JP2013019738A - Temperature measuring circuit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a temperature measuring circuit which enables accurate detection of temperature by using a semiconductor diode.SOLUTION: A temperature measuring circuit 1, which calculates a temperature T on the basis of a semiconductor diode 5 for temperature measurement, a power supply voltage HVcc, a constant current circuit 7, an AD converter 13 and a digital conversion value, comprises a reference power supply 9 for outputting a reference voltage Vref to the AD converter 13. The temperature measuring circuit 1 is configured to calculate a voltage value [HVcc+ΔHVcc] containing a variation component ΔHVcc, of the power supply voltage HVcc, on the basis of a digital conversion value X(Vref)real of the reference voltage Vref outputted from the AD converter 13, and a digital conversion value X(Vref)ideal of the reference voltage Vref outputted from the AD converter 13 when the power supply voltage HVcc does not contain the variation component ΔHVcc, and to revise a digital conversion value of a both-end voltage VF and a constant current value I of constant current according to the voltage value [HVcc+ΔHVcc] to calculate the temperature T.

Description

  The present invention relates to a technique for measuring temperature using a semiconductor diode for temperature measurement, and more particularly to a technique for measuring temperature with high accuracy.

In a semiconductor device such as an IGBT, it is necessary to accurately detect the temperature in order to protect the semiconductor device from overheating during operation. For temperature detection of semiconductor devices, pn junction type semiconductor diodes and resistors for temperature detection are used. In particular, semiconductor diodes are widely used because they can be manufactured in a chip of a semiconductor device such as an IGBT.
An outline of a temperature measurement circuit using a semiconductor diode is generally as follows. The temperature measurement circuit generally includes a semiconductor diode provided on the same substrate as the semiconductor device, a constant current circuit that applies a constant current in the forward direction of the semiconductor diode, and a constant current circuit. current applying time of the anode - cathode voltage comprises a detection circuit for detecting a (hereinafter, both ends referred to as the voltage V _F), and AD converter for converting a digital signal read an analog output of the detection circuit, and The temperature (absolute temperature) T is calculated based on the relational expression of the temperature dependence shown in the following equation between the both-end voltage VF indicated by the digital signal output of the AD converter and the constant current value I.

Here, k is Boltzmann's constant, q is the electron charge, I is a constant current value, I _s is the diode saturation current density.

  In the conventional temperature measurement circuit, the following four factors can be cited as factors that deteriorate the temperature detection accuracy. That is, characteristic variations (individual differences) of semiconductor diodes, small temperature coefficients (first factor), fluctuations in power supply voltage supplied to the detection circuit and constant current circuit (second factor), constant current of the constant current circuit This is a variation in values (individual difference) (third factor) and an AD conversion error (fourth factor) that occurs during AD conversion.

Variations in characteristics of a semiconductor diode of the first factor is mainly due to the individual difference of the diode saturation current density I _s. Therefore, two constant current circuits having different constant current values I and a switching circuit for selectively switching each constant current to apply to the semiconductor diode are provided in the temperature measurement circuit, and obtained when each constant current is applied. In addition, a technique has been proposed in which the temperature is calculated based on the difference between the voltages at both ends, thereby eliminating the influence of the diode saturation current density Is (see, for example, Patent Document 1).

Specifically, the constant current value of the two constant current circuit and _I 1, _{I 2,} the voltage across at each constant current value _I 1, _{I 2 V F1,} when the _{V F2,} these voltage across _{V F1} , V _F2 is expressed as the following equation.

And when the difference between both-end voltages V _F1 and V _F2 is obtained,



And when it develops at temperature T



Is obtained. In this equation since the term of the diode saturation current density I _s is erased, by suppressing the influence of the first factor, precisely the temperature T is determined.

JP 2004-134472 A

However, since the conventional temperature measurement circuit requires at least two constant current circuits and a switching circuit that switches them and applies them to the semiconductor diode, the number of components, the mounting area, and the cost increase are problems. .
Further, the larger constant current value of the two constant current values I ₁ and I ₂ is affected by the resistance component of the semiconductor diode if the value is large. From the above temperature characteristic equation of the ideal semiconductor diode, Therefore, a large value cannot be used. On the other hand, the smaller current value cannot be used because the temperature detection using the semiconductor diode is easily affected by the noise current. That is, the two constant current values I _1, I ₂ of lower, and since the upper limit is limited, these constant current value I _1, there is a limit to increase the difference in I _2, I both close values There is also a problem that the voltage between both ends generated by applying these constant current values I ₁ and I ₂ is so small that the difference between V _F1 and V _F2 becomes so small that it cannot be recognized by the AD converter.

  The present invention has been made in view of the above-described circumstances, and an object of the present invention is to provide a temperature measurement circuit that can accurately detect a temperature using a semiconductor diode.

  To achieve the above object, the present invention provides a semiconductor diode for temperature measurement, a constant current circuit for generating a constant current to be applied to the semiconductor diode based on a power supply voltage, and the semiconductor diode by applying the constant current. An AD converter that outputs a voltage between both ends of the anode and cathode of the power source voltage converted to a digital value having a predetermined number of bits using the power supply voltage as a full-scale voltage, and a constant current value of a constant current applied by the constant current circuit And a temperature measurement circuit for calculating a temperature based on a digital conversion value of the both-ends voltage output from the AD converter, comprising a reference power source for outputting a reference voltage to the AD converter, and by inputting the reference voltage When the digital conversion value of the reference voltage output from the AD converter and the power supply voltage do not include a fluctuation component, the AD conversion is performed. A voltage value of the power supply voltage including the fluctuation component is calculated based on a digital conversion value of the reference voltage output from a voltage generator, and the digital conversion value of the both-ends voltage and the constant current constant are calculated based on the voltage value. The temperature is calculated by correcting the current value.

According to the present invention, the digital conversion value of the reference voltage output from the AD converter by the input of the reference voltage and the digital conversion value of the reference voltage output from the AD converter when the fluctuation component is not included in the power supply voltage. Based on the above, the voltage value of the power supply voltage including the fluctuation component is calculated, and the digital conversion value of the both-end voltage output from the AD converter and the constant current circuit are output based on the voltage value of the power supply voltage including the fluctuation component. Since each of the constant current values of the constant current to be corrected is corrected, the AD conversion error caused by the fluctuation of the power supply voltage and the temperature corrected for the variation of the constant current value are accurately calculated.
Furthermore, compared to the conventional technique of measuring temperature using two constant current values, only one value needs to be used as the constant current value, so that each constant current value is increased so that the difference between the two constant current values is increased. Since there is no restriction of setting, the constant current value can be easily set.

  The present invention further includes an error component included in the constant current value due to the constant current circuit and the fluctuation component of the power supply voltage regardless of the presence or absence of the fluctuation component of the power supply voltage in the temperature measurement circuit. The constant current value is corrected based on the voltage value.

  According to the present invention, for the constant current value, the error component included in the constant current value due to the constant current circuit is also corrected regardless of the presence or absence of the fluctuation component of the power supply voltage. Desired.

  The present invention is also based on the error component included in the AD converter regardless of the presence or absence of the fluctuation component of the power supply voltage and the voltage value including the fluctuation component of the power supply voltage in the temperature measurement circuit. Then, the digital conversion value of the voltage between both ends is corrected.

  According to the present invention, for the digital conversion value of the both-end voltage, the error component included due to the AD converter is also corrected regardless of the presence / absence of the fluctuation component of the power supply voltage. It is done.

According to the present invention, the digital conversion value of the reference voltage output from the AD converter by the input of the reference voltage and the digital conversion value of the reference voltage output from the AD converter when the fluctuation component is not included in the power supply voltage. Based on the above, the voltage value of the power supply voltage including the fluctuation component is calculated, and the digital conversion value of the both-end voltage output from the AD converter and the constant current circuit are output based on the voltage value of the power supply voltage including the fluctuation component. Since each of the constant current values of the constant current to be corrected is corrected, the AD conversion error caused by the fluctuation of the power supply voltage and the temperature corrected for the variation of the constant current value are accurately calculated.
In the present invention, based on the error component included in the constant current value due to the constant current circuit regardless of the presence or absence of the fluctuation component of the power supply voltage and the voltage value including the fluctuation component of the power supply voltage. By correcting the constant current value, the error component included in the constant current value due to the constant current circuit is also corrected for the constant current value regardless of the presence or absence of the power supply voltage fluctuation component. Accurate temperature is required.
Further, in the present invention, the digital voltage of the both-end voltage is based on the error component included due to the AD converter regardless of the presence or absence of the fluctuation component of the power supply voltage and the voltage value including the fluctuation component of the power supply voltage. By correcting the conversion value, for the digital conversion value of the voltage at both ends, the error component included due to the AD converter is also corrected regardless of the presence or absence of the fluctuation component of the power supply voltage. Is required.

It is a figure which shows the structure of the temperature measurement circuit which concerns on embodiment of this invention. It is a circuit diagram of a reference power supply. It is a circuit diagram of a constant current circuit. It is a figure which shows the temperature measurement result of a temperature measurement circuit, (A) shows the case without correction | amendment, (B) shows the case with correction | amendment. It is a circuit diagram concerning the modification of a constant current circuit. It is a circuit diagram concerning the modification of a constant current circuit. It is a circuit diagram concerning the modification of a constant current circuit. It is a circuit diagram concerning the modification of a constant current circuit.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a diagram illustrating a configuration of a temperature measurement circuit 1 according to the present embodiment.
The temperature measurement circuit 1 is a circuit that is provided on a circuit board 3 of a semiconductor device such as an IGBT to be measured for temperature and measures a temperature T of the circuit board 3. That is, as shown in FIG. 1, the temperature measurement circuit 1 includes a semiconductor diode 5, a constant current circuit 7, a reference power supply 9, and a microcomputer 11. The constant current circuit 7, the reference power supply 9, and the microcomputer 11 include The power supply voltage HVcc is used as a power source.

The semiconductor diode 5 is a pn junction diode for temperature measurement mounted on the circuit board 3. The constant current circuit 7 is connected to the anode 5A, and the reference potential (ground) is connected to the cathode 5B. When a constant current is applied by the constant current circuit 7, an anode-cathode voltage (hereinafter, both-end voltage V _F ) is output to the microcomputer 11.
The constant current circuit 7 is a circuit that applies a constant current having a constant current value I in the forward direction of the semiconductor diode 5 during temperature measurement, and the circuit configuration thereof will be described later.

The reference power supply 9 generates a predetermined reference voltage V _ref without being affected by fluctuations in the power supply voltage HVcc and inputs it to the microcomputer 11.
Specifically, as shown in FIG. 2, the reference power supply 9 is configured as a constant voltage circuit. That is, the capacitor 20 that charges power and outputs the reference voltage V _ref , the transistor 21 that controls the amount of electricity stored in the capacitor 20, the voltage Vs obtained by dividing the power supply voltage HVcc by the resistors Ra and Rb, and the capacitor 20 An operational amplifier 22 is provided as an error amplifier that compares a reference voltage V _ref to be output and outputs a control signal to the transistor 21 so that they are always equal.
In the reference power supply 9, the output voltage is maintained at the reference voltage V _{ref under} the control of the operational amplifier 22 even when the fluctuation component ΔHVcc is included in the power supply voltage HVcc.

However, in the reference power supply 9, due to manufacturing errors of the resistors Ra and Rb that divide the power supply voltage HVcc, temperature characteristics of the resistor elements, and the like, errors ΔRa and ΔRb are included in the resistance values of the resistors Ra and Rb. For this reason, the actual value of the reference voltage V _ref output from the reference power supply 9 includes a slight error component ΔV _ref from the design value. The error component ΔV _ref of the reference voltage V _ref is measured in advance and stored in the memory 17 described later.

Returning to FIG. 1, the microcomputer 11 includes an AD converter 13 provided for each input of the both-end voltage V _F and the reference voltage V _ref , a CPU 15, and a memory 17. The CPU 15 and the AD converter 13 are The power supply voltage HVcc operates. The AD converter 13 performs AD conversion for converting the input voltage into a digital signal having a predetermined number of bits with the power supply voltage HVcc as a full scale. In this embodiment, the AD converter 13 converts the input voltage V _F and the reference voltage V _ref to digital. The conversion value X (V _F ) and the digital conversion value X (V _ref ) are output to the CPU 15.
In the following description, the resolution of the AD converter 13 is assumed to be 10 bits as an example.

CPU15 is the digital conversion value of the voltage across _{V F} X _{(V F),} errors due to characteristics of the AD converter 13, and performs a process of correcting an error due to variations in the power supply voltage HVcc, based on the fluctuation of the power supply voltage HVcc A process for correcting the constant current value I is performed, and the temperature T is calculated based on the digital conversion value X (V _F ) and the constant current value I after the correction, and another microcomputer mounted on the semiconductor device or a higher order Output to a computer. The calculation of the temperature T will be described in detail later. The memory 17 stores various parameters required for calculating the temperature T in advance.

Here, the power supply voltage HVcc is supplied from a power supply circuit (not shown) for converting the input power, but when the power input to the power supply circuit varies, the power supply voltage HVcc also varies accordingly.
For example, when the temperature measuring circuit 1 is provided in a semiconductor device such as an inverter circuit that controls a motor of an electric vehicle, the power supply voltage HVcc is obtained by converting the power of a battery mounted on the vehicle into a DC voltage by a DC / DC converter. It is generated by smoothing and transforming.
However, the voltage output by the battery varies depending on the remaining battery capacity, battery aging, load variation, battery temperature, and the like. In particular, in an in-vehicle battery, the voltage discharged from the battery is likely to fluctuate due to the characteristics of the use condition of the vehicle, and the power supply voltage HVcc generated through the DC / DC converter also fluctuates due to the above factors.

Since the AD converter 13 and the constant current circuit 7 operate based on the power supply voltage HVcc, when the power supply voltage HVcc varies, an error occurs in the digital conversion value and the constant current value.
Therefore, in this embodiment, the digital conversion value X by the AD converter 13 outputs, errors caused by fluctuations in the power supply voltage HVcc the AD converter 13 is corrected effectively by CPU15 for the voltage across V _F of the high-precision digital A conversion value X (V _F ) is obtained, a constant current value I obtained by correcting the fluctuation of the power supply voltage HVcc is obtained, and the temperature is calculated using these values without being affected by errors caused by fluctuations in the power supply voltage HVcc. I am going to do that.
Hereinafter, the configuration will be described in detail.

The resolution of the AD converter 13 is 10 bits as described above. When the reference voltage V _ref is input to the AD converter 13 as an input voltage, the AD converter 13 converts the reference voltage V _ref to digital data in 10 bits with the power supply voltage HVcc as a full scale, and the digital conversion value X (V _ref ) is output. This digital conversion value X (V _ref ) is expressed by the following equation (1).
However, as will be described later, in actuality, an error caused by the voltage fluctuation ΔHVcc of the power supply voltage HVcc occurs at the time of AD conversion by the AD converter 13, and therefore the digital conversion value X (ref) is caused by this error. The value of also changes. In the expression (1), by adding a subscript “ideal” to the digital conversion value X (V _ref ), it is clearly shown that the AD conversion is ideal without considering such an error.

Errors included in the AD conversion include errors caused regardless of fluctuations in the power supply voltage HVcc, such as quantization errors caused by characteristics of the circuit of the AD converter 13 in addition to errors caused by fluctuations in the power supply voltage HVcc. Is mentioned.
That is, when the power supply voltage HVcc is a power supply that is supplied for normal operation, the fluctuation component ΔHVcc is included, so the net value of the power supply voltage HVcc is expressed as HVcc + ΔHVcc, and also due to the characteristics of the circuit of the AD converter 13 Since the error to be performed is generally about ± 1 to 3 LSB (LSB: Least Significant Bit), the digital conversion value X (V _ref ) during normal operation is expressed by the following equation (2). The However, in equation (2), the error due to the circuit characteristics of the AD converter 13 is set to ± 3LSB.

  Note that the fluctuation component ΔHVcc is a value that can be either a positive value or a negative value, and of course, the power supply voltage HVcc for normal operation fluctuates in the decreasing direction, and “HVcc + ΔHVcc” becomes lower than HVcc. possible.

Further, considering that the error component ΔV _ref described with reference to FIG. 2 is included in the reference voltage V _ref due to an error in the circuit constant of the reference power supply 9, the expression (2) is expressed by the following expression ( It is expressed as 3).

  Using these formulas (1) and (3), the net power supply voltage HVcc + ΔHVcc including the fluctuation component ΔHVcc is expressed by the following formula (4).

In this equation (4), the power supply voltage HVcc, the reference voltage V _ref , and the reference voltage V _ref are design values, and an ideal digital conversion value X (V _ref ) ideal when AD conversion is performed based on the design value. Is uniquely determined by the number of bits of AD conversion and the like, and these values are stored in the memory 17 of the microcomputer 11 in advance. Moreover, the error component of the reference voltage _{V ref} [Delta] V _ref is previously sought by calculating the difference between the design value of the actually measured reference voltage _{V ref} the output voltage of the reference power source 9, the memory 17 is also the error component [Delta] V _ref Stored in advance.
Therefore, if the digital conversion value X (V _ref ) real obtained by actually measuring the reference voltage Vref is obtained, the net value (HVcc + ΔHVcc) including the fluctuation component ΔHVcc at that time is calculated for the power supply voltage HVcc based on the above equation (4). ) Will be required.

It should be noted that the digital conversion value X (V _ref ) ideal of the reference voltage V _ref is not obtained from a design value, but is accurate enough to be regarded as an ideal power supply voltage Vcc that does not include the fluctuation component ΔHVcc. The constant voltage source may be connected to the AD converter 13, and similarly, a high-precision reference voltage V _ref may be connected to the AD converter 13 and actually measured.

Now, at the time of temperature measurement, the voltage across V _F of the semiconductor diode 5 is AD converted is inputted to the AD converter 13 into a digital conversion value X (VF). At this time, as shown in equation (3), at the time of AD conversion, it becomes possible to AD conversion error occurs under the influence of the variation in power supply voltage HVcc, the digital conversion value _X (V F) real voltage across _{V F} Next Equation (5) is obtained.

On the other hand, the correction coefficient K for converting the digitally converted value X output from the AD converter 13 into a value having no error is expressed by the following equation (6) based on the equations (1) and (3). Is done. In this equation (6), the parameters other than “HVcc + ΔHVcc” are stored in the memory 17 in advance, and “HVcc + ΔHVcc” is based on the digital conversion value X (V _ref ) real obtained by actually measuring the reference voltage Vref. Since it is calculated | required from said Formula (4), the correction coefficient K which considered the fluctuation | variation of the power supply voltage HVcc at that time is calculated | required correctly.

Therefore, the digital conversion value X (V _F ) real of the both-end voltage V _F is obtained by using the correction coefficient K as shown in the following equation (7), the fluctuation component ΔHVcc of the power supply voltage HVcc, and the characteristics of the AD converter 13. Thus, the error included due to this is converted into a digital conversion value X (V _F ) correct corrected, and an accurate value is obtained for the both-end voltage V _F.

Next, correction of the constant current value I of the constant current circuit 7 will be described.
FIG. 3 is a circuit diagram showing an example of the constant current circuit 7.
The constant current circuit 7 includes a PNP transistor 31 that outputs a current flowing into the emitter to the collector, an operational amplifier 30 that controls the PNP transistor 31, and resistors Rc, Rd, Re, Rf, and Rg. A voltage Vt obtained by dividing the power supply voltage HVcc + ΔHVcc by the resistors Rc and Rd is input to the operational amplifier 30 as a set voltage, and the voltage at the emitter of the PNP transistor 31 is input as a feedback voltage. By controlling the PNP transistor 31 so that Vos becomes zero, the current output from the collector of the PNP transistor 31 is maintained at a constant constant current value I.

  The constant current value I is defined based on the resistance Re between the emitter and the power supply voltage HVcc + ΔHVcc and the voltage Vt which is the set voltage of the emitter. That is, since the voltage Vt is expressed by the following equation (8), the constant current value I is obtained by dividing the voltage Vt by the resistor Re.

  However, in practice, the constant current value I is an error caused by the input bias current Ib to the operational amplifier 30 and an error caused by the temperature characteristic of the input offset voltage Vos regardless of the presence or absence of the fluctuation component ΔHVcc of the power supply voltage HVcc. And, it is affected by various errors caused by the constant current circuit 7 such as an error caused by the base current IB of the PNP transistor 31. Considering these errors, the constant current value I is expressed as the following equation (9).

In Equation (9), the terms (A) to (C) correct errors caused by the constant current circuit 7, and the term (A) including the input bias current Ib as a parameter is the input bias current Ib. The term (B) including the input offset voltage Vos as a parameter indicates the error due to the temperature characteristic of the input offset voltage Vos. Further, the term (C) including the h _fe parameter of the PNP transistor 31 indicates an error caused by the base current IB.

  In this equation (9), the resistances Rc, Rd, Re, and the terms (A) to (C) are measured in advance, and the values are stored in the memory 17, whereby the above equation (4). Using the “HVcc + ΔHVcc” obtained from the above, it is possible to calculate the actual (actually varied) constant current value I reflecting the fluctuation component ΔHVcc of the power supply voltage HVcc.

At the time of temperature measurement, the microcomputer 11 takes in the reference voltage V _ref of the reference power supply 9 and performs AD conversion to obtain a digital conversion value X (V _ref ) real. The power supply voltage HVcc + ΔHVcc is obtained, and the corrected both-end voltage V _F and constant current value I are calculated based on the above formulas (7) and (9) using the power supply voltage HVcc + ΔHVcc, and these values are calculated. Is used to calculate the temperature T.

  Regarding the calculation of the temperature T, specifically, the characteristic equation of the semiconductor diode 5 is as follows: Is is a diode saturation current density, A is a fixed constant, and Eg is a band gap energy. The current density Is is expressed as Expressions (10) and (11), respectively, and Expression (12) is obtained from these Expressions (10) and (11).

  Then, when the equation (12) is developed for the temperature T and the current value Ia = the constant current value I, the equation (13) is obtained.

The microcomputer 11 calculates the temperature T by substituting the corrected both-ends voltage V _F and the constant current value I into the equation (13). As described above, the both-end voltage V _f and the constant current value I are corrected for errors caused by the AD converter 13 and the constant current circuit 7 in addition to errors based on fluctuations in the power supply voltage HVcc. Therefore, an accurate temperature T in which an error during AD conversion and variations in the constant current value I are corrected is obtained.

FIG. 4 is a diagram showing the temperature measurement result of the temperature measurement circuit 1. FIG. 4A shows the case where the temperature T is calculated without correcting the both-end voltage V _f and the constant current value I, and FIG. B) shows a case where the temperature T is calculated by correcting the both-end voltage V _f and the constant current value I. In the figure, “maximum temperature” indicates a positive error value during temperature measurement, and “minimum temperature” indicates a negative error value. That is, an error range between positive and negative detection errors is between “maximum temperature” and “minimum temperature”.
As is apparent from the comparison between FIG. 4A and FIG. 4B, the temperature T is calculated by correcting the voltage V _{f at} both ends and the constant current value I, compared with the case where the temperature T is calculated without correction. Thus, the error range of the detection error is reduced as a whole, and the accuracy of the temperature T is improved.

Thus, according to this embodiment, the digital conversion value _X (V ref) real reference voltage _{V ref} output from the AD converter 13 by the input of the reference voltage _{V ref,} the fluctuation component ΔHVcc to the supply voltage HVcc If not included, based on the digital conversion value X (V _ref ) ideal of the reference voltage V _ref output from the AD converter 13, the voltage value “HVcc + ΔHVcc” of the power supply voltage HVcc including the fluctuation component ΔHVcc is calculated, this voltage value "HVcc + ΔHVcc" digital conversion value _X (V F) across the voltage _{V F} which is output from the AD converter 13 real, and corrects the respective constant current value I of the constant current constant current circuit 7 outputs Because of this configuration, AD conversion errors caused by fluctuations in the power supply voltage HVcc and variations in the constant current value I are corrected. Temperature T is calculated accurately.
Furthermore, unlike the conventional temperature measurement circuit, there is no need for multiple constant current circuits that apply constant currents with different constant current values to the semiconductor diode, and no switching circuit for switching the applied constant current. The mounting area can be reduced. Further, when measuring the temperature using two different constant current values, there is a restriction that each constant current value is set so that the difference between the two constant current values becomes large. Since only one value needs to be used as the constant current value I, the constant current value I can be easily set.

  Further, according to the present embodiment, the constant current value I is an error component (for example, the expression (9) included in the constant current value I caused by the constant current circuit 7 regardless of the presence or absence of the fluctuation component ΔHVcc of the power supply voltage HVcc. ) (A) to (C)) are also corrected, so that a more accurate temperature is required.

Further, according to the present embodiment, error component for the digital conversion value _X (V F) real voltage across _{V F,} contained due to the AD converter 13 regardless of the presence or absence of fluctuation component ΔHVcc supply voltage HVcc Since (for example, ± 3LSB of Equation (3)) is also corrected, a more accurate temperature is obtained.

  Each embodiment described above shows only one aspect of the present invention, and can be arbitrarily modified and applied without departing from the spirit of the present invention.

  For example, in the above-described embodiment, the constant current circuit 7 may be any circuit as long as it generates and outputs a constant current using the power supply voltage HVcc as a power supply. For example, the following circuit is used. Can do.

A constant current circuit 107 shown in FIG. 5 includes an NPN transistor 40 and resistors Rm1, Rm2, and Rm3, and a constant current value that is a value obtained by dividing a voltage obtained by dividing the power supply voltage HVcc + ΔHVcc by the resistors Rm1 and Rm2 by the resistor Rm3. I is a circuit that outputs I to the emitter of the NPN transistor 40, and the semiconductor diode 5 is connected to the collector of the NPN transistor 40.
However, in consideration of an error caused by the base current IB of the NPN transistor 40 and the base-emitter voltage Vbe of the NPN transistor 40, the actual constant current value I is expressed as shown in Expression (14). 14) is used to calculate the actual constant current value I reflecting the variation.

The constant current circuit 207 shown in FIG. 6 is a so-called discharge type circuit that discharges a constant current to the semiconductor diode 5 (load) by replacing the NPN transistor 40 of the constant current circuit 107 shown in FIG. It is composed. That is, the constant current circuit 207 includes a PNP transistor 41 and resistors Rm1, Rm2, and Rm3, and a constant current value I obtained by dividing a voltage obtained by dividing the power supply voltage HVcc + ΔHVcc by the resistors Rm1 and Rm2 by the resistor Rm3. Output to the collector of the PNP transistor 41, and the semiconductor diode 5 is connected to the emitter of the PNP transistor 41.
Also in this constant current circuit 207, in consideration of an error caused by the base current IB of the PNP transistor 41 and the base-emitter voltage Vbe of the PNP transistor 41, the actual constant current value I is expressed by the equation of the constant current circuit 107. It is expressed as the same equation (15) as (14), and the actual constant current value I reflecting the variation is calculated using this equation (15).

  A constant current circuit 307 illustrated in FIG. 7 is a constant current circuit configured without using a bipolar transistor, and includes an operational amplifier 50 and resistors Rm1, Rm2, Rm3, and R. A voltage obtained by dividing the power supply voltage HVcc + ΔHVcc by the resistors Rm1 and Rm2 is input to the operational amplifier 50 as a set voltage, and a voltage Vi when an output current is passed through the resistor Rms is fed back to eliminate the difference between these voltages. By operating as described above, the output current is maintained at a constant constant current value I. The constant current value I of the constant current circuit 307 is expressed as in Expression (16), and an actual constant current value I reflecting the variation is calculated using Expression (16).

A constant current circuit 407 shown in FIG. 8 replaces the PNP transistor 31 of the constant current circuit 7 described in the above-described embodiment with an NPN transistor 43, thereby sucking a constant current from the semiconductor diode 5 (load). It is configured as a circuit. That is, the constant current circuit 407 includes an NPN transistor 43, an operational amplifier 51, and resistors Rm1, Rm2, and Rms. A constant current having a value obtained by dividing a voltage obtained by dividing the power supply voltage HVcc + ΔHVcc by the resistors Rm1 and Rm2 by the resistor Rms. The value I is extracted from the semiconductor diode 5 connected to the collector of the NPN transistor 43 and output to the emitter.
In this constant current circuit 407, an error caused by the input bias current Ib to the operational amplifier 51, an error caused by the temperature characteristic of the input offset voltage Vos, and an error caused by the base-emitter current Ibe of the NPN transistor 43, etc. In consideration of various errors caused by the constant current circuit 407, the actual constant current value I is expressed as shown in Expression (17). Using this Expression (17), the actual constant current value reflecting the variation is expressed. I is calculated.

Further, for example, in the above embodiment, the reference voltage V _ref of the reference power source _9, and it is configured to provide the AD converter 13 for each voltage across V _F of the semiconductor diode 5 is not limited to this, one AD converter The reference voltage V _ref of the reference power source 9 and the voltage V _F across the semiconductor diode 5 may be switched and input to 13.

1 Temperature Measurement Circuit 5 Semiconductor Diode 7, 107, 207, 307, 407 Constant Current Circuit 9 Reference Power Supply 11 Microcomputer 13 AD Converter 15 CPU
17 Memory HVcc Power supply voltage I Constant current value K Correction factor T Temperature V _F Both-ends voltage V _ref Reference voltage X Digital conversion value ΔHVcc Fluctuation component ΔV _ref Error component

Claims (3)

A semiconductor diode for temperature measurement;
A constant current circuit for generating a constant current applied to the semiconductor diode based on a power supply voltage;
An AD converter that converts a voltage across the anode and cathode of the semiconductor diode by applying the constant current into a digital value of a predetermined number of bits by using the power supply voltage as a full-scale voltage; and
In a temperature measurement circuit that calculates a temperature based on a constant current value of a constant current applied by the constant current circuit and a digital conversion value of the both-end voltage output from the AD converter,
A reference power supply for outputting a reference voltage to the AD converter;
The digital conversion value of the reference voltage output from the AD converter by the input of the reference voltage and the digital conversion value of the reference voltage output from the AD converter when the power supply voltage does not include a fluctuation component And calculating the voltage value of the power supply voltage including the fluctuation component, correcting the digital conversion value of the both-end voltage and the constant current value of the constant current by the voltage value, and calculating the temperature. A temperature measurement circuit characterized by that.   Regardless of the presence or absence of the fluctuation component of the power supply voltage, the constant current value is determined based on the error component included in the constant current value due to the constant current circuit and the voltage value including the fluctuation component of the power supply voltage. The temperature measurement circuit according to claim 1, wherein correction is performed.   Regardless of the presence or absence of the fluctuation component of the power supply voltage, the digital conversion value of the both-ends voltage is corrected based on the error component included due to the AD converter and the voltage value including the fluctuation component of the power supply voltage. The temperature measuring circuit according to claim 1 or 2, wherein