JPH0120368B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a multi-point temperature measurement method, and more particularly to a method for measuring temperatures at a large number of measurement points using a three-wire resistance temperature detector.

As a so-called multi-point temperature measurement method in which temperatures at a large number of measurement points are selected and measured in a time-sharing manner, the one shown in FIG. 1 is conventionally known.

In Fig. 1, 2 ₁ , . . . 2n are resistance temperature sensors provided one each for n measurement points, and 3 ₁ , . . . 3n and 4 ₁ , . Power supply terminal of body 2 ₁ ,...2n, 5 ₁ ,...5n
is the detection terminal. That is, the temperature measuring resistor 2
₁ , . . . 2n are generally called three-wire resistance temperature sensors. Then, each resistance temperature sensor 2 ₁ ,...
... 2n are connected to a constant voltage power supply +B via conductor wires _{6 1 ,} ... 6n and side resistors 7 ₁ , ... 7n, and the other power feeding terminals 4 ₁ , . ... 4n is grounded via the conductor wires 8 ₁ , ... 8n and the side resistances 9 ₁ , ... 9n, and each temperature sensing resistor 2 ₁ , ... 2n is connected to the side resistances 9 ₁ , ... 9n, Together with other side resistances 7 ₁ , . . . 7n and side resistances 1, 1' common to each measuring point, they form a bridge. Moreover, each of the above-mentioned detection terminals 5 ₁ , ... 5n is connected to a conductor 10
₁ , . . . 10n, and are connected to a bus 11 common to each measuring point via multiplexer switches 15 ₁ , . Furthermore, the bridge is connected to a data processing device 14 via an amplifier 12 and an analog/digital converter 13.

Now, to explain the first measurement point,
The unbalanced voltage of the bridge, that is, the measurement voltage signal, which is generated when the temperature at the measurement point changes and the resistance value of the resistance temperature detector 2 ₁ changes, is taken out via the detection terminal 5 ₁ and connected to the conductor 10 ₁ , The signal is input to the amplifier 12 via the multiplexer switch 15 ₁ and the bus 11, and is amplified. The analog/digital converter 13 converts this amplified measurement signal into a digital signal and sends it to the data processing device 14. The data processing device 14 has a function of instructing, recording, and controlling the temperature, and the data processing device 14 allows the temperature at the first measurement point to be known. Measurements are made in exactly the same way for other measurement points, but selection is made by switching the multiplexer switches 15 ₁ , . . . , 15n.

In this method, the resistances of the conductive wires 6 ₁ , . . . 6n and 8 ₁ , . . . 8n affect measurement accuracy. For this reason, we have taken the following structural measures.

That is, for example, regarding the first measurement point, the resistance values of the side resistances 7 ₁ , 1 and 1' are both r ₁ , the resistance value of the side resistance 9 ₁ is r ₂ , and the resistance temperature detector 2 ₁ at the center temperature of the measurement range is Assuming that the resistance value of conductor 7 ₁ is the same as that of conductor 9 ₁ ,
The resistance values r ₁ and r ₂ of the side resistances 7 ₁ and 9 ₁ are selected so that rtm=r ₁ −r ₂ . Then, if we do not take into account the effect of current changes due to temperature changes and changes in the resistance value of the resistance thermometer ₂₁ , the conductor resistance will be compensated for in a very narrow range centered around the above temperature, and the accuracy will be quite high. measurement becomes possible.

However, this requires two expensive precision side resistors for each measurement point, increasing costs. The cost increase due to this precision side resistance is due to the fact that this method requires an extremely large number of measurement points.
Sometimes there are thousands of points, so it cannot be ignored. Another drawback is that each time the measurement temperature range changes, it is necessary to replace the side resistance with one of an appropriate resistance value and readjust it so that the above-mentioned condition rtm = r ₁ - r ₂ is obtained.

The object of the present invention is to overcome the above-mentioned drawbacks of the conventional methods and to reduce the expensive precision side resistance to only one for each measuring point.
It is an object of the present invention to provide a multi-point temperature measurement method that not only can reduce the number of resistors to 100, but also can reduce costs and eliminates the need to replace the resistors on each side even if the measurement temperature range changes.

To achieve the above object, the present invention arranges one three-wire resistance temperature detector having a pair of power supply terminals and one detection terminal at each of a plurality of measurement points, and divides the measurement points in a time-division manner. In the method of selectively measuring the temperature at each measurement point, one power supply terminal of the resistance temperature detector is connected to a power source via a first conductor and a resistor, and the other power supply terminal is connected to a second conductor. grounded through the conductor, converts the voltage drop of the temperature measuring resistor including the voltage drop due to the first and second conductive wires into a digital voltage signal, and converts the voltage signal into a digital resistance signal,
It is characterized by a multi-point temperature measurement method in which the first and second conductor resistances included in the resistance signal are compensated based on a conductor resistance table prepared in advance and then converted into a digital temperature signal. .

To explain the method of the present invention in detail, in FIG. ₂ , 2 1 , . These resistance temperature detectors 2 ₁ , . . . 2n are connected to power supply terminals 3 ₁ , . . . 3n and 4 ₁ ,
. . . 4n and detection terminals 5 ₁ , . . . 5n.
It is a wire type resistance thermometer. One of the power supply terminals 3 ₁ , . . . 3n is connected to a constant voltage power supply +B via conducting wires 6 ₁ , . . . 6n and side resistors 7 ₁ , . Further, the connection points of the conductive wires 6 ₁ , . . . 6n and the side resistors 7 ₁ , . . . 7n are connected to multiplexer switches 16 ₁ , .
Each measuring point is connected to a common bus bar 11 via a bus line 11. On the other hand, the other power supply terminals 4 ₁ ,...4n are
They are grounded only through the conductive wires 8 ₁ , . . . , 8n without using side resistances. That is, from constant voltage power supply +B, side resistances 7 ₁ , . . . , 7n, conducting wires 6 ₁ , .
6n, resistance temperature detector 2 ₁ ,...2n, conductor 8 ₁ ,...
A resistance measuring circuit is configured that reaches a ground point (common potential point) through 8n. In addition, the detection terminal 5
₁ ,...5n are connected to conductors ₁₀₁ ,...10n and multiplexer switches 151, _15n selected in a time-sharing manner.
. . . is connected to the bus bar 11 via 15n. The bus bar 11 is connected to an amplifier 12, and a connection point between the side resistors 1 and 1' is connected to the amplifier 12. Furthermore, a data processing device 14 is connected to the output side of the amplifier 12 via an analog/digital converter 13 and a data converter 17. The data converter 17 includes a voltage-resistance conversion means 18, a conductor resistance correction means 19,
Resistance-temperature conversion means 20 and conductor resistance determination means 2
1, a conductor resistance table (memory) 22, a voltage-resistance linearization correction table (memory) 23, a resistance-temperature linearization correction table (memory) 24, and a measurement point selection means 25.
have.

In the above, the voltage-resistance linearization correction table 23 and the resistance-temperature linearization correction table 24 are created and incorporated when designing the device. However, in the present invention, it is not necessarily necessary to provide the resistance-temperature linearization correction table 24. The reason for providing these tables will be explained below.

Now, in the resistance measuring circuit described above, the voltage of the constant voltage power supply +B is es, the resistance value of the side resistances 7 ₁ , .
₁ ,...2n are rt, and the resistance values of conductors ₆₁ ,...6n and ₈₁ ,...8n are both ra (however,
In each resistance circuit, each conductor 6 ₁ and 8 ₁ , 6n
and 8n have the same resistance values), the side resistances 7 ₁ , . . . 7n and the _conductive wires 6 ₁ ,
The voltage at the connection point of ...6n is ex, the voltage at the detection terminals 5 ₁ , ...5n taken out via the multiplexer switch 15 ₁ , ...15n is ey, the current flowing through the resistance measurement circuit is i, and Assuming that no current flows through , the following equations (1) to (3) hold true.

ex=i(rt+2・ra)...(1) ey=i(rt+ra)...(2) i=es/(rs+rt+2・ra)...(3) Therefore, from equations (1) and (3) above, rt is rt＝[rs・ex/(es−ex)]−2・ra……(4) Also, from equations (2) and (3) above, rt＝[rs・ey/(es−ex)]−ra……(5) holds true, and further, from equations (4) and (5) above, ra＝〔rs・ex／(es−ex)〕 − [rs・ex・ey/ex (es−ex)] …(6) holds true.

In the above, equation (5) is a relational equation that is normally used when measuring the voltage at the detection terminal of a resistance temperature sensor and calculating its resistance value, but in order to calculate rt from this, ex and ey must be must be measured at the same time,
It's inconvenient. Therefore, in the present invention, in order to eliminate this inconvenience, the conductors 6 ₁ , . . . , 6n and 8 ₁ ,
...Measure the voltage ex including a resistance of 8n, and use the voltage-resistance conversion described later to calculate the resistance value rx= to the connection point of the conductor wires 6 ₁ , ... 6n and the side resistances 7 ₁ , ... 7n
After converting to rt+2・ra, rt is measured by subtracting the conductor resistance determined by the above equation (6), that is, 2・ra, from this value. Therefore, at the time of measurement, only ex needs to be measured, and ey only needs to be measured when creating the conductor resistance table 22.

By the way, the above rx is rx=rt+2・ra=rs・ex/(es−ex)...(7) Since es and rs are constants, if we measure ex, we can obtain from equation (7) above. rx is uniquely determined. However, the voltage actually input to the analog/digital converter 13 is a side resistance 1 corresponding to the resistance value rto of the resistance temperature detector at 0°C or the resistance value rtz of the resistance temperature detector at the zero point temperature tz. , 1' (hereinafter referred to as zero shift amount)
Since there is a difference between In this case, since ex is much smaller than es, not only high-precision calculation is required, but also the calculation time becomes longer. Therefore, in the present invention, in addition to performing the above correction, Rx
A voltage-resistance conversion means 18 and a voltage-resistance linearization correction table 23 are provided so as to be able to measure the voltage.

Next, the relationship between the resistance value rt of the resistance temperature detector and the temperature t is as follows: rt ₀ is the resistance value of the resistance temperature detector at 0°C, α and β are the temperature coefficients (α=0.39749×10 ^-2 , β
= −0.58937×10 ^-6 ), then there is a nonlinear relationship of rt=rt ₀ (1+α・t+β・t ₂ ) ...(8), and from this, the formula t[1+(t・β/α)]=( rt−rto)/α・rto
...It is necessary to find the temperature t based on (9). However, this is also very difficult. Therefore, in the present invention, the resistance-temperature conversion means 2
0 and low resistance-temperature linearization correction table 24 is provided.

Next, a method for creating the voltage-resistance linearization correction table 23 and the resistance-temperature linearization correction table 24 will be explained.

First, at the time of design, the type of resistance temperature detector to be used,
In other words, the analog/digital converter 13 is selected depending on the resistance value at 0° C. to be used and the desired measurement accuracy.
Furthermore, voltage es of constant voltage power supply +B, side resistance 7 ₁ ,
...Determine the resistance value rs of 7n and the resistance value of side resistance 1. The resistance value of side resistance 1 is set equal to rs. Also, the resistance value rtz of side resistance 2, the temperature tz at the zero point of the measurement range, and the temperature tfs at the full scale.
decide. In this case, the above tfs includes a temperature corresponding to the permissible value of the conductor resistance. Once these conditions are determined, the constants in equations (4), (7), (8), and (9) above are determined. Then, by adjusting the amplification degree of the amplifier 14, the output voltage signal Ex of the analog/digital converter 13 becomes Ex=Ke(ex−etz)
becomes. Here, Ex is a normalized quantization voltage, Ke is a quantization constant, and etz is the above-mentioned zero shift amount. Therefore, the normalized quantized voltage signal E, resistance signal R and temperature signal T are
E=Ke(et-etz) =K(et-etz)/(etfs-etz)...(10) R=Kr(rt-rtz) =K(rt-rtz)/(rtfs-rtz)... (11) T=Kt(t-tz) =K(t-tz)/(tfs-tz)...(12) Then, the resistance value rt of the resistance temperature detector corresponding to an arbitrary temperature t and its time The voltage et is the above E, R,
The above equations (10) to (12) are defined so that T becomes 0 and the above E, R, and T become K at rfs and etfs corresponding to the full-scale point temperature tfs. In the above, et, etz and etfs are
The equation obtained by setting ra to zero is et=es・rt/(rs+rt)...(13), which is obtained by substituting the resistance value rt of the resistance temperature detector.

Next, as the resistance of the resistance measurement circuit changes, the current i
The error △E due to changes in the measurement temperature range
Find rt from tz to tfz using equation (8), and calculate et obtained by substituting that value into equation (13) above.
Furthermore, by substituting into equations (10) and (11), we get the formula △E=E−R=K{[(et−etz) /(etfs−etz)]−[(rt−rtz) /(rtfs−rtz) ]} ...Calculate according to (14). For example, the correction amount ΔE in which the voltage signal E in equation (10) corresponds to 1° C. in terms of temperature is calculated over the entire measurement temperature range. This calculation result is recorded in the voltage-resistance linearization correction table 23. In this way, the voltage-resistance linearization correction table 23 is created.

Next, the resistance-temperature linearization correction table 24 is created. First, the resistance signal corresponding to the resistance value of the resistance temperature sensor is converted into a linearized temperature signal using the relationship in equation (9). The linearization correction amount △R is calculated using equation (8) from tz to tfs in the measurement temperature range, and using equations (11) and (12), the equation △R = R - T = K { [(rt−rtz) / (rtfs−rtz)]−[(r−tz) /(tfs−tz)]} ...Calculated according to (15). For example, the linearization correction amount ΔR is determined for the resistance signal R according to equation (11) at intervals corresponding to 1° C. in terms of temperature. Then, this calculation result is recorded in the resistance-temperature linearization correction table 24. Furthermore, Etz, Rtz, and Tz corresponding to the zero shift amounts etz, rtz, and tz are also recorded as normalized constants.

Next, a method of creating the conductor resistance table 22 will be explained.

First, multiplexer switch 16 ₁ ,...1
6n in a time-division manner, the voltage ex at the connection point of _the conductor wires 6 ₁ , . . . 6n and the side resistances 7 ₁ , . 5 ₁ ,...5n voltage ey
are guided to the data converter 17 as Ex and Ey, respectively, via the amplifier 12 and the analog/digital converter 13, and are temporarily stored. Regarding the Ex of each measurement point inputted to the data converter 17, the voltage-resistance conversion means 18 uses the input signal as the address signal Ex' to convert the correction amount △ from the voltage-resistance linearization correction table 23 prepared in advance. Read E, calculate Rx = Ex - △E, and obtain a resistance signal Rx corresponding to rx = rt - rtz + 2・ra, including the conductor resistance.
is sent to the conductor resistance determining means 21 along with the voltage signals Ex and Ey.

The conductor resistance determining means 21 detects the input resistance signal.
The conductor resistance is calculated from Rx and the voltage signals Ex and Ey using the relationship in equation (6) above. Here, the input Ex, Ey, and Rx do not include Etz, which corresponds to the zero shift amount etz, and Rtz, which corresponds to rtz, so add Etz and Rtz, and use the following formula to calculate the wire resistance equivalent to 2・ra. Find Ra.

Ra=2 {(Rx+Rtz) − [(Rx+Rtz) ・(Ey+Etz)/(Ex+Etz)]} ...(16) Perform this operation for all measurement points, and calculate the conductor resistance value Ran at each measurement point as the conductor resistance. Table 22
That is why we put it in .

All preparations for temperature measurement were thus completed. Next, temperature measurement will be explained.

The temperature is measured, for example, at a period of 1 second by measuring the voltage drop ex due to the resistance value rt of the resistance temperature sensor including the conductor resistance 2.ra, that is, the conductor wires 6 ₁ , ... 6n and the side resistances 7 ₁ , ... 7n. This is achieved by measuring only the voltage at the connection point.

That is, the voltage ex at _each measurement point is converted to a voltage signal Ex=Ke(ex-
etz) to the data converter 17.

The above Ex is converted by the voltage-resistance conversion means 18,
Using the voltage signal as the address signal Ex', read out the correction amount ΔE from the voltage-resistance linearization correction table 28,
A resistance signal equivalent to rx = rt-rtz + 2・ra including conductor resistance is performed by signal conversion Rx = Ex - △E.
It is led to the conductor resistance correction means 19 as Rx.

The conductor resistance correction means 19 uses the measurement point n as an address signal, reads the conductor resistance value Ran of the measurement point from the conductor resistance table 22, and performs the correction Rtx=Rx-Ran. In other words, it compensates for the conductor resistance. Then, the conductor resistance correction means 19 generates a resistance signal Rtx corresponding to the resistance value rtx=rt−rtz of the resistance temperature detector.
is sent to the resistance-temperature conversion means 20.

The resistance-temperature conversion means 20 uses the input resistance signal as an address signal Rtx', reads out the correction amount ΔR from the resistance-temperature linearization correction table 24, and calculates T=
After performing the correction Rtx-ΔR, the data is led to the data processing device 14.

The data processing device 14 has a function of instructing, recording, and controlling the temperature at the measurement point, and thereby can measure the temperature at the measurement point.

In order to guarantee measurement accuracy over a long period of time, the values in the conductor resistance table 22 are calculated using the method described above at a period that can accommodate changes in the conductor resistance value due to changes in ambient temperature, for example, at a period n times the temperature measurement period. Corrected by.

As explained above, the multi-point temperature measurement method of the present invention connects one power supply terminal of a three-wire resistance thermometer to a power supply via the first conductor and a resistor, and connects the other power supply terminal to a power supply via a first conductor and a resistor. It is grounded through a second conducting wire, converts the voltage drop of the temperature measuring resistor including the voltage drop due to the first and second conducting wires into a digital signal, converts the voltage signal into a digital resistance signal, Since the first and second conductor resistances included in the signal are compensated based on a pre-prepared conductor resistance table and then converted into a digital temperature signal, expensive precision side resistances are not required. Only one needs to be provided for each measurement point, reducing costs. That is, in the conventional method described above, compensation for the conductor resistance is performed in an analog manner using side resistances provided at each measurement point, so two precision side resistances are required.
In the present invention, the voltage drop of the resistance temperature detector, including the voltage drop due to the resistance of the conductor, is measured, and this is converted into a digital signal, and the resistance of the conductor is digitally subtracted from this. It is sufficient to provide only one resistor for the purpose of regulating the current flowing through the resistor. Moreover, since the conductor resistance table is used to compensate for the conductor resistance, calculation processing is not only simple and fast, but even if the measurement temperature range changes, the contents of the table can be updated to ensure that the sides of each measurement point are You can use the resistor as is.

Furthermore, since the method of the present invention first converts a digital voltage signal into a digital resistance signal, the voltage
A resistance linearization correction table can be used to eliminate measurement errors caused by changes in resistance values of the resistance temperature detector and conductor, and changes in the current flowing through them, and a resistance-temperature linearization correction table is also provided. In the case of Accurate measurement becomes possible.

[Brief explanation of drawings]

FIG. 1 is a schematic block diagram showing a conventional method, and FIG. 2 is a schematic block diagram showing an embodiment of the method of the present invention. 1, 1', 7 ₁ , ... 7n, 9 ₁ , ... 9n: side resistance, 2 ₁ , ... 2n: 3-wire resistance temperature detector, 3 ₁ ,
...3n, 4 ₁ , ... 4n: Power supply terminal, 5 ₁ , ...
5n: detection terminal, 6 ₁ ,...6n, 8 ₁ ,...8
n, 10 ₁ ,...10n: Conductor, 11: Bus bar, 1
2: Amplifier, 13: Analog/digital converter,
14: Data processing device, 15 ₁ , ... 15n, 1
6 ₁ ,...16n: Multiplexer switch, 1
7: data converter, 18: voltage-resistance conversion means,
19: Conductor resistance correction means, 20: Resistance-temperature conversion means, 21: Conductor resistance determination means, 22: Conductor resistance table, 23: Voltage-resistance linearization correction table, 24: Resistance-temperature linearization correction table, 25: Measurement point selection means.

Claims (1)

[Claims] 1 3 having a pair of power supply terminals and one detection terminal
In a method in which one wire resistance temperature detector is arranged at each of a plurality of measurement points and the temperature of each measurement point is measured by selecting the measurement points in a time-sharing manner, one power supply terminal of the resistance temperature measurement element is connected to a power supply via a first conducting wire and a resistor, the other power supply terminal is grounded via a second conducting wire, and the resistance temperature detector is Convert the voltage drop into a digital voltage signal, convert the voltage signal into a digital resistance signal, and calculate the resistance of the first and second wires included in the resistance signal based on the wire resistance table prepared in advance. A multi-point temperature measurement method characterized by converting to a digital temperature signal after compensation.