DE2721585B2 - Digitally displaying precision thermometer - Google Patents

Description

The invention relates to an electrical precision thermostat, mometer with a non-linear resistance sensor and an analog / digital converter. Such thermometers are usually constructed in such a way that the non-linearity of the sensor is initially eliminated by adding suitably dimensioned passive components is compensated. In particular, the circuit in shape a Wheatstone bridge has long been known (1) and is often used. After the sensor current has been linearized by the circuit measures mentioned from the field of analog technology, is it is fed to a commercially available analog / digital converter. However, the conventional arrangement has significant disadvantages.

The interconnection of the sensor with correction resistors must result in a considerable reduction the temperature coefficient, d. H. the sensitivity of the measuring arrangement must be accepted. In addition, circuit measures from the field of analog technology are used in precision measuring devices basically known to be particularly susceptible to failure. The rule applies that with increasing circuit complexity im Analog part the problems with temperature drift and component aging increase.

The task is to use a non-linear resistance sensor with a precision thermometer to construct digital signal representation. The linearization of the sensor characteristic with such Means are carried out that a reduction in the measurement sensitivity is avoided and that a Optimum freedom from drift, resistance to aging and low susceptibility to failure is achieved.

The object is achieved by the measures a) to d) characterized in claim 1.

The following is a summary of the advantages of the linearization method used in the invention will:

1. Circuit simplification compared to conventional methods. Where previously a linearizing element, e.g. B. a bridge circuit, had to be used together with a strictly linear analog / digital converter this task can now be taken over by the converter alone. In particular, the linearizing converters can be implemented with fewer components than a conventional one strictly linear converter with the same accuracy requirements. The reduction of the Component costs should not only be seen from the economic side, but also from the Construction of a device that z. B. the calibration requirement is subject to, the advantages in terms of temperature drift, component aging and Reliability matters

2. With the linearization method, the current of the resistance sensor goes into the full amount Analog / digital converter on.

3. When using the linearization method, both connection lines of the resistance sensor are at a constant electrical potential. This makes it particularly easy to construct sensors that are protected against by a protective conductor Leakage currents are secured.

4. According to the invention, an electrical thermometer can be constructed without a stabilized one Reference voltage is required. The stability of the arrangement depends primarily on the Constancy of the capacitance of a mica capacitor and the constancy of a time base.

The invention is illustrated by means of FIGS. 1 to 9 explained. F i g. 1 shows the block diagram of a complete one Precision thermometer. It is a sample device with the following properties:

Measuring range:

35.00-39.99 degrees Celsius
Implementation time:

0.32 s per measured value, integrating conversion
Temperature drift:

Drift with changed ambient temperature 0 ... 50 ⁰ C max. ± 0.01 ⁰ C from the value at 2O ⁰ C

Switch-on drift:

Difference between the measured values 3 s after switching on the device and 3 hours after switching on the device max. + 0.01 ^° C

Linearity error:

Linearity error when calibrating the scale at the start and end point max. ± 0.003 ° C

Mains voltage dependency:

Drift when the line voltage changes in the range -10% ... + 20% max. ± 0.003 ° C

Digitization uncertainty:

not applicable because the converter is synchronized with the measuring cycle

(I) WR Beakley: The design of thermistor thermometers with linear calibration. Journal of Scientific Instruments, Vol. 28, ] une \ 95 \ .OA76— \ 7S.

In Fig. 1 the resistance sensor is provided with the reference number 1, the analog part with the linearizing sawtooth converter with the reference

chen 2. The power supply unit has the reference number 3, the Control part the reference number 4, the decimal counter with fade-out the reference number 5, the other decimal counters the reference numerals 6, 7 and 3. The three The control lines shown have the following function:

ÜB = clock signal, which brings the current counter reading into the buffer memory and thus enables the counter reading to be displayed. In addition, the transmission signal is used for clock synchronization of the sawtooth converter

R0 = clear signal which, after the counter reading has been saved, sets the counter to zero again for the next measuring cycle.

BL = signal generated by the fade-out, which prevents temperature values from being displayed that do not belong to the linearized working range.

In Fig.2 characteristic diagrams are shown that can be explained as follows:

Fig.2a shows how a sawtooth converter with relative large discharge time maps the input current / to the pulse repetition frequency / generated by it

F i g. 2b shows how a non-linear resistance sensor, in particular an NTC sensor , maps the temperature # onto the measuring current I _n

F i g. 2c arises from FIG. 2b, if the compensation current / * is subtracted from the measuring current I _n. This subtraction is necessary if the working characteristic of the sawtooth converter (see Fig. 2d) is to go through the origin of the coordinates.

2d shows how a sawtooth converter with a relatively long discharge time, referred to below as a defined discharge time, maps the temperature Ό · onto the pulse repetition frequency ff &). The interrupted straight line is the working characteristic of the converter required for temperature measurement. This theoretical characteristic curve runs through the coordinate origin and through points A, B and C. At points A, B and C, the deviation of the actual characteristic curve (continuous line) from the theoretically required characteristic (broken line) is exactly zero. In the Aß area and in the ÖC area, this deviation is very small. The true magnitude of the deviation for the temperature range selected in the sample device is shown in FIG. 5 shown enlarged

In Fig. 3 is shown the circuit diagram of the analog part. It is a current / frequency converter in the form of a sawtooth converter. The ratio of charging and discharging time of the converter is dimensioned in such a way that the current of a non-linear resistance sensor (RO) is converted into the pulse repetition frequency f (&), which is strictly proportional to the temperature. As already explained in the description of FIG. 2, when a suitable compensation current h is subtracted from the measurement current I _n , the curvatures of the characteristics of the sensor and transducer are almost completely canceled out within the predetermined measurement range (area AC in FIG. 2d). . In the following, the signal flow is explained in more detail using the circuit diagram according to Figure 3:

The resistance sensor RO is located between the supply voltage (U ₊ ) and the inverting input of the current mirror (/ Cl). The voltage drop that occurs across it is therefore consistent under all operating conditions. The current I _n flowing through the resistance sensor (RO) is inverted by the current mirror (/ Cl) and passed on as ~ l _m to the integrator (IC2) . The inversion gives the possibility of deriving the measuring current I _m, the compensation current Ik, the switching _{current / s} and the reference voltage of the integrator formed by a voltage divider (R 3) from the same supply voltage (U ₊ ) . As a result, fluctuations in the supply voltage primarily have no influence on the pulse repetition frequency f (&) generated by the converter.

With the help of the capacitors C 4 and CS , the currents (-Im, Ih h) present at the input of the integrator (IC2 ) are integrated as soon as the upper limit switch (/ C3) detects that the

divider R 3 defined upper switching point is reached (see. Oscillograms according to Fig. 4a and 4b), then the RS flip-flop / C5 switches on the switching current Λ via the transistor T2 for discharge.

If the lower limit switch (/ C4)

:> o Reaching the lower switching point reported (see oscillograms according to Fig. 4a and 4c), the discharge is ended and the RS flip-flop is reset (see oscillogram according to Fig. 4d) and charging begins all over again. The oscillogram according to FIG. 4e shows how the transistor T2 is driven.

Switching times, residual currents and insufficient operating voltage suppression of the operational amplifiers / Cl to / C4 cannot be completely neglected. Fluctuations in the positive supply voltage have a slight effect on the pulse repetition frequency fffr) . In order to eliminate this influence, correction pulses are fed back to the integrator via the trimmer PS and the AC element R8 / C7 (cf. oscillogram according to FIG. 4f).

jj Furthermore, with the voltage divider from the resistors /? 10, Λ 11 and P 4 an adjustment option provided, which allows the measurement result to be completely protected from fluctuations in the negative supply voltage to make independent.

Finally, we would like to point out some special features in the practical design of the sample device: The interconnection of the capacitors C4 and C5 as well as the resistors P2, A4 and Λ 5 serves to achieve the lowest possible temperature drift.

The capacitors Cl, C2, C6, C9 and ClO suppress high-frequency interference pulses. The capacitors C3 and C8 as well as the resistor R 16 were introduced to shorten the switch-on and switch-off delay. A gate of / C5 is connected to the RS flip-flop in the sense of an OR link. This makes it possible to completely discharge the integration capacitor each time with the transmission pulse and thus to bring about a clock synchronization of the sawtooth converter. In order to prevent the integration capacitor from being inversely charged, the diode D 1 is provided. Another ICS gate serves as an output buffer.

The square flags in Fig. 3 are the measuring points for the oscillograms according to FIG. 4th

bo As already mentioned, the sawtooth converter gets its defined curved characteristic curve from the fact that the charging time and the discharging time of the integration capacitor (Ci + CS) are related to each other. Therefore, for a given measuring task

b5 those currents that influence charge and discharge, i.e. the compensation current / * and the switching current U as well as the integration capacitance (C4 + CS = C) determined by a system of equations. Successor

The calculation bases and the associated list of symbols are listed below:

Calculation basis:

t. =

_C_L ^

^h I _n , - Ii Ci U,

Is - Um -Um - Ik) -

Um - Ik?

C U ₁

(4I _b -31 _c ) l ² _a - {4I _a -2I _c ) ll (3I _a -2I _h ) I ² _c = 0

2 (/. - h? - Uc - hf

Is =

21. -Ic-h . , Uc- Ik) ²

I _n , I _k

_{N Vi}

Symool directory:

Pulse repetition frequency of the sawtooth converter Charging time of the integration capacitor (C4 + C5)

Discharge time of the integration capacitor (C4 + C5)

Measurement current of the resistance sensor at the voltage U ₊

Current through the compensation resistor (P2 + R5 + R4) at the voltage U ₊ (compensation current)

I _s = current through the switching resistor (R 6) at the voltage U ₊ (switching current)

/ a = measuring current of the resistance sensor (RO) at the lower end of the working range (35.00 ° in the sample device)

h = measuring current of the resistance sensor (Λ0) in the middle of the working area (37.50 ° in the sample device)

Ic = measuring current of the resistance sensor (RO) at the ίο upper end of the working range (40.00 ° im

Sample device)

Q = capacitance of the integration capacitor (C4 + C5)

L ^ / ~ integration voltage yGuercr minus lower Limit value according to FIG. 5a)

T = duration of a measuring cycle (cycle time)

N = number of pulses at the upper end of the measuring range

U ₊ = positive supply voltage

The error curve of the linearization in the temperature range of the sample device is shown in FIG. if the sawtooth converter is dimensioned according to the above equations.

In Fig. 6 is the circuit diagram of a normal decimal counter shown. The decimal counters of the sample device are standard circuits in TTL technology.

The decimal counter with fade out according to Fig. 7 generates the number "3" as the starting value of the most significant decimal and suppresses counter values that are not in the range from 3500 to 3999 with the help of the BL control line. It should also be mentioned that the counter readings 4500 to 4999 could also be displayed on the sample device in the temperature range of about 45-50 ° C. In order to prevent this, the current limitation from the transistor T1 and the variable resistor Pi in the analog part according to FIG. 3 is set in such a way that such current values cannot be reached.

Task of the in F i g. 8 is the control part shown, from the network frequency at certain time intervals

a pulse (ÜB) to transfer the counter readings to the buffer and shortly thereafter a pulse (Ä0) to reset the counter. This is the cycle for a measuring cycle.

The in F i g. 9 power supply unit shown schematically provides the two supply voltages and a 50Hz AC voltage as the time base for the control section.

For this purpose 10 sheets of drawings

Claims (2)

Patent claims: 1. Electric precision thermometer with non-linear working resistance sensor and an analog / digital converter, characterized in that that a) the sensor (R 0) has a constant voltage drop between the supply voltage (U +) and ground potential, b) a current mirror (/ Cl, Ä2, A3) for reversal the sensor current is provided, c) the analog / digital converter (IC2-IC5) has a non-linear characteristic, which straightens the non-linearity of the sensor characteristic, d) the analog / digital converter is a sawtooth converter, the loading time of which is not equal to the Discharge time is and that the ratio of charge time and discharge time the non-linearity of the Converter determined 2. Thermometer according to claim 1, characterized in that the sensor current (I _n ), a compensation current (4), a switching current (I ₅ ) determining the discharge time and a reference voltage for the integrator (IC2) formed by a voltage divider (A3) are derived from the same supply voltage ( U +) .