charge amplifier circuit showfl in
Fig. 2. A charge amplifier was chosen over a voltage amplifier
shielded wire reads to the inverting input of a MOS/FET operational amplifier. The biggest problem with any piezoelectric sensor iS that it does not respond to steady pressures. Small amplifier bias currents eventually causc saturation. A high RSistanCe, f (- 112 Mfg) was initially put across the feedback capacitor, Co (- 100 nF), in order to prevent saturation. This gave a time constant of 11.2 s, long enough to record the walking data.
The first stage of the op amp gave an output voltage in the range of 80—100 mV, when pressed with the finger. The gain of this stage could have been increased by decreasing the value of the feedback capacitor, C’y, which implied a much higher
resistor value of f I hiG11 WtlS itT1 JBctically high), in order to fflttifltlllfl II tilTlC constant ( Nf) Of about 10 s. Therefore, a second noninverting stage (with a gain of 40)was added to amplify this signal. Then /iIS a baseline drift, due to the pyroelectric effect and the bias current (= 5 pA) of the op amp, which eventually drifted towards saturation. The esistance, I , was added at the output to decouple the capacitance of the load.
We propose an automatic switch (CMOS analog switch), which we simulated with a mechanical switch. z. across the capacitor, Cy. With this switch open, the output drifted
towards saturation in about 10 h. To achieve an error Of 19r , we plan to reset the switch every 6 min using an interrupt from the microprocessor. We would have to ensure that the resetting occurred when the pressure was zero, such as when the foot is in the air.
With a typical bias current of 5 pA, the output would drift
supply an external current (by closing the switch S , in Fig.2) equal to the bias current of the op amp, then the output saturation time would be greatly increased. Initially, we used a capacitance (Cf) Of llXl pF, in order to observe the faster variation of thc output towards saturation (in less than 1 min). We then adjusted the external current using the potentiometer, 压力测量的压电传感器英文文献和中文翻译:http://www.youerw.com/fanyi/lunwen_80154.html