APPARATUS FOR THE MEASUREMENT OF SIGNALS FROM TENSORESISTIVE PRESSURE TRANSDUCERS A. I. Beklemishchev, Yu. K. Blokin-Mechtalin, V. M. Vlasenko, and V. A. Alekseenko
UDC 621.317.7:531..781.2.087.92
The study of pressure distribution using models is one of the basic methods of testing in experimental aeromechanics. A distinctive feature of this method involves, in the first place, the necessity of measuring the change in pressure over a wide range in a single experiment and, in the second place, a significant number of transducers for primary measurement. For such tests tensoresistive pressure transducers have been used successfully; these devices exhibit excellent metrological characteristics, wide measurement ranges, quick operation, small size, etc. The measurement apparatus used to record these signals should guarantee the required precision over the entire measurement range as well as high-speed operation for the purpose of conserving experimental time. In addition, the volume of data obtained experimentally is large, therefore, automation of data collection and processing is necessary. Since the number of tensoresitive transducers in any given test is relatively large (up to several hundred), it is reasonable to adopt an apparatus that has an input commutator. Subsequent conversion of signals into code is achieved by a measurement device that can be built using various schemes. Below, we examine an apparatus in which measurement is performed with the help of a low-level automatic digital processor (ADP) without preliminary signal amplification, which allows solution of the formulated problem. In order to work under various conditions and with various pressure transducers, two ADP configurations have been designed. An ADP that employs series connection of the pressure-transducer tensoresistive bridge (tensobridge, TB) and the code--voltage transducer (CVT) to the power supply (see Fig. i) [I] is used for the measurement of signals from the tensoresistive pressure transducers. In order to increase the precision and sensitivity of the apparatus, power to the tensobridge and CVT is supplied from a single bipolar pulsed voltage source U; the supply voltage had a square-wave shape with an amplitude of 40 V and a frequency of i0 kHz. The supply voltage is applied across the pressure transducer fo~ 2 msec, which is the time necessary to convert the tensobridge signal into code. A voltage drop across a resistance R, which is connected in series with the supply circuit of the tensobridge is used to operate the CVT. The output voltage of the tensobridge Uou t and the compensating voltage U c across the CVT are compared in the null sensor (NS), which sends command pulses to the code-equivalent shaper (CES). For the given circuit rlr 4
Uou t
=
--
ttl" s
- -
rz+rf~_ra+r 4 IR Uc= N'-KI
V+ ADP T=~ II a /I
~
z -
9 i;
'
~
(2)
~>~--~
ICglnrt nr~
r
IY
N
I
I
j K~r ---2 Fig. 1 Translated from Izmeritel'naya Tekhnika, No. 3, pp. 34-36, March, 1980.
230
0543-1972/80/2306-0230507.50
(i)
9 1980 Plenum Publishing Corporation
rcf ADP
I v - 1 roy
,
rCZ~!-" Fig. 2 where rl to r4 are the resistances of the tensobridge arms; I is the supply current to the tensobridge; N is the code equivalent of the measured quantity; K I is the proportionality coefficient. If Uou t = U c and r,=i&"=r+Ar;
(3)
r==ra= 9
where r is the nominal resistance of the tensoresistors in the bridges, and Ar is the absolute increase in resistance, then ~r
N = KI-R
(4)
From (4) it follows that the code equivalent N is proportional to Ar. Measurement of absolute Ar and not the relative increase Ar/r is reasonable in the application of semiconductor tensoresistors since in this case, errors caused by the change of tensoresistor resistance under the influence of temperature play a lesser role [2]. For the given circuit the relative value of this error 8~ =
ArT e+r
(5)
where Ar T is the temperature-related increase of the tensoresistor resistance. From (5) it follows that for a decreased 6T it is reasonable to approximate the supply to the tensobridge by a current source regime (R-~o), which is also preferable for semiconductor pressure transducers. Since the output signal of the bridge (i) and the output of the compensator (2) are proportional to the general supply current I, and, since the input resistance of the null sensor is high, the variations in the resistances in the connections between the pressure transducers and the apparatus and the commutator elements in them practically do not introduce error into the measurement. Therefore, in the given circuit we can use a contactless code-equivalent switch tensobridge commutator. In the presence of dynamic components in the signals from the pressure transducers at the output of the tensobridge, filters ~i-~ i are included. In this case, the group of tensobridges is connected to the general constant-voltage source U, and supply to the CVT is taken off of the corresponding diagonals of the tensobridges, i.e., this is done by connecting the CVT and the tensobridges in parallel (Fig. 2). Supply to the CVT along the lines of the circuit in Fig. 1 is not appropriate in this case because of the instability of R, owing to the flow of a significant resultant current through this resistor to the tensobridges. Without filters (in the absence of dynamic components in the signal) the circuit can be used with a pulsed supply as well, with commutation of supply voltage to each bridge and to the input of the CVT. For the compensator in the circuit shown in Fig. 2, for conditions account of the connection resistance,
(3) and not taking
N = Kuer,
where K U is the proportionality constant; cr = Ar/r.
231
The code equivalent N depends on the relative increase of tensoresistance, which is proportional to the pressure and does not depend on the nominal resistance of the tensoresistors. Therefore, we can measure the tensobridge signals that occur across the tensoresistors of various ratings without any loss in sensitivity or changes in the circuit of the compensator. The error due to the influence of resistances rc~-rc6 in the lines connecting the tensobridge and the compensator is, for equal values of these resistances, 5e = 2 ( r e •
)/Rim
where r c is the resultant resistance of the connection; A ~ is the variation in the resistance; Rin is the input resistance of the CVT, which for lowering the error ~c is chosen to be sufficiently high (tens of kilohms). The temperature error is 6T = ArT/r" The measurement apparatus that uses these digital compensating circuits (Fig. 3) a codeequivalent (CE), a commutator-control unit (CCU), a command-shaping unit (CSU), an indication unit (IU), a computer-interface unit (CIU), a puncher PL-150, a "Konsul" typewriter, a device for the metrological study and control of the apparatus: an automatic electrical-signal calibrator (AESC) [3], and an automatic pressure controller (APC). In order to increase the precision of the measurements over a wide range of input signals to the ADP, two measurement ranges with different conversion coefficients (quantization steps by level) have been provided. The ADP is constructed using the method of step-by-step balancing with automatic determination of the sign of the measured quantity and choice of measurement range [5]. In the ADP shown in Fig. i, a transformer voltage divider, comprising four inductive tetrads that effect binary-decimal encoding, is used as a CVT. In the compensator shown in Fig. 2 we can use a CVT based on transformers or on resistors for constant supply or pulsed voltage. For constant voltage at the input to the divider and the output of the commutator, full-wave modulators are used, which transform the constant voltage supplied to the tensobridge (U = 5 V) and the output voltage from the tensobridges (Uout = • mV) into a square-wave voltage of frequency i0 kHz. The CCU permits sequential interrogation of any number of pressure transducers from 1 to 256 (starting with any transducer), repetition of from 1 to 15 interrogation cycles, multiple (up to 20 times) measurement of signals from a transducer in the course of one interrogation cycle, connection of any transducer of particular import to the compensator, and computer control of the commutator. The CIU accomplishes exchange of data between the computer and the ADP in relation to a standardized system of connections and signals. The CIU fulfills the following basic functions: It transforms the binary-decimal code of a measurement having weighing coefficients 1-2-4-2 to a code having coefficients 1-2-4-8, it effects the sequential transfer of information to the computer by bytes, and it forms the evenness criterion for information bytes in order to monitor communication with the computer. [ " M e a s u r e m e n t apparatus
t
'
1 -..%1
I
"r'
r '
Controllir~ automatiol of the d e -
~ ~
i i
7 l_---------
Z--_~
~leet.rologieal m o n i t o r i n g vices
Fig. 3
232
I
J !
-
Zomputer
Synchronization of all units of the apparatus is performed by the command-shaping unit (CSU). At a signal from the controlling automation or the computer, the CSU sequentially works out commands for the apparatus to prepare for measurement, for the connection of the next transducer, and for the application of voltage to the tens.bridge, and it then issues the command "encode." Upon completion of an interrogation cycle, the CSU issues response commands for the experimental device and the computer. When working with devices for metrological monitoring (AESC, APC), the CSU accepts their commands and issues responses. The AESC is used for the study and monitoring of the ADP characteristics with respect to the input commutator, while the APC is used for the calibration of pressure transducers. An automated metrological complex using the AESC and APC as a basis results in increased reliability in the metrological characteristics of the apparatus and an increase in the productivity of metrological efforts [6]. The basic technological characteristics of the apparatus are: 256 measurement channels; 200-800-~ resistance in the arms of the pressure-transducer tens.bridges; • range for measureable signals from the tens.bridges; • equivalent capacitance for the apparatus scale (taking account of both measurement ranges); automatic determination of sign and measurement range; 200-transducer/sec rate of interrogation by the contact commutator; 8-pV maximum sensitivity; up to 100-m long measurement path between the pressure transducer and the ADP; an error of y=
•
U~Uou
t
1)]% J '
for the first and second ranges, where Uout,ma x and Uou t are the limiting and instantaneous values of the measured signals, respectively, on the chosen range. The apparatus is built on the basis of an ASETo LITERATURE CITED l. 2. 3. 4. 5. .
A. I. Beklemishchev, et al., Inventor's Certificate 405075, Byull. Izobr., No. 44 (1973). M. Din (editor), Semiconductor Tensile Pickups [in Russian], Energiya, Moscow--Leningrad (1965). Yu. K. Blokin-Mechtalin, Izmer. Tekh., No. 4 (1978). B. S. Dub.v, et al., Izmer. Tekh. No. 4 (1978). Yu. K. Blokin-Mechtalin, et al., Inventor's Certificate 309308, Byull. Izobr., No. 22 (1971). A. I. Beklemishchev, et al., Izmer. Tekho, No. 1 (1977).
233