Resistance device, integrated circuit device, in-vivo implant device, and correction coefficient determination method

Abstract

The resistance device (100) includes a field effect transistor (TN) and a voltage application circuit (1). The voltage application circuit (1) controls the resistance value (R) between the drain and source of the field effect transistor (TN) by applying a control voltage (Vgs) corresponding to the temperature (T) between the gate and source of the field effect transistor (TN). The control voltage (Vgs) is a voltage obtained by adding a correction voltage (Vc) to a reference voltage (Vgs0). The correction voltage (Vc) depends on the temperature (T) and is set to be zero at the first temperature (T1).

Description

Technical Field

[0001] The present invention relates to a resistance device, an integrated circuit device, an in-vivo implant device, and a correction coefficient determination method. Background Art

[0002] The amplifier circuit described in Patent Document 1 includes a field-effect transistor serving as a feedback resistor and a resistance correction unit. The resistance correction unit is used to correct the resistance fluctuation between the drain and source of the field-effect transistor caused by temperature changes.

[0003] That is, when the voltage Vgs between the gate and source of the field-effect transistor is constant, due to the change in the threshold voltage Vth, when the temperature increases, the resistance Rds between the drain and source becomes lower. On the other hand, when the temperature decreases, the resistance Rds between the drain and source becomes higher. Then, the resistance correction unit controls to apply a gate voltage Vg below the threshold voltage Vth to the gate terminal of the field-effect transistor, so that the difference (Vgs - Vth) between the voltage Vgs and the threshold voltage Vth reaches a certain value, thereby making the resistance Rds between the drain and source of the field-effect transistor reach a specified value.

[0004] Specifically, the resistance correction unit has a temperature detection unit. The temperature detection unit outputs a voltage or current that linearly changes with temperature changes. The resistance correction unit applies a gate voltage Vg below the threshold voltage to the gate terminal of the field-effect transistor based on the current or voltage output by the temperature detection unit, so that the resistance Rds between the drain and source of the field-effect transistor reaches a specified value.

[0005] More specifically, the resistance correction unit further has a storage unit and an arithmetic unit. The storage unit stores the relationship between the current or voltage output by the temperature detection unit and the temperature of the temperature detection unit. Also, the storage unit stores the relationship between the temperature of the temperature detection unit and the gate voltage Vg applied to the gate terminal of the field-effect transistor.

[0006] Regarding the relationship between the current or voltage output by the temperature detection unit and the temperature of the temperature detection unit, it is stored in the storage unit by previously measuring the relationship between the temperature and the current or voltage output by the temperature detection unit. Regarding the relationship between the temperature of the temperature detection unit and the gate voltage Vg applied to the gate terminal of the field-effect transistor, the gate voltage Vg is determined at each temperature so that the difference (Vgs - Vth) between the voltage Vgs and the threshold voltage Vth between the gate and source of the field-effect transistor reaches a certain value, and then stored in the storage unit.

[0007] The arithmetic unit calculates the temperature of the temperature detection unit based on the current or voltage output by the temperature detection unit stored in the storage unit and the relationship between the temperature of the temperature detection unit. Further, the arithmetic unit determines the gate voltage Vg applied to the gate terminal of the field effect transistor based on the calculated temperature according to the relationship between the temperature of the temperature detection unit stored in the storage unit and the gate voltage Vg applied to the gate terminal of the field effect transistor.

[0008] [Patent Document]

[0009] Patent Document 1: Japanese Unexamined Patent Application Publication No. 2015-122635 Summary of the Invention

[0010] However, in Patent Document 1, there is no description of a method for determining the gate voltage Vg, that is, there is no description of how to determine the gate voltage Vg so that the difference (Vgs - Vth) between the voltage Vgs and the threshold voltage Vth of the field effect transistor reaches a certain value at multiple different temperatures.

[0011] In other words, in Patent Document 1, there is no description of a method for determining the correction value for correcting the gate voltage Vg, that is, how to correct the gate voltage Vg to reduce the temperature dependence of the drain-source resistance Rds (an example of a physical quantity related to the field effect transistor) of the field effect transistor so that the resistance Rds reaches a specified value.

[0012] An object of the present invention is to provide a resistance device, an integrated circuit device, an in-vivo implant device, and a correction coefficient determination method that can efficiently determine a combination of a correction coefficient and a physical quantity required for a field effect transistor, where the correction coefficient is a correction coefficient when correcting the control voltage applied between the gate and source to reduce the temperature dependence of the physical quantity required for the field effect transistor.

[0013] According to one aspect of the present invention, a resistance device includes a field effect transistor and a voltage application circuit. The voltage application circuit controls the resistance value between the drain and source of the field effect transistor by applying a control voltage corresponding to the temperature between the gate and source of the field effect transistor. The control voltage is a voltage obtained by adding a correction voltage to a reference voltage. The correction voltage depends on the temperature and is set to be zero at a first temperature.

[0014] In the resistance device of the present invention, the voltage application circuit preferably includes a temperature detection unit and a control voltage application unit. The temperature detection unit preferably outputs a detection signal corresponding to the temperature. The control voltage application unit preferably: generates the control voltage according to the detection signal, makes the control voltage include the correction voltage that linearly changes with respect to the temperature, and applies the control voltage between the gate and source of the field effect transistor.

[0015] In the resistance device of the present invention, preferably: the first temperature represents the temperature when the physical quantity related to the field effect transistor is substantially constant with respect to the change of the correction coefficient, and the correction coefficient is the coefficient used to determine the correction voltage.

[0016] In the resistance device of the present invention, preferably: the correction coefficient is the coefficient used to determine the correction voltage. Regarding the value of the correction coefficient, based on the reference voltage when obtaining the target physical quantity related to the field effect transistor at the first temperature, the value when obtaining the target physical quantity at a second temperature different from the first temperature is the value of the correction coefficient.

[0017] In the resistance device of the present invention, the temperature detection unit preferably includes a first current source circuit that generates a first current and a second current source circuit that generates a second current. Preferably, the temperature dependence of the first current source circuit is different from that of the second current source circuit. The first current source circuit and the second current source circuit are preferably connected in series. Preferably, the difference current between the first current and the second current is the detection signal. Preferably: by changing the current value of the first current by the first current source circuit and / or changing the current value of the second current by the second current source circuit, the first temperature is changed.

[0018] In the resistance device of the present invention, the voltage application circuit preferably: applies the control voltage between the gate and source of the field effect transistor to control the resistance value between the drain and source in the first operation region of the field effect transistor. Preferably, the first operation region refers to the region where the magnitude of the voltage between the gate and source of the field effect transistor is greater than the magnitude of the threshold voltage.

[0019] In the resistance device of the present invention, the voltage application circuit preferably: applies the control voltage between the gate and source of the field effect transistor to control the resistance value between the drain and source in the second operation region of the field effect transistor. Preferably, the second operation region refers to the region where the magnitude of the voltage between the gate and source of the field effect transistor is smaller than the magnitude of the threshold voltage.

[0020] According to another aspect of the present invention, an integrated circuit device integrates the field effect transistor and the voltage application circuit of the above-described resistance device.

[0021] According to still another aspect of the present invention, an in-vivo implant device is implanted in the body. The in-vivo implant device includes at least one of a stimulation device and a measurement device, the stimulation device being configured to issue a stimulation signal to living tissue, and the measurement device being configured to measure a living body signal. At least one of the stimulation device and the measurement device includes the above-described integrated circuit device.

[0022] According to still another aspect of the present invention, a correction coefficient determination method is used to determine a correction coefficient when correcting a control voltage applied between the gate and the source of a field effect transistor. In the correction coefficient determination method, in the following formula, "Vgs" represents the control voltage, "Vgs0" represents a reference voltage, "Vc" represents a correction voltage, "β" represents the correction coefficient, "T" represents a temperature as a variable, and "T1" represents a first temperature, that is, the temperature when the correction voltage Vc is zero. The correction coefficient determination method includes a specific voltage value determination step and a specific coefficient value determination step, the specific voltage value being a voltage value of the reference voltage Vgs0 when obtaining a target physical quantity related to the field effect transistor at the first temperature T1, and the specific coefficient value being a value of the correction coefficient β when obtaining the target physical quantity under conditions of a second temperature different from the first temperature T1 and the specific voltage value of the reference voltage Vgs0.

[0023] Vgs = Vgs0 + Vc = Vgs0 + β(T - T1)

[0024] In the method for determining the correction coefficient of the present invention, preferably, the determination process of the specific voltage value of the reference voltage Vgs0 includes: a physical quantity measurement process of measuring a physical quantity related to the field effect transistor while changing the voltage value of the reference voltage Vgs0 at the first temperature T1; and a voltage value determination process of determining the voltage value of the reference voltage Vgs0 when the physical quantity measured is substantially the same as the target physical quantity during the process of measuring a plurality of the physical quantities while changing the voltage value of the reference voltage Vgs0 as the specific voltage value of the reference voltage Vgs0. Preferably, the determination process of the specific coefficient value of the correction coefficient β includes: a physical quantity measurement process of measuring a physical quantity related to the field effect transistor while changing the value of the correction coefficient β under the conditions of the second temperature and the specific voltage value of the reference voltage Vgs0; and a coefficient value determination process of determining the value of the correction coefficient β when the physical quantity measured is substantially the same as the target physical quantity during the process of measuring a plurality of the physical quantities while changing the value of the correction coefficient β as the specific coefficient value of the correction coefficient β.

[0025] In the method for determining the correction coefficient of the present invention, preferably, the first temperature T1 refers to the temperature at which the physical quantity related to the field effect transistor is substantially constant with respect to the change of the correction coefficient β.

[0026] In the method for determining the correction coefficient of the present invention, the correction voltage Vc preferably has a value based on the difference current between the first current and the second current. The first current preferably represents a current that linearly changes with respect to the change in temperature. The second current preferably represents a current that linearly changes with respect to the change in temperature. Preferably, the temperature dependence of the first current is different from that of the second current. The method for determining the correction coefficient preferably further includes a process of changing the first temperature T1 by changing at least one of the current values of the first current and the second current.

[0027] In the method for determining the correction coefficient of the present invention, the target physical quantity preferably can be measured from an electronic circuit including the field effect transistor and includes the resistance value of the field effect transistor, and is a physical quantity set as a target value.

[0028] 〔Advantages of the Invention〕

[0029] According to the present invention, a resistance device, an integrated circuit device, an in-vivo implant device, and a correction coefficient determination method can be provided, which can efficiently determine a combination of a correction coefficient and a physical quantity related to a field-effect transistor, where the correction coefficient is a correction coefficient when correcting a control voltage applied between a gate and a source in order to reduce the temperature dependence of the physical quantity related to the field-effect transistor. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 is a diagram of the resistance device in Embodiment 1 of the present invention.

[0031] Figure 2 is a graph showing the temperature dependence of the correction voltage generated inside the voltage application circuit in Embodiment 1.

[0032] Figure 3 (a) is a graph showing the Ids-Vgs characteristics in the saturation region where Vgs > Vth of a normal NMOS transistor. Figure 3 (b) is a graph showing the Ids-Vds characteristics when Vgs > Vth of a normal NMOS transistor.

[0033] Figure 4 (a) is a graph showing the Ids-Vgs characteristics in the linear region where Vgs > Vth of a normal NMOS transistor. Figure 4 (b) is a graph showing the relationship between the resistance value and Vgs in the linear region where Vgs > Vth of a normal NMOS transistor.

[0034] Figure 5 is a graph showing the temperature dependence of the drain current in the linear region where Vgs > Vth of a normal NMOS transistor.

[0035] Figure 6 (a) is a graph showing the Ids-Vgs characteristics in the saturation region where Vgs > Vth of a normal NMOS transistor. Figure 6 (b) is a graph showing the temperature dependence of the drain current in the saturation region where Vgs > Vth of a normal NMOS transistor.

[0036] Figure 7 (a) is a circuit diagram of the first example of the voltage-controlled voltage source in Embodiment 1. Figure 7 (b) is a circuit diagram of the second example of the voltage-controlled voltage source in Embodiment 1.

[0037] Figure 8 is a circuit diagram of an example of the temperature detection unit and the correction voltage generation unit in Embodiment 1.

[0038] Figure 9 (a) is a graph showing the temperature dependence of the first current and the second current in Embodiment 1. Figure 9(b) is a graph of the temperature dependence of the differential current in Embodiment 1. Figure 9 (c) is a graph of the temperature dependence of the correction voltage in Embodiment 1.

[0039] Figure 10 (a) is a graph of the temperature dependence of the first current and the second current when changing the current value of the first current in Embodiment 1. Figure 10 (b) is a graph of the temperature dependence of the first current and the second current when changing the current value of the second current in Embodiment 1. Figure 10 (c) is a graph of the temperature dependence of the differential current when changing the current value of the differential current in Embodiment 1.

[0040] Figure 11 (a) is a graph showing the relationship between the correction coefficient and the transistor resistance value at multiple different temperatures in Embodiment 1. Figure 11 (b) is a graph of the temperature correction effect of the temperature dependence of the transistor resistance value in Embodiment 1.

[0041] Figure 12 (a) is the R-β curve in Embodiment 1, showing the transistor resistance values relative to the correction coefficient at any two different temperatures respectively. Figure 12 (b) is the R-β curve in Embodiment 1, showing the transistor resistance values relative to the correction coefficient at the first temperature and the second temperature respectively.

[0042] Figure 13 (a) is a graph showing the relationship between the reference voltage and the transistor resistance value at the first temperature in Embodiment 1. Figure 13 (b) is a graph showing the relationship between the correction coefficient and the transistor resistance value at the second temperature in Embodiment 1. Figure 13 (c) is a graph of the temperature correction effect of the temperature dependence of the transistor resistance value in Embodiment 1.

[0043] Figure 14 (a) is a semi-logarithmic graph of the Ids-Vgs characteristics of a common NMOS transistor. Figure 14 (b) is a graph of the Ids-Vds characteristics in the subthreshold region of a common NMOS transistor.

[0044] Figure 15 (a) is a graph of the Ids-Vgs characteristics in the subthreshold region of a common NMOS transistor. Figure 15 (b) is a semi-logarithmic graph of the Ids-Vgs characteristics in the subthreshold region of a common NMOS transistor. Figure 15 (c) is a graph showing the relationship between the resistance value in the subthreshold region of a common NMOS transistor and Vgs.

[0045] Figure 16(a) is a graph showing the temperature dependence of the drain current in the subthreshold region of a general NMOS transistor. Figure 16 (b) is a semi-logarithmic graph of the temperature dependence of the drain current in the subthreshold region of a general NMOS transistor.

[0046] Figure 17 (a) is a graph showing the relationship between the reference voltage and the transistor resistance value at the first temperature in the subthreshold region in Embodiment 1. Figure 17 (b) is a graph showing the relationship between the correction coefficient and the transistor resistance value at the second temperature in the subthreshold region in Embodiment 1. Figure 17 (c) is a graph showing the temperature correction effect of the temperature dependence of the transistor resistance value in the subthreshold region in Embodiment 1.

[0047] Figure 18 (a) is an explanatory diagram of the first example of the method for measuring the transistor resistance value in Embodiment 1. Figure 18 (b) is an explanatory diagram of the second example of the method for measuring the transistor resistance value in Embodiment 1.

[0048] Figure 19 (a) is a diagram of an electronic circuit device including a resistance device in Embodiment 1. Figure 19 (b) is a general diagram showing the relationship between the physical quantity measurable from the electronic circuit and the correction coefficient in Embodiment 1.

[0049] Figure 20 (a) is a circuit diagram of an integrating filter as an RC filter circuit in Embodiment 1. Figure 20 (b) is a circuit diagram of a differentiating filter as an RC filter circuit in Embodiment 1.

[0050] Figure 21 is a circuit diagram of an active filter circuit in Embodiment 1.

[0051] Figure 22 is a flowchart of the method for determining the correction coefficient in Embodiment 1.

[0052] Figure 23 (a) is Figure 22 a flowchart of step S3. Figure 23 (b) is Figure 22 a flowchart of step S4.

[0053] Figure 24 (a) to Figure 24 (d) are diagrams of resistance devices related to the first to fourth modified examples of Embodiment 1.

[0054] Figure 25 is a diagram of a resistance device in Embodiment 2 of the present invention.

[0055] Figure 26 It is a graph of the temperature dependence of the correction voltage in Embodiment 2.

[0056] Figure 27 (a) It is a graph of the temperature dependence of the first current and the second current in Embodiment 2. Figure 27 (b) It is a graph of the temperature dependence of the difference current in Embodiment 2. Figure 27 (c) It is a graph of the temperature dependence of the correction voltage in Embodiment 2.

[0057] Figure 28 (a) It is the R-β curve in Embodiment 2, showing the transistor resistance values with respect to the correction coefficient at the first temperature and the second temperature respectively. Figure 28 (b) It is a graph showing the relationship between the reference voltage and the transistor resistance value at the first temperature in Embodiment 2. Figure 28 (c) It is a graph showing the relationship between the correction coefficient and the transistor resistance value at the second temperature in Embodiment 2.

[0058] Figure 29 (a) It is an example circuit diagram of the voltage application circuit in Embodiment 2. Figure 29 (b) It is a graph of the temperature dependence of the PTAT current in Embodiment 2.

[0059] Figure 30 It is an example circuit diagram of the PTAT circuit in Embodiment 2.

[0060] Figure 31 It is a graph showing the changeability of the first temperature when the PTAT current is zero in Embodiment 2.

[0061] Figure 32 (a) to Figure 32 (d) are resistor device diagrams related to the first to fourth modified examples of Embodiment 2.

[0062] Figure 33 It is a resistor device diagram related to Embodiment 3 of the present invention.

[0063] Figure 34 It is a brain-computer interface device diagram related to Embodiment 4 of the present invention.

[0064] Figure 35 It is an example circuit diagram of an integrated circuit device related to Embodiment 4. Specific Embodiments

[0065] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals, and description thereof will not be repeated. Further, a reference symbol attached to a current is sometimes used as a symbol for the "current value" of the current, a reference symbol attached to a voltage is sometimes used as a symbol for the "voltage value" of the voltage, and a reference symbol attached to a resistor or a resistance element is sometimes used as a symbol for the "resistance value" of the resistor or the resistance element. Moreover, the same symbols in a plurality of mathematical expressions have the same definitions, and description of the same symbols is appropriately omitted. Further, in the drawings, for ease of understanding, the gate terminal is sometimes denoted by "g", the drain terminal is sometimes denoted by "d", and the source terminal is sometimes denoted by "s".

[0066] (Embodiment 1)

[0067] Refer to Figures 1 to 23 (b), and the resistance device 100 in Embodiment 1 of the present invention will be described.

[0068] Figure 1 is the resistance device 100 in Embodiment 1. As Figure 1 shown, the resistance device 100 includes a field effect transistor TN and a voltage application circuit 1. As Figure 1 shown, in Embodiment 1, the field effect transistor TN is an N-type field effect transistor. Specifically, the field effect transistor TN is an n-type MOSFET (n-type Metal-Oxide-Semiconductor Field-Effect Transistor), that is, an NMOS transistor.

[0069] Hereinafter, the field effect transistor TN is sometimes referred to as "transistor TN".

[0070] Further, the back gate terminal of the field effect transistor TN may be connected to the source terminal or the drain terminal of the field effect transistor TN, or may be connected to the ground or grounded.

[0071] The field-effect transistor TN functions as a resistance element. Specifically, the field-effect transistor TN functions as a resistance element by utilizing the resistance between the drain and source of the field-effect transistor TN. That is to say, the field-effect transistor TN functions as a MOS resistor. The fact that the field-effect transistor functions as a MOS resistor is described, for example, in "C.A. Mead, 'Analog VLSI and Neural Systems', Addison-Wesley Publishing Company, 1989." or "T. Delbruck and C.A. Mead, 'Adaptive photoreceptor with wide dynamic range', Proceedings of IEEE International Symposium on Circuits and Systems, 1994."

[0072] Specifically, the field-effect transistor TN functions as a resistance element by utilizing the resistance between the drain and source in the region (linear region and saturation region) where the voltage between the gate and source of the field-effect transistor TN is greater than the threshold voltage. Also, the field-effect transistor TN functions as a resistance element by utilizing the resistance between the drain and source in the region (subthreshold region) where the voltage between the gate and source of the field-effect transistor TN is less than the threshold voltage.

[0073] In Figure 1 the region AR, an equivalent circuit is shown when the field-effect transistor TN functions as the MOS resistor MR. The MOS resistor MR has a resistance value R, which corresponds to the resistance value between the drain and source of the field-effect transistor TN. When a voltage Vds corresponding to the voltage between the drain and source of the field-effect transistor TN is applied to the MOS resistor MR, a current Ids flows through the MOS resistor MR, which corresponds to the drain current flowing between the drain and source of the field-effect transistor TN.

[0074] Hereinafter, the resistance value R between the drain and source of the field-effect transistor TN may sometimes be referred to as "the resistance value R of the field-effect transistor TN".

[0075] The voltage application circuit 1 applies a control voltage Vgs corresponding to the temperature T between the gate and source of the field-effect transistor TN to control the resistance value R between the drain and source of the field-effect transistor TN. "Between the gate and source of the field-effect transistor TN" means "between the gate terminal and the source terminal of the field-effect transistor TN". The temperature T is the ambient temperature of the resistance device 100. The control voltage Vgs has a positive value. The control voltage Vgs represents the voltage between the gate and source of the field-effect transistor TN.

[0076] Hereinafter, the control voltage Vgs is sometimes referred to as "gate-source voltage Vgs". Also, the voltage Vds between the drain and the source is sometimes referred to as "drain-source voltage Vds".

[0077] The control voltage Vgs refers to the voltage obtained by adding the correction voltage Vc to the reference voltage Vgs0. Specifically, the control voltage Vgs is represented by Equation (1).

[0078] Vgs = Vgs0 + Vc…(1)

[0079] The correction voltage Vc is a voltage added to the reference voltage Vgs0 to reduce the temperature dependence of the physical quantity related to the field effect transistor TN.

[0080] The physical quantity related to the field effect transistor TN is a physical quantity including the resistance value of the field effect transistor TN and can be measured from an electronic circuit including the field effect transistor TN. The physical quantity related to the field effect transistor TN is, for example, the resistance value R between the drain and the source of the field effect transistor TN or the cut-off frequency fc of a filter circuit including the field effect transistor TN. The "physical quantity including the resistance value R" refers to a physical quantity that depends on the resistance value R. Hereinafter, the physical quantity required for the field effect transistor TN is sometimes referred to as the "target physical quantity". Therefore, the target physical quantity can be measured from an electronic circuit including the field effect transistor TN, is a physical quantity including the resistance value R of the field effect transistor TN, and is a physical quantity set as a target value.

[0081] Specifically, the correction voltage Vc is represented by Equation (2). In Equation (2), β is a correction coefficient, T is the temperature, and T1 is the first temperature. The correction coefficient β is a coefficient used to determine the correction voltage Vc. In the first embodiment, the correction coefficient β has a negative value. Therefore, the higher the temperature T, the smaller the correction voltage Vc. Specifically, the correction coefficient β is a coefficient for correcting the control voltage Vgs applied between the gate and the source of the field effect transistor TN to reduce the temperature dependence of the physical quantity required for the field effect transistor TN.

[0082] Vc = β(T - T1)…(2)

[0083] As shown in Equation (2), the correction voltage Vc depends on the temperature T and is set to be zero at the first temperature T1. In other words, the first temperature T1 is the temperature at which the correction voltage Vc is zero. Therefore, according to Embodiment 1, there is no correction effect at the first temperature T1. By using the voltage application circuit 1 that has no correction effect at the first temperature T1, it is possible to efficiently determine the combination of the correction coefficient β and the physical quantities related to the field effect transistor TN. The correction coefficient β is the correction coefficient when correcting the control voltage Vgs applied between the gate and source of the field effect transistor TN. This will be described in detail later.

[0084] Figure 2 is a graph of the correction voltage Vc. The vertical axis represents the correction voltage Vc [V], and the horizontal axis represents the temperature T [K]. As Figure 2 shown, the correction voltage Vc changes linearly with respect to the temperature T. The slope of the straight line representing the correction voltage Vc is the correction coefficient β.

[0085] Next, with reference to Figures 3 to 6 explain the reason for being able to correct the temperature dependence, that is, like the control voltage Vgs shown in Equation (1), by adding the correction voltage Vc that linearly changes according to the temperature T shown in Equation (2) to the reference voltage Vgs0, it is possible to correct the temperature dependence of the field effect transistor TN as the MOS resistance MR. In this case, for the sake of easy understanding, focus on the drain-source resistance value R in the "Vgs > Vth" region of the NOMS transistor. The "Vgs > Vth" region refers to the operating region of the NOMS transistor when the magnitude of the gate-source voltage Vgs (the voltage between the gate and source) is larger than the magnitude of the threshold voltage Vth. The "Vgs > Vth" region is an example of "the first operating region of the field effect transistor".

[0086] Figure 3 (a) is a graph of the Ids-Vgs characteristics in the saturation region in the "Vgs > Vth" region of a normal NMOS transistor. The horizontal axis represents the gate-source voltage Vgs [V], and the vertical axis represents the drain current Ids [μA].

[0087] Figure 3 (b) is a graph of the Ids-Vds characteristics of a normal NMOS transistor. The horizontal axis represents the drain-source voltage Vds [V], and the vertical axis represents the drain current Ids [μA]. Figure 3 (b) shows the drain currents Ids for Vgs = 0.8V, 1.0V, and 1.2V.

[0088] In the region of “Vgs>Vth”, the region of “Vds>Vgs-Vth” is the saturation region of the NMOS transistor. The drain current Ids in the saturation region is represented by Equation (3). In Equation (3), Cox represents the gate capacitance of the NMOS transistor, μn is the electron mobility of the NMOS transistor. L represents the gate length of the NMOS transistor, and W represents the gate width of the NMOS transistor.

[0089] Figure 3 (a) and Figure 3 (b) show the simulation results using the standard circuit parameters obtained from Equation (3) in the “saturation region”. As Figure 3 (a) shows, after the gate-source voltage Vgs becomes greater than the threshold voltage Vth, the drain current Ids flows. Then, as Figure 3 (b) shows, in the saturation region, the drain current Ids is almost a constant value and reaches saturation. That is, in the saturation region, the drain current Ids does not depend on the drain-source voltage Vds. The region generally used in NMOS transistors is the saturation region.

[0090]

Number 1

[0091]

[0092] On the other hand, in the region of “Vgs>Vth”, the region of “Vds<Vgs-Vth” is the linear region of the NMOS transistor. The drain current Ids in the linear region is represented by Equation (4). Figure 3 (b) shows the simulation results using the standard circuit parameters obtained from Equation (4) in the “linear region”. As Figure 3 (b) shows, in the “linear region”, the drain current Ids changes linearly with respect to the drain-source voltage Vds.

[0093]

Number 2

[0094]

[0095] That is, in the linear region, the Ids-Vds characteristic can be linearly approximated. Therefore, as a preferred example, the NMOS transistor can be easily used as a MOS resistor. Specifically, through the tangent equation of Equation (4) based on Vds = 0, the drain current Ids can be linearly approximated as in Equation (5).

[0096]

Number 3

[0097]

[0098] As can be seen from Equation (5), the drain current Ids can be controlled by the gate-source voltage Vgs. Controlling the drain current Ids by Vgs has the same meaning as controlling the resistance value R of the NMOS transistor by Vgs. The reason is that R = Vds / Ids. Figure 3 The straight lines shown by the dashed lines in (b) represent the simulation results using the standard circuit parameters obtained from Equation (5) when Vgs = 0.8V, 1.0V, and 1.2V. The smaller the Vgs, the smaller the slope of the straight line shown by the dashed line. That is, the smaller the Vgs, the higher the resistance value R.

[0099] Among them, the relationship between the gate-source voltage Vgs [V], the drain current Ids [nA], and the resistance value R [MΩ] is shown in Figure 4 (a) and Figure 4 (b). Figure 4 (a) and Figure 4 (b) are the simulation results using the standard circuit parameters based on Equation (4).

[0100] As Figure 4 (a) shows, as an example, the drain current Ids increases linearly corresponding to the increase in the gate-source voltage Vgs within the range of several hundred mV. That is, the drain current Ids is proportional to the gate-source voltage Vgs within the range of several hundred mV.

[0101] Since the resistance value R of the NMOS transistor as the MOS resistor is represented by Equation (6), as Figure 4 (b) shows, the resistance value R is inversely proportional to the gate-source voltage Vgs.

[0102] R = Vds / Ids…(6)

[0103] In addition, Figure 3 (a) to Figure 4 (b) the standard circuit parameter simulation conditions are μn·Cox = 170.0 μA / V 2 , W = 0.6 μm, L = 60.0 μm, αth = -1.7 mV / K, αμ = -1.5, Vth = 0.7V, T0 = 300K, C = 800.0 fF.

[0104] As described above with reference to Figure 3 (a) to Figure 4(b) As described, the MOS resistance of the NMOS transistor can be used as a linear resistor in the linear region of the "Vgs > Vth" region. For example, the current flowing through the MOS resistance is proportional to the gate-source voltage Vgs, and the cut-off frequency fc of the RC filter composed of the MOS resistance is proportional to the gate-source voltage Vgs. That is, the current flowing through the MOS resistance and the cut-off frequency fc of the RC filter can be linearly controlled by Vgs.

[0105] However, the threshold voltage Vth and electron mobility μn of the NMOS transistor have temperature dependence. The threshold voltage Vth considering temperature dependence is expressed by Equation (7). Here, αth is the temperature coefficient of the threshold voltage Vth. Vth_T0 is the threshold voltage Vth at temperature T0. That is, as shown in Equation (7), with the threshold voltage Vth_T0 obtained as an observation value at an arbitrary temperature T0 as a boundary, corresponding to the change in temperature starting from T0, the threshold voltage Vth changes from Vth_T0.

[0106]

Equation 4

[0107] V th =V th _T 0 +α th (T-T 0 )

[0108] …(7)

[0109] The electron mobility μn considering temperature dependence is expressed by Equation (8). Here, αμ is the temperature coefficient of the electron mobility μn. μn_T0 is the electron mobility μn at temperature T0. That is, as shown in Equation (8), with the electron mobility μn_T0 obtained as an observation value at an arbitrary temperature T0 as a boundary, corresponding to the change in temperature starting from T0, the electron mobility μn changes from μn_T0.

[0110]

Equation 5

[0111]

[0112] Based on Equation (4), Equation (7), and Equation (8), in the linear region of the NMOS transistor, the drain current Ids reflecting the temperature dependence of both the threshold voltage Vth and the electron mobility μn is expressed by Equation (9).

[0113]

Equation 6

[0114]

[0115] Figure 5It is a graph showing the temperature dependence of the drain current Ids in the linear region of the "Vgs > Vth" region of a general NMOS transistor. The horizontal axis represents the temperature [K], and the vertical axis represents the drain current Ids [nA]. Figure 5 In it, line 500 is the drain current Ids when αμ = 0 in Equation (9) and only the temperature dependence of the threshold voltage Vth is reflected. Line 501 is the drain current Ids when αth = 0 in Equation (9) and only the temperature dependence of the electron mobility μn is reflected. Line 502 is the drain current Ids represented by Equation (9) when the temperature dependencies of both the electron mobility μn and the threshold voltage Vth are reflected.

[0116] The temperature coefficient αth of the threshold voltage Vth in Equation (7) and Equation (9) takes a negative value. Therefore, as Figure 5 shown by line 500, when only the temperature dependence of the threshold voltage Vth is considered, the drain current Ids increases linearly as the temperature rises. Qualitatively, the reason is that the number of movable charged particles increases as the temperature rises, resulting in a decrease in the threshold voltage Vth.

[0117] Also, the temperature coefficient αμ of the electron mobility μn in Equation (8) and Equation (9) takes a negative value. Therefore, as Figure 5 shown by line 501, when only the temperature dependence of the electron mobility μn is considered, the drain current Ids basically decreases linearly as the temperature rises. Qualitatively, the reason is that the thermal vibration of the silicon lattice hinders the movement of charged particles when the temperature rises.

[0118] That is to say, as shown by line 500 and line 501, the drain current Ids after reflecting the temperature dependence of the threshold voltage Vth and the drain current Ids after reflecting the temperature dependence of the electron mobility μn can be linearly approximated.

[0119] Therefore, as shown by line 502, the drain current Ids reflecting both the temperature dependence of the threshold voltage Vth and the temperature dependence of the electron mobility μn can also be linearly approximated.

[0120] Specifically, as shown by line 502, the drain current Ids basically increases linearly as the temperature rises. The reason is that the temperature dependence of the threshold voltage Vth has a greater influence on the drain current Ids than the temperature dependence of the electron mobility μn. In other words, the influence of the temperature dependence of the threshold voltage Vth is suppressed by the influence of the temperature dependence of the electron mobility μn.

[0121] As an example, the temperature coefficient αth of the standard threshold voltage Vth is -1.7 mV / K. Corresponding to a temperature rise of 50 degrees, the threshold voltage Vth drops by 85 mV. Moreover, the amount of decrease in the threshold voltage Vth (85 mV) becomes narrower (less than 85 mV) due to the suppression effect of the temperature dependence of the electron mobility μn.

[0122] On the other hand, as shown in Figure 4 (a), as an example, when the gate-source voltage Vgs is increased by 100 mV, the increase in the drain current Ids can be linearly approximated.

[0123] Therefore, corresponding to the drain current Ids ( Figure 5 of line 502) that linearly increases with temperature rise, the gate-source voltage Vgs linearly decreases, thereby linearly reducing the drain current Ids, and thus the temperature dependence of the drain current Ids can be suppressed. As a result, in the linear region of the "Vgs > Vth" region, the temperature dependence of the NMOS transistor as a MOS resistance can be corrected.

[0124] As described above, in the linear region of the "Vgs > Vth" region, like the control voltage Vgs shown in Equation (1) applied between the gate and source of the transistor TN, by adding the correction voltage Vc that linearly changes corresponding to the temperature T shown in Equation (2) to the reference voltage Vgs0, the temperature dependence of the transistor TN as a MOS resistance MR can be corrected.

[0125] Similarly, in the saturation region of the "Vgs > Vth" region of the transistor TN, like the control voltage Vgs shown in Equation (1), by adding the correction voltage Vc that linearly changes corresponding to the temperature T shown in Equation (2) to the reference voltage Vgs0, the temperature dependence of the transistor TN as a MOS resistance can also be corrected. Regarding this point, refer to Figure 6 (a) and Figure 6 (b) for explanation.

[0126] Figure 6 (a) shows the drain current Ids in the saturation region of the "Vgs > Vth" region of a general NMOS transistor obtained according to Equation (3) with respect to the gate-source voltage Vgs. As shown in Figure 6 (a), compared with the linear region of Figure 4 (a), although the drain current Ids is non-linear, it can be linearly approximated within a range of about 100 mV.

[0127] On the other hand, the drain current Ids in the saturation region of the NMOS transistor reflecting the temperature dependence is represented by Equation (10) based on Equation (3), Equation (7), and Equation (8).

[0128] [Number 7]

[0129]

[0130] Figure 6 (b) is a graph showing the temperature dependence of the drain current Ids in the saturation region of the "Vgs > Vth" region of a normal NMOS transistor. The horizontal axis represents the temperature [K], and the vertical axis represents the drain current Ids [nA]. Figure 6 In (b), line 503 represents the drain current Ids when αμ = 0 in Equation (10) and only the temperature dependence of the threshold voltage Vth is reflected. Line 504 represents the drain current Ids when αth = 0 in Equation (10) and only the temperature dependence of the electron mobility μn is reflected. Line 505 is the drain current Ids represented by Equation (10) when the temperature dependencies of both the electron mobility μn and the threshold voltage Vth are reflected.

[0131] As shown by line 505, in the saturation region of the "Vgs > Vth" region, similar to the linear region, the drain current Ids that reflects both the temperature dependence of the threshold voltage Vth and the temperature dependence of the electron mobility μn can also be linearly approximated. As an example, the temperature coefficient αth of the standard threshold voltage Vth is -1.7 mV / K. Due to the suppression effect caused by the temperature dependence of the electron mobility μn, in the saturation region as well as in the linear region, with a 50-degree temperature increase, the decrease in the threshold voltage Vth is less than 85 mV. On the other hand, as Figure 6 (a) shows, as an example, the increase in the drain current Ids when the gate-source voltage Vgs is increased by 100 mV can be linearly approximated.

[0132] Therefore, corresponding to the drain current Ids ( Figure 6 (line 505 in (b)) that linearly increases with temperature rise, the gate-source voltage Vgs is linearly decreased, thereby linearly reducing the drain current Ids, and thus the temperature dependence of the drain current Ids can be suppressed. As a result, in the saturation region of the "Vgs > Vth" region, the temperature dependence of the NMOS transistor as a MOS resistor can be corrected.

[0133] As can be seen from the above, in the saturation region of the "Vgs > Vth" region, similar to the control voltage Vgs shown in Equation (1) applied between the gate and source of the transistor TN, by adding the correction voltage Vc that linearly varies with the temperature T shown in Equation (2) to the reference voltage Vgs0, the temperature dependence of the transistor TN as a MOS resistor MR can also be corrected.

[0134] In addition, Figures 5 to 6The standard circuit parameter simulation conditions in (b) are μn_T0·Cox = 170.0 μA / V 2 , W = 0.6 μm, L = 60.0 μm, αth = -1.7 mV / K, αμ = -1.5, Vth_T0 = 0.7 V, T0 = 300 K, C = 800.0 fF.

[0135] As described with reference to Figure 3 (a) to Figure 6 (b), in both the linear region and the saturation region, in the region of "Vgs > Vth" where the gate-source voltage Vgs is greater than the threshold voltage Vth, like the control voltage Vgs shown in Equation (1) applied between the gate and source of the transistor TN, by adding the correction voltage Vc that linearly varies corresponding to the temperature T shown in Equation (2) to the reference voltage Vgs0, the temperature dependence of the transistor TN as the MOS resistance MR can be corrected.

[0136] According to the method based on Equation (1) and Equation (2), an example of a correction circuit for the temperature dependence of the MOS resistance MR composed of the transistor TN is Figure 1 the voltage application circuit 1. Referring again to Figure 1 , the voltage application circuit 1 will be described in detail. The voltage application circuit 1 is arranged between the gate terminal and the source terminal of the transistor TN. The potential of the source terminal (hereinafter, sometimes referred to as "source potential Vs") can take any value. Specifically, since the two terminals of the drain terminal and the source terminal of the transistor TN are used as the two terminals at both ends of the MOS resistance MR, it is used in a state independent of being connected to the ground or grounded (0 [V]) (for example, a floating ground state). As a result, the source potential Vs can take any value.

[0137] The voltage application circuit 1 includes a control voltage application unit 9 and a temperature detection unit 13. The temperature detection unit 13 detects the temperature T and outputs a detection signal TM corresponding to the temperature T to the control voltage application unit 9. The temperature detection unit 13 only needs to be able to detect the temperature T and output the detection signal TM, and the structure of the temperature detection unit 13 is not particularly limited. The detection signal TM represents a physical quantity indicating the temperature T (for example, current or voltage), or the detection signal TM represents a physical quantity related to the temperature T (for example, current or voltage). For example, the temperature detection unit 13 may also include a temperature sensor such as a thermistor. For example, the temperature detection unit 13 may also include a field effect transistor or a bipolar transistor, and utilize the temperature-dependent characteristics of the field effect transistor or the bipolar transistor. For example, the temperature detection unit 13 may also include a PTAT (Proportional To Absolute Temperature) circuit. The PTAT circuit outputs a current proportional to the absolute temperature as the detection signal TM. For example, the temperature detection unit 13 is composed of a temperature detection circuit that detects the temperature T and outputs the detection signal TM.

[0138] The control voltage application unit 9 applies a control voltage Vgs corresponding to the detection signal TM indicating the temperature T between the gate and source of the transistor TN. Specifically, the control voltage application unit 9 generates the control voltage Vgs corresponding to the detection signal TM, and the control voltage Vgs includes a correction voltage Vc that linearly changes with respect to the temperature T. Then, the control voltage application unit 9 applies the control voltage Vgs between the gate and source of the transistor TN. Therefore, according to the first embodiment, the temperature dependence of the resistance value R of the transistor TN can be appropriately reduced according to the detection signal TM indicating the temperature T.

[0139] Specifically, the control voltage application unit 9 includes a control voltage generation unit 10 and a voltage-controlled voltage source 19. Moreover, the control voltage generation unit 10 includes a reference voltage generation unit 11, a correction voltage generation unit 15, and an addition calculation unit 17. The reference voltage generation unit 11 generates the reference voltage Vgs0 in Equation (1) and outputs the reference voltage Vgs0 to the addition calculation unit 17. The correction voltage generation unit 15 generates a correction voltage Vc that linearly changes with respect to the temperature T based on the detection signal TM of the temperature and outputs the correction voltage Vc to the addition calculation unit 17. The addition calculation unit 17 performs an addition calculation on the reference voltage Vgs0 and the correction voltage Vc to generate a control voltage Vgsa as the addition calculation result. Thus, the control voltage generation unit 10 generates the control voltage Vgsa. The control voltage Vgsa may also be referred to as the "reference control voltage Vgsa". The control voltage Vgsa is represented by Equation (11).

[0140] Vgsa = Vgs0 + Vc…(11)

[0141] It can be clearly seen from Equation (1) and Equation (11) that the control voltage Vgsa has the same voltage components (reference voltage Vgs0 and correction voltage Vc) and the same voltage value as the control voltage Vgs. In Equation (11), the correction voltage Vc is also represented by Equation (2). As Figure 2 shown, the correction voltage Vc shown in Equation (2) linearly decreases as the temperature T rises. That is to say, in the case of the transistor TN, the correction coefficient β is negative. Then, by changing the correction coefficient β, the ratio of reducing the control voltage Vgsa according to the temperature T can be adjusted.

[0142] Specifically, the control voltage Vgsa is a voltage based on 0 [V] (that is, a potential difference based on 0 [V]). On the other hand, the source potential Vs of the transistor TN can take any value. Therefore, when the control voltage Vgsa is directly applied between the gate and the source, when the source potential Vs of the transistor TN takes any value independent of being connected to the ground or grounded (0 [V]), "Vgsa - Vs" becomes the voltage between the gate and the source of the transistor TN. As a result, the resistance value R of the transistor TN can vary according to the source potential Vs. Among them, in Embodiment 1, the control voltage Vgsa is indirectly applied between the gate and the source of the transistor TN. As a preferred example of this case, in Embodiment 1, the control voltage application unit 9 has a voltage-controlled voltage source 19. However, for example, in the case where the source potential Vs of the transistor TN is a certain value and / or when the change in the resistance value R with the source potential Vs can be tolerated, the control voltage Vgsa can also be directly applied between the gate and the source of the transistor TN.

[0143] The voltage-controlled voltage source 19 is connected between the gate terminal and the source terminal of the transistor TN. The voltage-controlled voltage source 19 has two input terminals and two output terminals. Therefore, the voltage-controlled voltage source 19 is a voltage source that determines the potential difference between the two output terminals according to the potential difference between the two input terminals. In the voltage-controlled voltage source 19, the control voltage Vgsa based on 0 [V] and the reference voltage 0 [V] are input by the control voltage generation unit 10, and thus the control voltage Vgsa is input as a potential difference. Then, by connecting the two output terminals of the voltage-controlled voltage source 19 to the gate terminal and the source terminal of the transistor TN respectively, a control voltage Vgs having the same voltage value as the control voltage Vgsa is applied between the gate and the source of the transistor TN. At this time, even if the source potential Vs changes, the potential difference between the two output terminals of the voltage-controlled voltage source 19 does not change, that is, the control voltage Vgs does not change.

[0144] Further, in the voltage-controlled voltage source 19, it is only necessary to input the control voltage Vgsa as a potential difference, and the reference voltage of 0 [V] can also be any value. In this case, if the reference voltage is Vref and the output voltage from the control voltage generation unit 10 is "Vgsa + Vref", then the potential difference input to the voltage-controlled voltage source 19 is Vgsa of "Vgsa + Vref - Vref".

[0145] As described above, Figure 1 the voltage application circuit 1 can apply any control voltage Vgs shown in Equation (1) between the gate and source of the transistor TN. Specifically, the drain current Ids of the MOS resistor MR composed of the transistor TN increases approximately linearly as the temperature T rises ( Figure 5 line 502 of Figure 6 (b)), by appropriately setting the correction coefficient β of the correction voltage Vc ( Figure 2 ) in the voltage application circuit 1, the voltage application circuit 1 linearly reduces the control voltage Vgs according to the temperature T, and linearly reduces the drain current Ids. As a result, the temperature dependence of the drain current Ids can be suppressed. By suppressing the temperature dependence of the drain current Ids, the temperature dependence of the physical quantity related to the transistor TN can be suppressed.

[0146] Regarding the specific circuit, first refer to Figure 7 (a) and Figure 7 (b) to describe an example of the voltage-controlled voltage source 19. Figure 7 (a) is a circuit diagram of the first example of the voltage-controlled voltage source 19. As Figure 7 (a) shows, the voltage-controlled voltage source 19 according to the first example includes a first switch circuit 191, a second switch circuit 192, and a capacitor 193.

[0147] The first switch circuit 191 includes terminals t1 to t3. Terminal t1 is connected to the control voltage generation unit 10. Terminal t2 is connected to the gate terminal of the transistor TN. Terminal t3 is connected to one terminal of the capacitor 193.

[0148] The second switch circuit 192 includes terminals t4 to t6. Terminal t4 is connected to the control voltage generation unit 10. Terminal t5 is connected to the source terminal of the transistor TN. Terminal t6 is connected to the other terminal of the capacitor 193.

[0149] The control voltage generation unit 10 generates a control voltage Vgsa. The first switch circuit 191 connects the terminal t3 to the terminal t1. Also, the second switch circuit 192 connects the terminal t6 to the terminal t4. As a result, the capacitor 193 holds the control voltage Vgsa. Then, the first switch circuit 191 connects the terminal t3 to the terminal t2. Also, the second switch circuit 192 connects the terminal t6 to the terminal t5. As a result, the control voltage Vgsa held by the capacitor 193 is applied as the control voltage Vgs between the gate and source of the transistor TN.

[0150] Figure 7 (b) is a circuit diagram of a second example of the voltage-controlled voltage source 19. As Figure 7 (b) shows, the voltage-controlled voltage source 19 according to the second example further includes an operational amplifier 194 on the basis of the structure of the voltage-controlled voltage source 19 according to the first example. Hereinafter, the differences from the first example in the second example will be mainly described.

[0151] In the second example, the output terminal of the operational amplifier 194 is connected to the node N. The node N is located on the line connecting the terminal t2 and the gate terminal of the transistor TN. The inverting input terminal of the operational amplifier 194 is connected to the terminal t5. The non-inverting input terminal of the operational amplifier 194 is connected to the source terminal of the transistor TN. When the operational amplifier 194 forms a feedback circuit as Figure 7 (b), the potentials of the non-inverting input terminal and the inverting input terminal are substantially equal (virtual short circuit). Therefore, the potential of the inverting input terminal is equal to the source potential Vs of the transistor TN. As a result, the capacitor 193 holds the control voltage Vgsa. After the capacitor 193 is connected to the operational amplifier 194, the potential of the node N at the output terminal of the operational amplifier 194 becomes "Vgsa + Vs". Thus, the gate-source voltage Vgs of the transistor TN is substantially equal to the control voltage Vgsa.

[0152] That is, the operation of the voltage-controlled voltage source 19 according to the second example is the same as the operation of the voltage-controlled voltage source 19 according to the first example. In particular, in the second example, since the source terminal of the transistor TN is connected to the non-inverting input terminal of the operational amplifier 194, the influence of the capacitive load of the gate terminal and the source terminal of the transistor TN can be reduced as compared with the first example.

[0153] In addition, as long as the voltage-controlled voltage source 19 can be inserted between any two floating ground terminals, the structure of the voltage-controlled voltage source 19 is not particularly limited.

[0154] Regarding the structure of the control voltage generation unit 10, as long as it can generate the control voltage Vgsa represented by the formula (11), there is no particular limitation, and it can be constituted by any control voltage generation circuit.

[0155] In addition, Figure 1 shows the physical or logical structure of the control voltage generation unit 10. Therefore, in Figure 1 when showing the physical structure of the control voltage generation unit 10, for example, the reference voltage generation unit 11 is composed of a reference voltage generation circuit that generates the reference voltage Vgs0, the correction voltage generation unit 15 is composed of a correction voltage generation circuit that generates the correction voltage Vc based on the detection signal TM of the temperature detection unit 13, and the addition calculation unit 17 is composed of an addition calculation circuit that adds the correction voltage Vc to the reference voltage Vgs0.

[0156] Furthermore, in Figure 1 when showing the logical structure of the control voltage generation unit 10, for example, even when the reference voltage generation unit 11, the correction voltage generation unit 15, and the addition calculation unit 17 are not clearly distinguishable in terms of physical structure, as long as the control voltage generation unit 10 can generate the control voltage Vgsa shown in Equation (11), the circuit constituting the control voltage generation unit 10 is not particularly limited.

[0157] Furthermore, the same applies to the structures of the temperature detection unit 13 and the correction voltage generation unit 15. Figure 1 shows the physical or logical structure of the temperature detection unit 13 and the correction voltage generation unit 15. Therefore, in the case where the temperature detection unit 13 and the correction voltage generation unit 15 show a logical structure, for example, even when the temperature detection unit 13 and the correction voltage generation unit 15 are not clearly distinguishable in terms of physical structure, as long as they can generate the correction voltage Vc shown in Equation (2), there is no particular limitation, and they can be constituted by any temperature detection circuit and correction voltage generation circuit.

[0158] Referring to Figure 8 , an example of the temperature detection unit 13 and the correction voltage generation unit 15 will be described. Figure 8 is an example circuit diagram of the temperature detection unit 13 and the correction voltage generation unit 15. The temperature detection unit 13 and the correction voltage generation unit 15 include a first current source circuit 131, a second current source circuit 133, and a variable resistor Ro. The first current source circuit 131 and the second current source circuit 133 are connected in series between the first power supply line PL1 and the second power supply line PL2. One terminal of the variable resistor Ro is connected to the node Nc between the first current source circuit 131 and the second current source circuit 133. The other terminal of the variable resistor Ro is grounded. In this case, the potential of the first power supply line PL1 is positive. For example, the first power supply line PL1 is connected to a positive power supply that provides a positive power supply voltage. On the other hand, the potential of the second power supply line PL2 is negative. For example, the second power supply line PL2 is connected to a negative power supply that provides a negative power supply voltage.

[0159] The first current source circuit 131 generates a first current Ip. The second current source circuit 133 generates a second current Im. A differential current Io flows through the variable resistor Ro. The differential current Io is a current representing the difference between the first current Ip and the second current Im. Specifically, the differential current Io is the current obtained by subtracting the second current Im from the first current Ip (Io = Ip - Im). By the differential current Io flowing through the variable resistor Ro, a potential difference Vc is generated between both ends of the variable resistor Ro. The potential difference Vc is the value obtained by multiplying the resistance value Ro by the current value Io (Vc = Ro × Io).

[0160] Figure 8 In the circuit of, the temperature detection unit 13 and the correction voltage generation unit 15 are not clearly distinguished, and the temperature dependencies of the first current source circuit 131 and the second current source circuit 133 are used in temperature detection. The temperature dependency graphs of the first current Ip and the second current Im are shown in Figure 9 (a). The horizontal axis is the temperature T [K], and the vertical axis is the current value I [A] of each current source circuit. As Figure 9 (a) shows, both the first current Ip and the second current Im change linearly with respect to the change in temperature T. Also, the temperature dependency of the first current Ip is different from that of the second current Im. That is, the temperature dependency of the first current source circuit 131 of the temperature detection unit 13 is different from that of the second current source circuit 133.

[0161] Figure 9 In the example of (a), the temperature dependency of the first current Ip is lower than that of the second current Im. That is, the temperature dependency of the first current source circuit 131 is lower than that of the second current source circuit 133. At this time, the temperature dependency of the differential current Io (= Ip - Im) flowing through the variable resistor Ro is shown in Figure 9 (b). The horizontal axis is the temperature T [K], and the vertical axis is the differential current value Io [A]. As Figure 9 (b) shows, the differential current Io has a negative temperature characteristic. That is, the slope A of the straight line representing the differential current Io has a negative value. Also,[[]] Figure 9 in (a), the temperature T at which the first current Ip and the second current Im are the same is the first temperature T1. At the first temperature T1, the differential current Io is zero as shown in Figure 9 (b). Therefore, the differential current Io is represented by Equation (12). Among them, it can be considered that the first current source circuit 131 and the second current source circuit 133 connected in series constitute Figure 1 the temperature detection unit 13 in. Also, it can be considered that the differential current Io is Figure 1 the detection signal TM of the temperature detection unit 13 in.

[0162] Io = A × (T - T1)…(12)

[0163] The correction voltage Vc when the differential current Io is input to the variable resistor Ro is represented by Equation (13). The temperature dependence graph of the correction voltage Vc is shown in Figure 9 (c). The horizontal axis is the temperature T [K], and the vertical axis is the correction voltage Vc [V]. Since the differential current Io has a negative temperature characteristic, as Figure 9 (c) shows, the correction voltage Vc also has a negative temperature characteristic. That is, it is consistent with the temperature dependence graph of the correction voltage Vc shown in Figure 2 . Also, when the temperature is the first temperature T1, the correction voltage Vc is zero as shown in Figure 9 (c). Among them, the correction coefficient β is "Ro × A" as shown in Equation (13), and like the slope A of Equation (12), it takes a negative value. Also, it can be seen that when it is desired to change the correction coefficient β, it is only necessary to change the variable resistor Ro. It can also be considered that the variable resistor Ro is the Figure 1 correction voltage generation unit 15 in. Also, it is clear from Equation (13) that the correction voltage Vc has a value based on the differential current Io.

[0164] Vc = Ro × Io = Ro × A × (T - T1) = β(T - T1)

[0165] …(13)

[0166] Among them, the first temperature T1 when the differential current Io and the correction voltage Vc are zero can be changed by the following method. Refer to Figure 8 and Figure 10 (a) to Figure 10 (c), the method for changing the first temperature T1 shown in Equation (2), Equation (12), and Equation (13) will be described.

[0167] The temperature dependence graph of the first current Ip and the second current Im when changing the current value of the first current Ip is shown in Figure 10 (a). As shown in Figure 10 (a), the temperature when the first current Ip is equal to the second current Im is the first temperature T1. Therefore, when the current value of the first current Ip increases, the first temperature T1 also increases. That is, by changing the current value of the first current Ip through the first current source circuit 131, the first temperature T1 can be changed.

[0168] By changing the current value of the second current Im through the second current source circuit 133, the first temperature T1 can also be changed. Figure 10 (b) shows the temperature dependence graph of the first current Ip and the second current Im when changing the current value of the second current Im. Among them, as Figure 10 (b) shows, when the current value of the second current Im increases, the first temperature T1 decreases.

[0169] When changing the current values of the first current Ip and / or the second current Im, the temperature dependence of the differential current Io also changes according to the current values. Figure 10 (c) represents a graph of the temperature dependence of the differential current Io when changing the current value of the differential current Io. As Figure 10 shown in (c), as the current value of the differential current Io increases due to the change in the current values of the first current Ip and / or the second current Im, the temperature T (i.e., the first temperature T1) at which the differential current Io becomes zero also increases. Also, when the differential current Io is zero, the correction voltage Vc is also zero.

[0170] That is to say, in the Figure 8 circuit example of Embodiment 1, by providing the first current source circuit 131 and the second current source circuit 133, changing the first current Ip and / or the second current Im, and changing the differential current Io, it is possible to easily change the temperature T at which the correction voltage Vc becomes zero, that is, it is possible to easily change the first temperature T1.

[0171] In addition, the second power supply line PL2 may also be grounded. In this case, one terminal of the variable resistor Ro is connected to the node Nc, and the other terminal of the variable resistor Ro is connected to a reference voltage source that generates a reference voltage Vref (0 < Vref < potential of PL1). At this time, the output voltage of the correction voltage generation unit 15 is "β(T - T1) + Vref". Among them, by setting the output voltage of the reference voltage generation unit 11 to "Vgs0 - Vref", the same as in Equation (11), Vgsa = Vgs0 + β(T - T1).

[0172] Also, Figure 9 (a) and Figure 10 in the example of, the slopes of the two straight lines representing the first current Ip and the second current Im are both positive values, but they do not have to be positive values. In Embodiment 1, for correcting the temperature dependence of the transistor TN, as long as the correction coefficient β of the correction voltage Vc is negative, as long as the slope of the second current Im is greater than the slope of the first current Ip, the sign of the slope is irrelevant.

[0173] As described above, through the Figure 1 structure of the resistance device 100 in Embodiment 1 shown, it is illustrated that the temperature dependence of the transistor TN can be corrected by appropriately setting the correction coefficient β of the correction voltage Vc. According to this structure, substituting Equation (1) and Equation (2) into Equation (9), in the linear region (Vds < Vgs - Vth) in the "Vgs > Vth" region, the drain current Ids of the transistor TN is represented by Equation (14) that reflects the correction coefficient β.

[0174]

Equation 8

[0175]

[0176] Further, by substituting Equation (1) and Equation (2) into Equation (10), in the saturation region (Vds > Vgs - Vth) in the "Vgs > Vth" region, the drain current Ids of the transistor TN is represented by Equation (15) that reflects the correction coefficient β.

[0177]

Equation 9

[0178]

[0179] In Equation (14) and Equation (15), a correction factor that linearly varies with respect to the temperature T is included, that is, the correction voltage Vc (= β(T - T1)). Further, in the linear region and the saturation region in the "Vgs > Vth" region, the resistance value R of the transistor TN is represented by Equation (16).

[0180] R = Vds / Ids…(16)

[0181] By appropriately setting the correction coefficient β, the temperature dependence of the drain current Ids of the transistor TN is eliminated. In other words, by appropriately setting the correction coefficient β, the temperature dependence of the resistance value R (Equation (16)) based on the drain current Ids is eliminated. Refer to Figure 11 (a), an example of a method for determining the correction coefficient β for determining the correction voltage Vc is described. In this example, the saturation region of the transistor TN is focused on.

[0182] Figure 11 (a) is a relationship chart of the correction coefficient β and the resistance value R of the transistor TN at multiple different temperatures. The horizontal axis represents the correction coefficient β, and the vertical axis represents the resistance value R [MΩ].

[0183] As Figure 11 (a) shows, based on Equation (15) and Equation (16) in the saturation region in the "Vgs > Vth" region, the resistance value R is simulated to obtain the R-β curves G10 to G16. The simulation conditions for the standard circuit parameters are μn_T0·Cox = 170.0 μA / V 2 , W = 0.6 μm, L = 60.0 μm, αth = -1.7 mV / K, αμ = -1.5, Vth_T0 = 0.7 V, T0 = 300 K, T1 = 320 K. Further, the reference voltage Vgs0 is 0.9 V, and the drain-source voltage Vds is 1.8 V.

[0184] The R-β curves G10, G11, G12, G13, G14, G15, G16 are the relationships between the resistance value R and the correction coefficient β at T = 330 K, 320 K, 310 K, 300 K, 290 K, 280 K, 270 K, respectively.

[0185] The slopes of the R-β curves G10 to G16 are different from each other depending on the temperature T. Also, the R-β curves G10 to G16 basically intersect at a point P. As shown in Figure 11 (a), at the intersection point P, the resistance value R of the transistor TN basically does not depend on the temperature T.

[0186] At the intersection point P, the resistance value R of the transistor TN is "Rp", and the correction coefficient β is "β(Rp)". By setting the correction coefficient to β(Rp) in Equation (2), the temperature dependence of the transistor TN can be canceled by the correction voltage Vc containing the correction coefficient β(Rp). As a result, the temperature dependence of the resistance value R of the transistor TN can be reduced. In other words, with respect to the change in the temperature T, the resistance value R of the transistor TN can be maintained substantially constant. Hereinafter, the resistance value R that basically does not depend on the temperature is sometimes described as "independent of temperature" or "having no temperature dependence".

[0187] In order to obtain the correction coefficient β(Rp), it is sufficient to calculate at least two R-β curves. Also, as can be seen from Equations (15) and (16), when the reference voltage Vgs0 changes, the R-β curve also changes. Therefore, when the reference voltage Vgs0 changes, the position of the intersection point P changes. As a result, when the reference voltage Vgs0 changes, the correction coefficient β(Rp) also changes.

[0188] Among them, it is preferable to actually measure the resistance value R of the transistor TN of the resistance device 100 while changing the correction coefficient β at two or more temperatures T Figure 1 to obtain two or more R-β curves. Then, the correction coefficient β(Rp) at the intersection point P of the two or more R-β curves is obtained. Moreover, the correction coefficient β(Rp) is set as the correction coefficient β of the correction voltage generation unit 15 of the resistance device 100. In particular, since the R-β curve is actually measured, the correction coefficient β(Rp) suitable for actual use of the transistor TN can be determined. As a result, the temperature dependence of the resistance value R of the transistor TN can be further reduced.

[0189] In addition, the temperature T represents the ambient temperature of the resistance device 100. For example, the resistance device 100 is placed in a thermostat, and the temperature T is set by the thermostat.

[0190] As described above with reference to Figure 11 (a), in the first embodiment, based on the intersection point of the R-β curves G10 to G16, the correction coefficient β(Rp) when the resistance value R is independent of the temperature T can be obtained.

[0191] Next, referring to Figure 12 (a) and Figure 12(b) A further preferred example of the method for determining the correction coefficient β. Figure 12 (a) R-β curves G21 and G22 at two arbitrarily different temperatures T11 and T12 are respectively shown. The R-β curves G21 and G22 represent the resistance value R of the transistor TN with respect to the correction coefficient β. The horizontal axis represents the correction coefficient β, and the vertical axis represents the resistance value R.

[0192] First, in each of the temperatures T11 and T12, while changing the correction coefficient β, the resistance value R of the transistor TN of the resistance device 100 is measured, and two R-β curves G21 and G22 are obtained. Then, based on the intersection point P of the two R-β curves G21 and G22, the correction coefficient β(Rr) is obtained. The resistance value R when the correction coefficient β is the correction coefficient β(Rr) is the resistance value Rr without temperature dependence. Therefore, when the correction coefficient β(Rr) is set as the correction coefficient β of the correction voltage generation unit 15 of the resistance device 100, the temperature dependence of the resistance value R of the transistor TN can be effectively reduced by the correction voltage Vc. As a result, the resistance value R of the transistor TN can be maintained as the resistance value Rr.

[0193] However, the resistance value Rr without temperature dependence at the intersection point P is basically inconsistent with the required resistance value Rd (hereinafter, sometimes referred to as "target resistance value Rd"). The reason is that, as Figure 12 (a) shows, since the resistance value R of the transistor TN at any temperatures T11 and T12 changes according to the correction coefficient β, when obtaining the R-β curve at each temperature, what value the resistance value Rr at the intersection point P is is unknown.

[0194] In such a case, in order to make the resistance value Rr consistent with the target resistance value Rd, it is necessary to measure the resistance value R with respect to the correction coefficient β at two different temperatures while changing the reference voltage Vgs0, and repeatedly and exploratively obtain the intersection point of the R-β curve with respect to the reference voltage Vgs0 until the resistance value Rr at the intersection point is consistent with the target resistance value Rd.

[0195] Then, in order to more efficiently determine the correction coefficient β(Rd) corresponding to the target resistance value Rd, focus on Figure 11 the R-β curve G11 at the temperature T = 320K in (a). Figure 11 In the example of (a), the resistance value R of the transistor TN is Rp which is substantially constant and does not depend on the correction coefficient β as shown by the R-β curve G11 when the temperature T is 320K.

[0196] Therefore, when the temperature T is 320K, there is no correction effect on the correction voltage Vc with the correction coefficient β. In other words, when the temperature T is 320K, the correction voltage Vc is zero. In other words, the temperature T of 320K is the temperature when the correction voltage Vc is zero. Therefore, in Equation (2), the temperature T of 320K corresponds to the first temperature T1.

[0197] Figure 12 (b) is a relationship graph of the R-β curve G31 and the R-β curve G32. The R-β curve G31 represents the resistance value R of the transistor TN with respect to the correction coefficient β at the first temperature T1, and the R-β curve G32 represents the resistance value R of the transistor TN with respect to the correction coefficient β at the second temperature T2. The horizontal axis represents the correction coefficient β, and the vertical axis represents the resistance value R. The second temperature T2 is different from the first temperature T1.

[0198] As Figure 12 (b) shows, the R-β curve G31 is the resistance value R of the transistor TN at the first temperature T1. The first temperature T1 is the temperature when the physical quantity related to the transistor TN is basically constant with respect to the change of the correction coefficient β. Figure 12 In the example of (b), the first temperature T1 is the temperature when the resistance value R of the transistor TN is basically constant with respect to the change of the correction coefficient β. That is to say, the first temperature T1 is the temperature when the correction voltage Vc is zero. At the first temperature T1, the resistance value R is the resistance value Rr without temperature dependence.

[0199] Among them, at the temperature when the correction voltage Vc is zero, that is, the first temperature T1, while changing the reference voltage Vgs0, the resistance value R of the resistance device 100 is actually measured, and the reference voltage Vgs0(Rd) when the resistance value R is the target resistance value Rd is obtained.

[0200] Then, in the resistance device 100, the reference voltage Vgs0 is set to the reference voltage Vgs0(Rd) and the temperature T is set to the second temperature T2 different from the first temperature T1. By actually testing the resistance value R while changing the correction coefficient β, the R-β curve G32 is obtained. The R-β curve G32 is the resistance value R of the transistor TN at the second temperature T2.

[0201] Then, obtain the correction coefficient β(Rr) at the intersection point P of the R-β curve G31 at the first temperature T1 and the R-β curve G32 at the second temperature T2. The resistance value Rr without temperature dependence corresponding to the correction coefficient β(Rr) must be consistent with the target resistance value Rd. At this time, the correction coefficient β(Rr) at the intersection point P is also consistent with the correction coefficient β(Rd) with respect to the target resistance value Rd. Therefore, according to the first embodiment, when the correction coefficient β(Rd) is set as the correction coefficient β of the correction voltage generation unit 15 of the resistance device 100, the temperature dependence of the resistance value R of the transistor TN can be effectively reduced by the correction voltage Vc, and the resistance value R of the transistor TN can be maintained as the target resistance value Rd.

[0202] In particular, as described with reference to Figure 11 (a), due to the influence of non-linearity, several R-β curves G10 to G16 do not strictly intersect at one point. Among them, it is preferable to obtain the correction coefficient β(Rd) by setting the second temperature T2 to a value near the temperature at which the transistor TN is actually used. For example, in the case of using the transistor TN near the body temperature in an electronic device for detecting biological information, the second temperature T2 is set to a value near the human body temperature, that is, 310K, and the first temperature T1 is set to a value near the second temperature T2, that is, 320K.

[0203] In addition, for example, Figure 11 (a), at the intersection point of the R-β curve G11 corresponding to the first temperature T1 of 320K and the R-β curve G12 corresponding to the second temperature T2 of 310K, the correction coefficient β(Rp) is -0.00115. Then, in Equation (15), the correction coefficient β(Rp) is set to -0.00115, and the resistance value R of the transistor TN is calculated according to Equations (15) and (16).

[0204] Figure 11 (b) is a temperature correction effect chart of the temperature dependence of the resistance value R of the transistor TN. The curve 506 represents the case where temperature correction is not implemented, and the curve 507 represents the case where temperature correction is implemented. The horizontal axis represents the temperature T [K], and the vertical axis represents the resistance value R [MΩ]. In the case of not implementing temperature correction, the correction coefficient β in Equation (15) is zero, and the resistance value R is calculated according to Equation (16), and the curve 506 is plotted. It can be clearly seen from the curve 506 that the temperature dependence of the resistance value R is strong without performing correction.

[0205] On the other hand, in the case of implementing temperature correction, the correction coefficient β in Equation (15) is set according to Figure 11(a) The obtained correction coefficient β(Rp) is -0.00115. According to Equation (16), the resistance value R is calculated, and the curve 507 is plotted. It can be clearly seen from the curve 507 that the resistance value R is basically constant and is the value Rp. That is, through the correction based on the correction coefficient β(Rp), the temperature dependence of the resistance value R of the transistor TN is strongly suppressed.

[0206] As described above, according to the first embodiment, in the method for finding the combination of the target resistance value Rd without temperature dependence and the correction coefficient β(Rd), compared with the case Figure 12 (a) described, that is, the case of continuously and repeatedly searching while changing the reference voltage Vgs0 to find the intersection points of the R-β curves at any two different temperatures, Figure 12 (b) described, the combination can be determined at high speed, that is, uniquely determined by the intersection point P of the R-β curve at the first temperature T1 independent of the correction coefficient β and the R-β curve at the second temperature T2 different from the first temperature T1. In particular, in order to implement the method for determining the target resistance value Rd without temperature dependence and the correction coefficient β(Rd) described in Figure 12 (b), the voltage application circuit 1 having the correction voltage generation unit 15 is preferred, and the correction voltage generation unit 15 generates a correction voltage Vc that has no correction effect at the first temperature T1.

[0207] In addition, the target resistance value Rd is an example of the "target physical quantity related to the field effect transistor". That is, the target resistance value Rd can be measured from an electronic circuit including the field effect transistor TN, is the resistance value of the field effect transistor TN, and is the resistance value set as the target value.

[0208] Next, with reference to Figure 13 (a) to Figure 13 (c), a further preferred example of the method for determining the correction coefficient will be described. Figure 13 (a) is a graph showing the relationship between the reference voltage Vgs0 and the resistance value R of the transistor TN at the first temperature T1 (=320K). Figure 13 (b) is a graph showing the relationship between the correction coefficient β and the resistance value R of the transistor TN at the second temperature T2 (=310K). Figure 13 In (a), the horizontal axis represents the reference voltage Vgs0 [V]. Figure 13 In (b), the horizontal axis represents the correction coefficient β. Figure 13 (a) and Figure 13 (b), the vertical axis represents the resistance value R [MΩ]. The correction coefficient β is determined by the following (Step 1) and (Step 2).

[0209] (Step 1) As Figure 13As shown in (a), at the first temperature T1 where the correction voltage Vc is zero, determine the reference voltage Vgs0(Rd) when the resistance value R of the transistor TN of the resistance device 100 reaches the target resistance value Rd ( Figure 13 in the example of (a) is 40 MΩ). In addition, at the first temperature T1, since the correction voltage Vc is zero, the correction coefficient β can be any value, and the reference voltage Vgs0(Rd) is consistent with Figure 1 the control voltage Vgsa in Equation (11).

[0210] Specifically, first, place the resistance device 100 in a thermostat, and set the ambient temperature of the resistance device 100 to the first temperature T1 through the thermostat. Next, while changing the voltage value of the reference voltage Vgs0, measure the resistance value R. Next, determine the reference voltage Vgs0(Rd) when the resistance value R is the target resistance value Rd.

[0211] (Step 2) As shown in Figure 13 (b), at the second temperature T2 and the reference voltage Vgs0(Rd), determine the correction coefficient β(Rd) when the resistance value R of the transistor TN reaches the target resistance value Rd ( Figure 13 in the example of (b) is 40 MΩ).

[0212] Specifically, first, place the resistance device 100 in a thermostat, and set the ambient temperature of the resistance device 100 to the second temperature T2 through the thermostat. Next, set the reference voltage Vgs0 to the reference voltage Vgs0(Rd) determined in (Step 1). Next, while changing the value of the correction coefficient β, measure the resistance value R. Then, determine the correction coefficient β(Rd) when the resistance value R is the target resistance value Rd.

[0213] Performing (Step 2) is equivalent to obtaining the correction coefficient β(Rd) at the intersection point P of the R-β curve G31 at the first temperature T1 and the R-β curve G32 at the second temperature T2 in Figure 12 (b). The reason is as follows. That is, as shown in Figure 12 (b), at the second temperature T2, only at the intersection point P, the resistance value R is consistent with the target resistance value Rd. Therefore, the correction coefficient β when the resistance value R is the target resistance value Rd at the second temperature T2 must be consistent with the correction coefficient β(Rd) at the intersection point P. Also, at the intersection point P, the resistance value R is the resistance value Rr without temperature dependence, and the target resistance value Rd is consistent with the resistance value Rr without temperature dependence.

[0214] As described above with reference to Figure 13 (a) and Figure 13(b) According to the description of Embodiment 1, through (Step 1) and (Step 2), the correction coefficient β(Rd) for obtaining the target resistance value Rd is determined, and the target resistance value Rd is consistent with the resistance value Rr without temperature dependence. Therefore, it is not necessary to repeatedly find the intersection points of the R-β curves at any two different temperatures while changing the reference voltage Vgs0. As a result, the combination of the target resistance value Rd without temperature dependence and the correction coefficient β(Rd) can be determined uniquely and at high speed.

[0215] That is to say, in Embodiment 1, the value of the correction coefficient β(Rd) refers to the value obtained when the target resistance value Rd of the transistor TN is obtained at the second temperature T2 different from the first temperature T1 based on the reference voltage Vgs0(Rd) when the target resistance value Rd of the transistor TN is obtained at the first temperature T1.

[0216] In particular, in order to implement (Step 1) and (Step 2), the voltage application circuit 1 having the correction voltage generation unit 15 is preferable, and the correction voltage generation unit 15 generates a correction voltage Vc having no correction effect at the first temperature T1.

[0217] Among them, Figure 13 (a) and Figure 13 The resistance value R in (b) is based on the simulation results of Equation (15) and Equation (16). The simulation conditions for the standard circuit parameters are μn_T0·Cox = 170.0 μA / V 2 , W = 0.6 μm, L = 60.0 μm, αth = -1.7 mV / K, αμ = -1.5, Vth_T0 = 0.7 V, T0 = 300 K. Also, Figure 13 (a), T = T1 = 320 K. Moreover, Figure 13 (b), T = T2 = 310 K.

[0218] Figure 13 (a) and Figure 13 (b), Vgs0(Rd) = 0.8368 V, β(Rd) = -0.00130.

[0219] Figure 13 (c) is a temperature correction effect chart of the temperature dependence of the resistance value R of the transistor TN. The curve 508 represents the case where temperature correction is not performed, and the curve 509 represents the case where temperature correction is performed. The horizontal axis represents the temperature T [K], and the vertical axis represents the resistance value R [MΩ]. In the case where temperature correction is not performed, the correction coefficient β of Equation (15) is set to zero, the reference voltage Vgs0 is set to 0.8368 V of Vgs0(Rd) obtained in (Step 1), the resistance value R is calculated by Equation (16), and the curve 508 is plotted. It can be clearly seen from the curve 508 that the temperature dependence of the resistance value R is strong when correction is not performed.

[0220] On the other hand, in the case of performing temperature correction, the correction coefficient β in Equation (15) is set to -0.00130 of β(Rd) obtained in (Step 2), and the resistance value R is calculated by Equation (16), and the curve 509 is plotted. It can be clearly seen from the curve 509 that the resistance value R is substantially constant and is the value Rd. That is, by the correction based on the correction coefficient β(Rd), the temperature dependence of the resistance value R of the transistor TN is strongly suppressed.

[0221] As described above with reference to Figure 11 (b) and Figure 13 (c), in the saturation region (Vds > Vgs - Vth) with strong non-linearity in the "Vgs > Vth" region, it was confirmed that the temperature correction of the drain current Ids and the resistance value R (MOS resistance) can be appropriately performed based on the correction coefficient β(Rp) or the correction coefficient β(Rd).

[0222] That is, Figure 1 the voltage application circuit 1 in controls the resistance value R between the drain and source of the transistor TN in the saturation region by applying a control voltage Vgs (= Vgs0 + β(T - T1)) between the gate and source of the transistor TN. As a result, the temperature dependence of the resistance value R of the transistor TN can be strongly suppressed.

[0223] As described with reference to Figure 11 (b) and Figure 13 (c), even in the saturation region with strong non-linearity, it was confirmed that the temperature correction of the drain current Ids and the resistance value R (MOS resistance) can be appropriately performed. Therefore, in the linear region (Vds < Vgs - Vth) with high linearity in the "Vgs > Vth" region, the temperature correction of the drain current Ids and the resistance value R (MOS resistance) can be performed more accurately. In such a case as well, by the same steps as those described with reference to Figure 11 (a) to Figure 13 (c), the correction coefficient β(Rp) corresponding to the resistance value Rp without temperature dependence or the correction coefficient β(Rd) corresponding to the target resistance value Rd without temperature dependence is determined.

[0224] That is, Figure 1 the voltage application circuit 1 in is preferably configured to control the resistance value R between the drain and source of the transistor TN in the linear region by applying a control voltage Vgs (= Vgs0 + β(T - T1)) between the gate and source of the transistor TN.

[0225] As described above, Figure 1The voltage application circuit 1 in the embodiment controls the resistance value R between the drain and the source of the transistor TN in the “Vgs>Vth” region by applying a control voltage Vgs between the gate and the source of the transistor TN.

[0226] Among them, if the subthreshold region of the transistor TN is used, a higher resistance value R can be achieved without changing the size of the transistor TN (gate length L and gate width W). The subthreshold region refers to the operating region of the transistor TN when the gate-source voltage Vgs (gate-source voltage) is smaller than the threshold voltage Vth (Vgs < Vth). In the subthreshold region, for example, when the aspect ratio (W / L) of the transistor TN is 0.01, a resistance value R of several MΩ to several tens of TΩ can be achieved. The subthreshold region is equivalent to an example of the "second operating region of the field effect transistor".

[0227] The case of using the subthreshold region of the transistor TN can also be adopted Figure 1 Therefore, the voltage applying circuit 1 controls the resistance value R between the drain and the source of the transistor TN in the subthreshold region by applying a control voltage Vgs (= Vgs0 + β (T-T1)) between the gate and the source of the transistor TN.

[0228] In the subthreshold region, the drain current Ids increases exponentially with respect to the gate-source voltage Vgs. Therefore, the operating characteristics of the transistor TN as a MOS resistor are different from the saturation region and the linear region in the "Vgs>Vth" region.

[0229] Below, refer to Figure 14 (a)~ Figure 16 (b) explains why the control voltage Vgs can also be corrected by the linear function shown in equation (2) in the subthreshold region.

[0230] Figure 14 (a) is a semi-logarithmic graph of the Ids-Vgs characteristic of a normal NMOS transistor. The horizontal axis represents the gate-source voltage Vgs [V], and the vertical axis represents the drain current Ids [μA] in a logarithmic scale. Figure 14 As shown in (a), the region indicated by "Vgs<Vth" is the subthreshold region. In the subthreshold region, the logarithm of the drain current Ids is 10 Ids is proportional to the gate-source voltage Vgs.

[0231] Figure 14 (b) is a graph showing the Ids-Vds characteristics of a subthreshold region of a general NMOS transistor. The horizontal axis represents the drain-source voltage Vds [V], and the vertical axis represents the drain current Ids [fA]. Figure 14(b) shows the drain current Ids for Vgs = 0.20 V, 0.25 V, and 0.30 V.

[0232] The drain current Ids in the subthreshold region is represented by Equation (17). The Vt in Equation (17) is called the thermal voltage and is represented by Equation (18). In Equation (18), k is the Boltzmann constant and q is the elementary charge. Also, η in Equation (17) is represented by Equation (19). The Cd in Equation (19) is the depletion layer capacitance. Figure 14 (a) shows the simulation results of Equation (17). From Equation (17), it can be seen that the drain current Ids of the transistor TN increases exponentially with the increase in the gate-source voltage Vgs. Therefore, as Figure 14 (a) shows, in the subthreshold region, the logarithm of the drain current Ids is proportional to the gate-source voltage Vgs.

[0233]

Number 10

[0234]

[0235] Vt = kT / q…(18)

[0236] η = 1 + (Cd / Cox)…(19)

[0237] In the subthreshold region, the range where the Ids-Vds characteristic can be linearly approximated is Vds < several tens of mV, which is relatively narrow. For example, an NMOS transistor is used as a non-linear MOS resistor with a high resistance value. Specifically, according to the tangent equation of Equation (17) based on Vds = 0, the drain current Ids can be linearly approximated as in Equation (20). Figure 14 The straight line represented by the dashed line in (b) is the simulation result of the standard circuit parameters of Equation (20) when Vgs = 0.20 V, 0.25 V, and 0.30 V. Figure 14 In the example of (b), it can be seen that a linear approximation can be made in the range of Vds < several tens of mV. The smaller Vgs is, Figure 14 the smaller the slope of the straight line represented by the dashed line in (b) is. The Ids-Vds characteristic in the range of Vds < several tens of mV corresponds to the linear region in the “Vgs > Vth” region.

[0238]

Number 11

[0239]

[0240] On the other hand, when Vds becomes large, the drain current Ids of the transistor TN saturates to a certain value. Figure 14In the example of (b), it can be seen that in the range where Vds > 100 mV, the drain current Ids saturates to a certain value. The Ids-Vds characteristic in the range where Vds > 100 mV corresponds to the saturation region in the "Vgs > Vth" region.

[0241] That is to say, Equation (17) representing the drain current Ids of the transistor TN in the subthreshold region (Vgs < Vth) expresses the characteristics of both Equation (4) representing the drain current Ids of the transistor TN in the linear region and Equation (3) representing the drain current Ids of the transistor TN in the saturation region in the "Vgs > Vth" region with a single formula. Similarly in the subthreshold region, the resistance value R of the transistor TN is represented by Equation (21).

[0242] R = Vds / Ids…(21)

[0243] From Equation (17), it can be seen that the drain current Ids can be controlled exponentially by the gate-source voltage Vgs. Controlling the drain current Ids by Vgs has the same meaning as controlling the resistance value R of the NMOS transistor by Vgs. The reason is that, as shown in Equation (21), R = Vds / Ids.

[0244] Among them, the relationship between the gate-source voltage Vgs, the drain current Ids, and the resistance value R in the subthreshold region of a general NMOS transistor is shown in Figure 15 . Figure 15 is the simulation result using standard circuit parameters based on Equation (17). Figure 15 (a)~ Figure 15 The horizontal axis of (c) represents the gate-source voltage Vgs [V]. Figure 15 The vertical axis of (a) represents the drain current Ids [fA]. Figure 15 The vertical axis of (b) represents Figure 15 the drain current Ids [A] of a semi-logarithmic graph with the vertical axis of (a) on a logarithmic scale. Figure 15 (c) is Figure 15 the graph of the drain current Ids of (a) transformed into the resistance value R through Equation (21), and the vertical axis represents the resistance value R [TΩ]. At this time, Vds is 0.1 V.

[0245] As Figure 15 (a) shows, in the subthreshold region, the drain current Ids increases exponentially with respect to the gate-source voltage Vgs. Therefore, as Figure 15 (b) shows, in the subthreshold region, the logarithm log 10 Ids of the drain current is proportional to the gate-source voltage Vgs. Also, as Figure 15As shown in (c), in the subthreshold region, a high resistance value R (= Vds / Ids) of, for example, more than teraohm (T)Ω can be achieved.

[0246] On the other hand, in the subthreshold region, the threshold voltage Vth and the electron mobility μn of the NMOS transistor also have temperature dependencies. The electron mobility μn after considering the temperature dependency is represented by Equation (8). Further, based on Equation (8) and Equation (17), in the subthreshold region, the drain current Ids reflecting the temperature dependencies of both the electron mobility μn and the threshold voltage Vth is represented by Equation (22). The temperature dependency of the threshold voltage Vth is included in Vt (= q / kT) of Equation (22).

[0247]

Number 12

[0248]

[0249] Figure 16 (a) is a graph showing the temperature dependency of the drain current Ids in the subthreshold region of the NMOS transistor obtained from Equation (22). The horizontal axis represents temperature [K], and the vertical axis represents the drain current Ids [fA]. Figure 16 (b) is Figure 16 a semi-logarithmic graph with the vertical axis of (a) on a logarithmic scale. The horizontal axis represents temperature [K], and the vertical axis represents the drain current Ids [A] on a logarithmic scale. Figure 16 (a) and Figure 16 (b) represent simulation results using standard circuit parameters.

[0250] As Figure 16 shown in (a), the drain current Ids reflecting the temperature dependencies of both the electron mobility μn and the threshold voltage Vth increases exponentially as the temperature T rises. Therefore, as Figure 16 shown in (b), the logarithm log 10 Ids of the drain current is approximately proportional to the temperature T.

[0251] On the other hand, as Figure 15 shown in (a), the drain current Ids increases exponentially as the gate-source voltage Vgs increases. Therefore, as Figure 15 shown in (b), the logarithm log 10 Ids of the drain current is proportional to the gate-source voltage Vgs.

[0252] Among them, corresponding to the drain current Ids that increases exponentially as the temperature rises, by linearly reducing the gate-source voltage Vgs to exponentially decrease the drain current Ids, the temperature dependency of the drain current Ids can be suppressed. As a result, in the subthreshold region, the temperature dependency of the NMOS transistor as a MOS resistor can be corrected.

[0253] As described above, in the subthreshold region, the control voltage Vgs applied between the gate and source of the transistor TN can be corrected by the linear function (correction voltage Vc) shown in Equation (2). The same is true in this case. By the same steps as those described with reference to Figure 11 (a) to Figure 13 (c), the correction coefficient β corresponding to the resistance value Rp without temperature dependence or the correction coefficient β corresponding to the target resistance value Rd without temperature dependence is determined.

[0254] In addition, Figure 14 (a) to Figure 16 (b), the simulation conditions using standard circuit parameters are μn·Cox = 170.0 μA / V 2 , μn_T0·Cox = 170.0 μA / V 2 , W = 0.6 μm, L = 60.0 μm, η = 1 / 0.7, αμ = -1.5, η = 1 / 0.7, k = 1.381×10 -23 J / K, q = 1.602×10 -19 C, Vth = 0.7 V, T0 = 300 K. In addition, Figure 14 (a) to Figure 15 (c), T = 310 K.

[0255] Among them, after reflecting the correction coefficient β for correcting the temperature dependence of the transistor TN into Equation (22), in the subthreshold region, the drain current Ids of the transistor TN is represented by Equation (23).

[0256]

Equation 13

[0257]

[0258] Next, with reference to Figure 17 (a) to Figure 17 (c), an example of a method for determining the correction coefficient β in the subthreshold region will be described. Figure 17 (a) is a graph showing the relationship between the reference voltage Vgs0 and the resistance value R of the transistor TN at the first temperature T1 (= 320 K). Figure 17 (b) is a graph showing the relationship between the correction coefficient β and the resistance value R of the transistor TN at the second temperature T2 (= 310 K). Figure 17 In (a), the horizontal axis represents the voltage value [V]. Figure 17 In (b), the horizontal axis represents the correction coefficient β. Figure 17 (a) and Figure 17 (b), the vertical axis represents the resistance value R [TΩ]. The correction coefficient β is determined by the following (Step 1) and (Step 2).

[0259] (Step 1) As shown in Figure 17 (a), at the first temperature T1 where the correction voltage Vc is zero, while changing the voltage value of the reference voltage Vgs0, measure the resistance value R of the resistance device 100, and determine the reference voltage Vgs0(Rd) when the resistance value R is the target resistance value Rd ( Figure 17 in the example of (a) is 10 TΩ). Additionally, at the first temperature T1, since the correction voltage Vc is zero, the correction coefficient β can be any value.

[0260] (Step 2) As shown in Figure 17 (b), set the reference voltage Vgs0 in the resistance device 100 to the reference voltage Vgs0(Rd) determined in (Step 1). Next, at the second temperature T2, while changing the value of the correction coefficient β, measure the resistance value R, and determine the correction coefficient β(Rd) when the resistance value R is the target resistance value Rd ( Figure 17 in the example of (b) is 10 TΩ).

[0261] Among them, Figure 17 (a) and Figure 17 (b), the resistance value R is based on the simulation results of Equation (21) and Equation (23). The simulation conditions using standard circuit parameters are μn_T0·Cox = 170.0 μA / V 2 , W = 0.6 μm, L = 60.0 μm, αμ = -1.5, η = 1 / 0.7, k = 1.381×10 -23 J / K, q = 1.602×10 -19 C, Vth = 0.7 V, T0 = 300 K. Also, Figure 17 in (a), T = T1 = 320 K. Also, Figure 17 in (b), T = T2 = 310 K.

[0262] Figure 17 (a) and Figure 17 (b), Vgs0(Rd) = 0.2746 V, β(Rd) = -0.00138.

[0263] Figure 17(c) is a graph showing the temperature correction effect of the temperature dependence of the resistance value R of the transistor TN. Curve 510 represents the case where temperature correction is not performed, and curve 511 represents the case where temperature correction is performed. The horizontal axis represents the temperature T [K], and the vertical axis represents the resistance value R [TΩ]. In the case where temperature correction is not performed, the correction coefficient β in Equation (23) is set to zero, the reference voltage Vgs0 is set to 0.2746 V of Vgs0(Rd) obtained in (Step 1), the resistance value R is calculated by Equation (21), and curve 510 is plotted. It can be clearly seen from curve 510 that the temperature dependence of the resistance value R is strong when correction is not performed.

[0264] On the other hand, in the case where temperature correction is performed, the correction coefficient β in Equation (23) is set to -0.00138 of β(Rd) obtained in (Step 2), the resistance value R is calculated by Equation (23), and curve 511 is plotted. It can be clearly seen from curve 511 that the resistance value R is substantially constant and is the value Rd. That is, by the correction based on the correction coefficient β(Rd), the temperature dependence of the resistance value R of the transistor TN is strongly suppressed.

[0265] As described above with reference to Figure 17 (c), in the subthreshold region (Vgs < Vth), it is confirmed that the temperature correction of the drain current Ids and the resistance value R (MOS resistance) can be appropriately performed based on the correction coefficient β(Rd). That is, with respect to the gate-source voltage, temperature correction can be performed over the entire range of Vgs < Vth (subthreshold region) and Vgs > Vth ("Vgs > Vth" region). Moreover, the same applies to the drain-source voltage, and temperature correction can be performed over the entire range from the region where the drain current Ids of the transistor TN can be linearly approximated (linear region) to the region where it saturates to a constant value (saturation region).

[0266] Next, with reference to Figure 18 (a) and Figure 18 (b), an example of the measurement method of the resistance value R of the transistor TN when determining the reference voltage Vgs0 and the correction coefficient β (for example, Figure 11 (a), Figure 12 (b), Figure 13 (a), Figure 13 (b), Figure 17 (a), Figure 17 (b)) will be described. In addition, Figure 18 (a) and Figure 18 (b), the temperature detection unit 13 is omitted in order to simplify the drawings.

[0267] Figure 18 (a) is used to illustrate the first example of the measurement method of the resistance value R of the transistor TN in the first embodiment. AsFigure 1 As shown in (a), the resistance device 100 is installed in the electronic circuit device 200. The electronic circuit device 200 is, for example, an integrated circuit device.

[0268] The electronic circuit device 200 includes a resistance device 100, switches SW1 to SW4, an electronic circuit 3, a monitoring terminal Mt1, and a monitoring terminal Mt2. The transistor TN of the resistance device 100 is connected to the electronic circuit 3 through switches SW3 and SW4.

[0269] The switch SW1 and the switch SW3 are connected in series between the monitoring terminal Mt1 and the electronic circuit 3. At a node N1 between the switch SW1 and the switch SW3, one terminal (drain terminal) of the transistor TN of the resistance device 100 is connected. The switch SW2 and the switch SW4 are connected in series between the monitoring terminal Mt2 and the electronic circuit 3. At a node N2 between the switch SW2 and the switch SW4, the other terminal (source terminal) of the transistor TN of the resistance device 100 is connected.

[0270] When measuring the resistance value R of the transistor TN and determining the reference voltage Vgs0 and the correction coefficient β, the switches SW1 and SW2 connect the transistor TN to the monitoring terminals Mt1 and Mt2. Also, the switches SW3 and SW4 disconnect the transistor TN from the electronic circuit 3.

[0271] When the electronic circuit device 200 is operated alone, the determined reference voltage Vgs0 and the correction coefficient β are set in the control voltage application unit 9, and the switches SW1 and SW2 disconnect the transistor TN and the electronic circuit 3 from the monitoring terminals Mt1 and Mt2. Moreover, the switches SW3 and SW4 connect the transistor TN to the electronic circuit 3.

[0272] The measurement system SYS performs the measurement of the resistance value R of the transistor TN and the processing of the measurement data. The measurement system SYS includes a computer 300 and a measuring device 400.

[0273] The measuring device 400 is connected to the monitoring terminals Mt1 and Mt2. Then, the measuring device 400 applies a voltage Vds across the resistance generated by the transistor TN through the monitoring terminals Mt1 and Mt2, and measures the resistance value R of the transistor TN by measuring the current Ids flowing through the resistance.

[0274] The computer 300 sets the voltage value of the reference voltage Vgs0 in the reference voltage generation unit 11 of the control voltage application unit 9. Also, when the computer 300 searches for the reference voltage Vgs0(Rd) relative to the target resistance value Rd, it changes the voltage value of the reference voltage Vgs0 set in the reference voltage generation unit 11.

[0275] The computer 300 sets the value of the correction coefficient β for the correction voltage generation unit 15 of the control voltage application unit 9. Further, when the computer 300 searches for the correction coefficient β(Rd) with respect to the target resistance value Rd, the computer 300 changes the value of the correction coefficient β set in the correction voltage generation unit 15.

[0276] The computer 300 controls the measuring device 400. Further, the computer 300 obtains measurement data representing the resistance value R of the transistor TN from the measuring device 400. Moreover, the computer 300 determines the reference voltage Vgs0(Rd) and the correction coefficient β(Rd) with respect to the target resistance value Rd by processing the measurement data.

[0277] As described above with reference to ​ (a), in the first example of the measurement method, the transistor TN is connected to the monitoring terminals Mt1 and Mt2, and the resistance value R of the transistor TN is directly measured. Therefore, the resistance value R can be measured with high precision. In particular, the first example of the measurement method is effective when connecting the monitoring terminals Mt1 and Mt2 to the transistor TN and connecting the switches SW3 and SW4 to between the transistor TN and the electronic circuit 3 do not affect the characteristics of the transistor TN and the operation of the electronic circuit 3.

[0278] ​ (b) The second example of the measurement method for the resistance value R of the transistor TN in the first embodiment is described. Hereinafter, the points different from the first example in the second example are mainly described.

[0279] As ​ (b) shows, the resistance device 100 is mounted in the electronic circuit device 200A. The electronic circuit device 200A is, for example, an integrated circuit device.

[0280] The electronic circuit device 200A includes a resistance device 100, an electronic circuit 3, a monitoring terminal Mt1, a monitoring terminal Mt2, a voltage control voltage source 19x, and a transistor TND.

[0281] The structure of the transistor TND is the same as that of the transistor TN. The transistor TND is arranged close to the transistor TN. In the second example, the resistance value of the transistor TND having the same structure as the transistor TN (hereinafter, referred to as "resistance value Rx") is measured, and it is inferred that the resistance value Rx is the resistance value R of the transistor TN.

[0282] One terminal (drain terminal) of the transistor TND is connected to the monitoring terminal Mt1, and the other terminal (source terminal) is connected to the monitoring terminal Mt2. On the other hand, the transistor TN is connected to the electronic circuit 3.

[0283] The structure of the voltage-controlled voltage source 19x is the same as that of the voltage-controlled voltage source 19. The voltage-controlled voltage source 19x is connected to the control voltage application unit 9. Therefore, based on the control voltage Vgsa, the voltage-controlled voltage source 19x generates a control voltage (hereinafter referred to as "measurement control voltage Vgsx") having the same voltage value as the control voltage Vgs generated by the voltage-controlled voltage source 19, and applies the measurement control voltage Vgsx between the gate and source of the transistor TND.

[0284] The measuring device 400 applies a voltage Vds across the resistance generated by the transistor TND through the monitoring terminals Mt1 and Mt2, and measures the drain current flowing through the resistance to measure the resistance value Rx of the transistor TND. Then, the computer 300 obtains measurement data representing the resistance value Rx of the transistor TND from the measuring device 400. Moreover, the computer 300 determines the reference voltage Vgs0(Rd) and the correction coefficient β(Rd) with respect to the target resistance value Rd by processing the measurement data. That is, the computer 300 infers that the resistance value Rx of the transistor TND is the resistance value R of the transistor TN, and thereby determines the reference voltage Vgs0(Rd) and the correction coefficient β(Rd).

[0285] As described above with reference to ​ (b), in the second example of the measurement method, the resistance value R of the transistor TN is indirectly measured by measuring the resistance value Rx of the transistor TND having the same structure as the transistor TN connected to the electronic circuit 3. Therefore, it is possible to prevent the monitoring terminals Mt1 and Mt2 from affecting the characteristics of the transistor TN and the operation of the electronic circuit 3.

[0286] In particular, in the second example of the measurement method, it is preferable to place the transistor TND as close as possible to the transistor TN. The reason is that it is possible to suppress the characteristic deviation between the transistor TND and the transistor TN, and further improve the consistency between the resistance value Rx and the resistance value R.

[0287] Among them, ​ (a) and ​ (b), the structure of the electronic circuit 3 is not particularly limited as long as the transistor TN is connected to the electronic circuit 3. The electronic circuit 3 may include, for example, at least one of a transistor, a diode, a capacitor, an inductor, and a resistor.

[0288] So far, for the purpose of temperature correction of the resistance value R of the transistor TN of the resistance device 100, for example, the method of measuring the resistance value R of the target transistor TN has been described. However, temperature correction of the resistance value R of the transistor TN can also be performed using physical quantities other than the resistance value R. In other words, by using a physical quantity (hereinafter referred to as "physical quantity G") that can be measured from an electronic circuit including the resistance device 100 for which temperature correction is performed, the reference voltage Vgs0 and the correction coefficient β for temperature correction can be determined. Hereinafter, the physical quantity G that can be measured from the electronic circuit is sometimes referred to as the "physical quantity G of the electronic circuit".

[0289] Refer to ​ (a) to describe determining the reference voltage Vgs0 and the correction coefficient β and performing temperature correction using the physical quantity G that can be measured from an electronic circuit including the resistance device 100. ​ (a) is an electronic circuit device 200B including the resistance device 100. In addition, ​ in (a), the temperature detection unit 13 is omitted for simplifying the drawing.

[0290] As ​ shown in (a), the electronic circuit device 200B includes an electronic circuit 3B and a resistance device 100. The transistor TN (MOS resistor MR) of the resistance device 100 is built in the electronic circuit 3B as a circuit element of the electronic circuit 3B. The input terminal In and the output terminal Out of the electronic circuit 3B are connected to the terminal Mt1 and the terminal Mt2 of the electronic circuit device 200B, respectively. Among them, ​ in (a), one terminal each of the input terminal In and the output terminal Out of the electronic circuit 3B and the terminal Mt1 and the terminal Mt2 of the electronic circuit device 200B are illustrated, but the number of terminals is not particularly limited, and the number of terminals required to measure the physical quantity G of the electronic circuit 3B can be used to connect the measuring device 400 to the electronic circuit 3B.

[0291] The measurement system SYS includes a computer 300 and a measuring device 400. The measuring device 400 measures the physical quantity G of the electronic circuit 3B by being connected to the input terminal In and the output terminal Out of the electronic circuit 3B through the terminal Mt1 and the terminal Mt2. The computer 300 controls the measuring device 400 and obtains the measurement data from the measuring device 400.

[0292] Regarding the physical quantity G that can be measured by the measuring instrument 400 from the electronic circuit 3B including the resistance device 100 of the transistor TN (MOS resistor MR), it can be generally expressed as a function of the resistance value R of the transistor TN as shown in Equation (24). Specifically, the resistance value R can be generally expressed as a function of the temperature T, the reference voltage Vgs0, and the correction coefficient β. Therefore, the physical quantity G, which is a function of the resistance value R, can also be generally expressed as a function of the temperature T, the reference voltage Vgs0, and the correction coefficient β.

[0293] G = G(R) = G(R(T, Vgs0, β))

[0294] …(24)

[0295] By implementing the temperature correction of the resistance value R of the transistor TN in the resistance device 100, the temperature dependence of the physical quantity G, which is a function of the resistance value R of the transistor TN shown in Equation (24), is also corrected. For example, in the electronic circuit 3B, when the temperature dependence of the transistor TN of the resistance device 100 is dominant and the temperature dependence of other circuit elements constituting the electronic circuit 3B can be completely ignored, if the temperature correction of the resistance value R of the transistor TN is appropriately implemented, it is possible to effectively correct the temperature dependence of the physical quantity G, which is a function of the resistance value R of the transistor TN shown in Equation (24).

[0296] At this time, the relationship between the physical quantity G that can be measured from the electronic circuit 3B and the correction coefficient β of the resistance value R of the transistor TN of the resistance device 100 included in the electronic circuit 3B is shown in a general graph in ​ (b). The curve G91 represents the physical quantity G with respect to the correction coefficient β at the first temperature T1 when the correction voltage Vc of the transistor TN is zero, that is, the G-β curve. Among them, two second temperatures T2 different from the first temperature T1 are introduced. One of the two second temperatures T2 is denoted as the second temperature T21, and the other of the two second temperatures T2 is denoted as the second temperature T22. The second temperature T21 is different from the second temperature T22. The curves G92 and G93 represent the G-β curves of the second temperature T21 and the second temperature T22, respectively. Thus, the G-β curve represents the relationship between the physical quantity G and the correction coefficient β.

[0297] In the electronic circuit 3B, for example, when the circuit element with temperature dependence is only the transistor TN of the resistance device 100, after appropriately correcting the temperature dependence of the resistance value R of the transistor TN, the temperature dependence of the physical quantity G of the electronic circuit 3B is eliminated. That is to say, after appropriately correcting the temperature dependence of the resistance value R of the transistor TN, ​In (b), not only is the physical quantity G at the first temperature T1 where the correction voltage Vc is zero the same as the physical quantity G at the second temperature T21, but the physical quantity G at the first temperature T1 where the correction voltage Vc is zero is also the same as the physical quantity G at the second temperature T22. In other words, the intersection point of the curve G91 at the first temperature T1 and the curve G92 at the second temperature T21 coincides with the intersection point of the curve G91 at the first temperature T1 and the curve G93 at the second temperature T22 at the intersection point P. In other words, the intersection point of the curve G91 at the first temperature T1 and any G-β curve at any temperature different from the first temperature T1 coincides at one point, that is, at the intersection point P. In other words, the curve G91 at the first temperature T1 coincides with two or more G-β curves corresponding to two or more arbitrary temperatures different from the first temperature T1 at one point, that is, at the intersection point P. Therefore, based on the intersection point P of the curve G91 at the first temperature T1 where the correction voltage Vc is zero and the G-β curve (for example, curve G92) at a temperature (for example, the second temperature T21) different from the first temperature T1, the physical quantity Gd independent of temperature and the correction coefficient β(Gd) are uniquely determined.

[0298] Also, as ​ shown in (b), at the first temperature T1, the correction voltage Vc is zero. Therefore, the resistance value R of the transistor TN is independent of the correction coefficient β and is basically constant, and the physical quantity G of the electronic circuit 3B shown in Equation (24) is also basically constant. Therefore, by finding the reference voltage Vgs0(Gd) that becomes the target physical quantity Gd at the first temperature T1 and setting the reference voltage Vgs0 of the resistance device 100 to Vgs0(Gd), and by finding the correction coefficient β(Gd) for the target physical quantity Gd at a second temperature T2 different from the first temperature T1, the reference voltage Vgs0 and the correction coefficient β can be uniquely determined.

[0299] That is to say, referring to ​ and ​ the method for determining the reference voltage Vgs0 and the correction coefficient β described, by replacing the "resistance value R" of the transistor TN of the resistance device 100 with the "physical quantity G" of the electronic circuit 3B including the resistance device 100 and replacing the "target resistance value Rd" with the "target physical quantity Gd", in the case of the "physical quantity G", the reference voltage Vgs0 and the correction coefficient β can also be determined by the same steps as in the case of the "resistance value R". In addition, referring to ​ and ​ the method for determining the reference voltage Vgs0 and the correction coefficient β described, replace the "resistance value" with the "physical quantity" and the "R-β curve" with the "G-β curve".

[0300] In the case where other circuit elements constituting the electronic circuit 3B, which are different from the transistor TN of the resistance device 100, have temperature dependence​ (b), the curves G91, G92, and G93 do not intersect at a single point P. That is, there are multiple intersection points. For example, there are intersection points between curve G91 and curve G92, between curve G91 and curve G93, and between curve G92 and curve G93. In other words, there are multiple intersection points of the G-β curves at any two different temperatures. However, when the temperature dependence of other circuit elements is sufficiently small compared to the temperature dependence of the transistor TN of the resistance device 100, the multiple intersection points of the G-β curves at any two different temperatures take similar values. Therefore, for example, by finally determining the average value or the median value of the multiple correction coefficients β determined according to the multiple intersection points as the correction coefficient β, the temperature dependence of the physical quantity G of the electronic circuit 3B can be suppressed by appropriately determining the correction coefficient β.

[0301] In addition, the value of the correction coefficient β(Gd) refers to the value obtained when the target physical quantity Gd related to the transistor TN is obtained at the second temperature T2 different from the first temperature T1 based on the reference voltage Vgs0(Gd) when the target physical quantity Gd related to the transistor TN is obtained at the first temperature T1.

[0302] Among them, the use of the RC filter circuit is described ​ A specific example of the temperature correction method of the electronic circuit 3B including the transistor TN of the resistance device 100 shown in (a). ​ (a) is a circuit diagram of the RC integrator 110A in the first embodiment. ​ (b) is a circuit diagram of the RC differentiator 110B in the first embodiment. The RC integrator 110A and the RC differentiator 110B are each an example of an RC filter circuit. Hereinafter, the RC integrator 110A and the RC differentiator 110B are sometimes collectively referred to as the "RC filter circuit 110X".

[0303] As ​ (a) and ​ As shown in (b), the RC filter circuit 110X includes a resistor R with a resistance value R and a capacitor C with a capacitance value C, and is a basic electronic circuit. The RC filter circuit 110X has an input terminal In and an output terminal Out. In addition, the voltage input to the input terminal In is sometimes referred to as the "input voltage In", and the voltage output from the output terminal Out is sometimes referred to as the "output voltage Out".

[0304] ​ A specific circuit example of the electronic circuit 3B in (a) is ​ the RC integrator 110A in (a) or ​(b)'s RC differentiator filter 110B. In other words, the input terminals In and output terminals Out of the RC integrator filter 110A and the RC differentiator filter 110B respectively correspond to ​ (a)'s input terminal In and output terminal Out of the electronic circuit 3B. Also, the resistance elements R of the RC integrator filter 110A and the RC differentiator filter 110B respectively correspond to ​ (a)'s transistor TN (MOS resistor MR) of the resistance device 100.

[0305] Regarding ​ (a) and ​ (b)'s physical quantity representing the basic characteristics of the RC filter circuit 110X has a cut-off frequency fc [Hz]. ​ In the RC integrator filter 110A shown in (a), when the frequency of the sine wave voltage signal applied to the input terminal In continuously increases, after the frequency exceeds the cut-off frequency fc, the circuit gain (= amplitude of the output voltage Out / amplitude of the input voltage In) decreases. Theoretically, compared with the case of inputting a sine wave with a frequency much lower than the cut-off frequency fc, the circuit gain at the cut-off frequency fc decreases by 3 dB, which is called the high cut-off frequency. ​ In the RC differentiator filter 110B shown in (b), when the frequency of the sine wave voltage signal applied to the input terminal In continuously decreases, after the frequency is lower than the cut-off frequency fc, the circuit gain decreases. Theoretically, compared with the case of inputting a sine wave with a frequency much higher than the cut-off frequency fc, the circuit gain at the cut-off frequency fc decreases by 3 dB, which is called the low cut-off frequency.

[0306] The cut-off frequency fc is common to the RC integrator filter 110A and the RC differentiator filter 110B, and is expressed by the resistance value R and the capacitance value C as shown in Equation (25). In the case where the MOS resistor generated by the transistor TN constitutes the resistance element R of the RC filter circuit 110X and the ordinary parallel plate capacitor constitutes the capacitor C of the RC filter circuit 110X, the temperature dependence of the capacitor C is sufficiently low compared with the MOS resistor. Therefore, regarding the temperature dependence, the cut-off frequency fc can be expressed as a function of the resistance value R with high temperature dependence. In other words, as can be seen from Equation (25), the cut-off frequency fc is expressed by the reciprocal of the resistance value R (= 1 / R), and 1 / R is multiplied by a constant 1 / (2πC) with sufficiently low temperature dependence.

[0307] fc = 1 / (2πRC) = fc(R(T, Vgs0, β))…(25)

[0308] The formula (25) representing the cut-off frequency fc of the RC filter circuit 110X is, similarly to the formula (24) obtained by generalizing the physical quantity G that can be measured from the electronic circuit 3B, expressed as a function of the temperature T, the reference voltage Vgs0, and the correction coefficient β. Therefore, referring to formula (24) and ​ the physical quantity G that can be measured from the electronic circuit 3B in the description of (a) is the same as the cut-off frequency fc of the RC filter circuit 110X. That is to say, the cut-off frequency fc is an example of the physical quantity G that can be measured from the electronic circuit 3B. Therefore, by measuring the cut-off frequency fc, the reference voltage Vgs0 and the correction coefficient β at the required cut-off frequency fd (hereinafter, sometimes referred to as "target cut-off frequency fd") can be determined. The target cut-off frequency fd corresponds to an example of the "field-effect transistor-related target physical quantity". That is to say, the target cut-off frequency fd can be measured from the electronic circuit (RC filter circuit 110X) including the field-effect transistor TN, is a physical quantity including the resistance value of the field-effect transistor TN, and is the cut-off frequency fc set as the target value.

[0309] Also, according to formula (25), when the physical quantity on the vertical axis of the ​ and ​ charts used in the description of the temperature correction step is transformed into the cut-off frequency fc, it becomes a diagram for explaining the temperature correction step by the cut-off frequency fc of the RC filter circuit 110X. Therefore, similarly to the steps described with reference to the ​ and ​ charts, it can be known that the reference voltage Vgs0 and the correction coefficient β at the target cut-off frequency fd can be determined.

[0310] For example, ​ (a), ​ (a) and, ​ (b) in the charts, after the reciprocal (1 / R) of the resistance value R becomes the vertical axis, although the shape of the curve changes, the value of the correction coefficient β(Rp) or β(Rr) at the intersection of the two curves at two different temperatures remains unchanged. Therefore, in such a case, the value of 1 / R on the vertical axis at the intersection of the charts is 1 / Rp or 1 / Rr. Also, in such a case, by multiplying 1 / R on the vertical axis of the chart by the constant 1 / (2πC), even if the physical quantity on the vertical axis of the chart is transformed into the cut-off frequency fc, the value of the correction coefficient β(Rp) or β(Rr) at the intersection of the two curves at two different temperatures remains unchanged. Therefore, in such a case, the value of the cut-off frequency fc on the vertical axis at the intersection of the charts is 1 / (2πC×Rp) or 1 / (2πC×Rr). Therefore, 1 / (2πC×Rp) or 1 / (2πC×Rr) is the cut-off frequency without temperature dependence.

[0311] Therefore, the same applies to the cut-off frequency fc expressed as the reciprocal of the resistance value R, and the reference voltage Vgs0(fd) and the correction coefficient β(fd) at the target cut-off frequency fd without temperature dependence can be determined by the same steps as those described ​ and ​ For the steps described.

[0312] Furthermore, for example, ​ (a), ​ (b), ​ (a) and ​ (b) in the graphs, by transforming the physical quantity on the vertical axis into the cut-off frequency fc using Equation (25), the reference voltage Vgs0(fd) and the correction coefficient β(fd) at the target cut-off frequency fd without temperature dependence can be determined by the steps described with reference to ​ (a), ​ (b), ​ (a) and ​ (b).

[0313] As described above, in the case of determining the reference voltage Vgs0(fd) and the correction coefficient β(fd) at the target cut-off frequency fd, in the method for determining the reference voltage Vgs0 and the correction coefficient β described with reference to ​ and ​ if the "resistance value R" of the transistor TN of the resistance device 100 is replaced with the "cut-off frequency fd" of the electronic circuit 3B including the transistor TN, and the "target resistance value Rd" is replaced with the "target cut-off frequency fd", even in the case of the "cut-off frequency fd", the reference voltage Vgs0 and the correction coefficient β can be determined by the same steps as in the case of the "resistance value R". Additionally, in the method for determining the reference voltage Vgs0 and the correction coefficient β described with reference to ​ and ​ the "resistance value" is replaced with the "cut-off frequency", and the "R-β curve" is replaced with the "fc-β curve". The fc-β curve represents the relationship between the cut-off frequency fc and the correction coefficient β.

[0314] Using ​ (a), an example of the method for measuring the cut-off frequency fc of the RC filter circuit 110X will be described. The measuring device 400 of the measuring system SYS applies a sinusoidal voltage signal to the electronic circuit 3B through the terminal Mt1, that is, to ​ (a) or ​A sinusoidal voltage signal is applied to the input terminal In of the RC filter circuit 110X in (b). At this time, the measuring device 400 of the measurement system SYS measures the electronic circuit 3B through the terminal Mt2, that is, measures the output waveform from the output terminal Out of the RC filter circuit 110X. The measuring device 400 repeats such measurements while changing the frequency of the sinusoidal wave, and measures (calculates) the cut-off frequency fc based on the relationship between the gain of the RC filter circuit 110X (=amplitude of the output voltage Out / amplitude of the input voltage In) and the input frequency.

[0315] The computer 300 changes the voltage value of the reference voltage Vgs0 set in the reference voltage generation unit 11 according to the measurement purpose of the cut-off frequency fc. Also, the computer 300 changes the value of the correction coefficient β set in the correction voltage generation unit 15 according to the measurement purpose of the cut-off frequency fc. Also, the computer 300 obtains measurement data from the measuring device 400 for calculating the cut-off frequency fc of the electronic circuit 3B, that is ​ (a) or ​ the cut-off frequency fc of the RC filter circuit 110X shown in (b). Then, the computer 300 processes the measurement data to determine the reference voltage Vgs0 and the correction coefficient β.

[0316] Among them, as ​ a specific example of the temperature correction method of the electronic circuit 3B of the transistor TN including the resistance device 100 shown in (a), the most basic filter circuit, that is ​ the RC filter circuit 110X has been described, but the structure of the filter circuit is not particularly limited. For example, in an active filter circuit 110C including an operational amplifier 90 as ​ shown, temperature correction can also be performed.

[0317] ​ is an example of the active filter circuit 110C in the first embodiment. As ​ shown, the active filter circuit 110C includes an operational amplifier 90, input capacitors Ci1 and Ci2, feedback capacitors Cf1 and Cf2, and feedback resistors Rf1 and Rf2. The feedback resistors Rf1 and Rf2 are each composed of the transistor TN (MOS resistor MR) of the resistance device 100. In this way, ​ the electronic circuit 3B shown in (a) may also include two or more transistors TN (MOS resistors MR).

[0318] ​ In the example of, it is a two-input, two-output fully differential structure, and the input terminals In1 and In2 correspond to ​ the input terminal In of the electronic circuit 3B in (a), and the output terminals Out1 and Out2 correspond to ​The output terminal Out of the electronic circuit 3B in (a).

[0319] The operational amplifier 90 has a fully differential structure and inputs differential voltage signals In1 and In2 through input capacitors Cin1 and Cin2, and outputs differential voltage signals Out1 and Out2.

[0320] A first feedback circuit 91 composed of a feedback capacitor Cf1 and a feedback resistor Rf1 is connected in parallel between the inverting input terminal and the positive output terminal of the operational amplifier 90.

[0321] A second feedback circuit 92 composed of a feedback capacitor Cf2 and a feedback resistor Rf2 is connected in parallel between the non-inverting input terminal and the negative output terminal of the operational amplifier 90.

[0322] Generally, in ​ the case of the active filter circuit 110C with such a fully differential structure, the circuit parameters on the input terminal In1 and output terminal Out1 sides and the circuit parameters on the input terminal In2 and output terminal Out2 sides are made to have the same numerical values and characteristics as much as possible. Therefore, generally, the capacitance values of the input capacitors Cin1 and Cin2 are set to be equal. Similarly, generally, the capacitance values of the feedback capacitors Cf1 and Cf2 are set to be equal, and the resistance values of the feedback resistors Rf1 and Rf2 are set to be equal.

[0323] Figure 21 The active filter circuit 110C shown operates as a differential filter similar to the RC differential filter 110B shown in Figure 20 (b). Then, the cut-off frequency fc of the active filter circuit 110C is determined by the capacitance values of the feedback capacitors Cf1 and Cf2 and the resistance values of the feedback resistors Rf1 and Rf2. If the capacitance values of the feedback capacitors Cf1 and Cf2 are C and the resistance values of the feedback resistors Rf1 and Rf2 are R, the cut-off frequency fc is expressed by Equation (26). Equation (26) is the same as the equation expressing the cut-off frequency fc of the RC filter circuit 110X in Figure 20 (a) and Figure 20 (b).

[0324] fc = 1 / (2πRC)…(26)

[0325] Therefore, Equation (26) representing the cut-off frequency of the active filter circuit 110C, similar to Equation (24) that generalizes the physical quantity G that can be measured from the electronic circuit 3B, is expressed as a function of the temperature T, the reference voltage Vgs0, and the correction coefficient β. Therefore, the cut-off frequency fc of the active filter circuit 110C, referring to Equation (24) and Figure 19(a) The physical quantity G that can be measured from the electronic circuit 3B is the same. As a result, by measuring the cut-off frequency fc, the reference voltage Vgs0 and the correction coefficient β at the target cut-off frequency fd can be determined.

[0326] It can also be understood from the above that as Figure 19 an example of (a), the structure of the electronic circuit 3B is used as Figure 20 (a) or Figure 20 the RC filter circuit 110X of (b) or Figure 21 the active filter circuit 110C. Taking the cut-off frequency fc as the physical quantity to be measured, in such a case, it is also possible to perform temperature correction on the transistor TN of the resistance device 100 included in the RC filter circuit 110X or the active filter circuit 110C.

[0327] More generally, in the case where a physical quantity including the resistance value R of the transistor TN including the resistance device 100 can be measured from the electronic circuit 3B, it can be understood that temperature correction of the transistor TN included in the electronic circuit 3B can be performed. In such a case, the resistance value R of the transistor TN of the resistance device 100 can also be measured indirectly or directly as in Figure 18 (a) and Figure 18 (b).

[0328] Among them, Figure 19 in (a) (in this specification), the physical quantity related to the transistor TN, that is, the physical quantity that can be measured from the electronic circuit 3B including the transistor TN and includes the resistance value R of the transistor TN is not limited to the resistance value R and the cut-off frequency fc. For example, the physical quantity related to the transistor TN is the resistance value, current value, voltage value, frequency, gain, phase, sound, light, pressure, or energy of the circuit elements or circuits included in the electronic circuit 3B. For example, the physical quantity related to the transistor TN is a physical quantity obtained by combining two or more of the resistance value, current value, voltage value, frequency, gain, phase, sound, light, pressure, and energy of the circuit elements or circuits included in the electronic circuit 3B. For example, the physical quantity related to the transistor TN is a frequency distribution, spatial distribution, or time distribution of one or more of the resistance value, current value, voltage value, frequency, gain, phase, sound, light, pressure, and energy of the circuit elements or circuits included in the electronic circuit 3B.

[0329] For example, in the case where the electronic circuit 3B has a current source including a transistor, the "current value" as a physical quantity is the current value of the drain current Ids of the transistor TN. For example, in the case where the electronic circuit 3B has a current source including a transistor TN and an operational amplifier operating on the current source, the "physical quantity obtained by the combination of the voltage value and the current value" is the slew rate of the operational amplifier. The slew rate is expressed as C×(Vout / Iout). C is a constant, Vout represents the output voltage of the operational amplifier, and Iout represents the output current of the operational amplifier.

[0330] For example, in the case where the electronic circuit 3B has a current source including a transistor TN and an oscillator operating on the current source, the "frequency" as a physical quantity is the emission frequency of the oscillator. For example, in the case where the electronic circuit 3B has a current source including a transistor TN, an oscillator, and an electromagnetic buzzer connected to the oscillator, the "sound" as a physical quantity is the pitch of the sound output by the electromagnetic buzzer according to the emission frequency of the oscillator.

[0331] For example, in the case where the electronic circuit 3B has an operational amplifier and a phase compensation circuit for performing phase compensation of the operational amplifier, the "gain frequency distribution" as a physical quantity is the frequency characteristic of the gain of the operational amplifier. The phase compensation circuit includes, for example, a transistor TN and a capacitor. For example, in the case where the electronic circuit 3B has an operational amplifier and a phase compensation circuit, the "frequency distribution of the phase" as a physical quantity is the frequency characteristic of the phase of the operational amplifier.

[0332] For example, in the case where the electronic circuit 3B has a current source including a transistor TN and a light-emitting element (e.g., a light-emitting diode) operating on the current source, the "light" as a physical quantity is the light emission amount of the light-emitting element. The light emission amount is represented, for example, by luminance (intensity of light) (Cd), illuminance (lux), or irradiance (W / m 2 )

[0333] For example, in the case where the electronic circuit 3B has a resistor network, the "spatial distribution of voltage values" as a physical quantity is the output voltage distribution from multiple output terminals of the resistor network. Specifically, the resistor network includes multiple input terminals that respectively input different input voltages and multiple output terminals that respectively output different output voltages. Further, the resistor network has multiple first MOS resistors and multiple second MOS resistors. One terminal of the multiple first MOS resistors is the multiple input terminals, and the other terminal of the multiple first MOS resistors is the multiple output terminals. On the other hand, the multiple second MOS resistors are connected in series. Further, at the node between adjacent second MOS resistors, the output terminal that is the other terminal of the first MOS resistor is connected. At least one of the first MOS resistor and the second MOS resistor is constituted by a transistor TN. Additionally, for example, a spatial filter such as an image filter can be constituted by the resistor network. In such a case, the distribution of the output voltages from the multiple output terminals of the resistor network represents the spatial filtering characteristics.

[0334] Next, with reference to Figure 22 、 Figure 23 (a) and Figure 23 (b), the method for determining the correction coefficient in the first embodiment of the present invention will be described. The method for determining the correction coefficient determines the correction coefficient β, which is the correction coefficient when correcting the control voltage Vgs applied between the gate and source of the field effect transistor TN serving as a MOS resistor. The method for determining the correction coefficient is executed, for example, by the measurement system SYS ( Figure 18 (a) to Figure 19 (a)).

[0335] Further, in the method for determining the correction coefficient, in Equation (27), the control voltage Vgs is represented by "Vgs", the reference voltage Vgs0 is represented by "Vgs0", the correction voltage Vc is represented by "Vc", the correction coefficient β is represented by "β", the temperature T as a variable is represented by "T", and the first temperature T1 as the temperature when the correction voltage Vc is zero is represented by "T1". That is, Equation (27) is the same as Equation (1).

[0336] Vgs = Vgs0 + Vc = Vgs0 + β(T - T1)…(27)

[0337] Figure 22 is the flowchart of the method for determining the correction coefficient in the first embodiment. As Figure 22 shown, the method for determining the correction coefficient includes Step S1 to Step S4.

[0338] In Step S1, the measurement system SYS determines whether an instruction to change the first temperature T1 has been received.

[0339] In the case where a negative determination is made in process S1 (No), the process proceeds to process S3.

[0340] On the other hand, in the case where an affirmative determination is made in process S1 (Yes). The process proceeds to process S2.

[0341] In process S2, the measurement system SYS controls the temperature detection unit 13 and changes the first temperature T1 by changing at least one of the current value of the first current Ip and the current value of the second current Im. Then, the process proceeds to process S1.

[0342] In process S3, the measurement system SYS determines the voltage value of the reference voltage Vgs0, that is, the specific voltage value X, when the target physical quantity Gd related to the field effect transistor TN is obtained at the first temperature T1. The target physical quantity Gd represents, for example, the target resistance value Rd of the field effect transistor TN or the target cut-off frequency fd of a filter circuit (for example, the RC filter circuit 110X or the active filter circuit 110C) including the field effect transistor TN. The reference voltage Vgs0 having the specific voltage value X is the reference voltage Vgs0(Gd). The reference voltage Vgs0(Gd) is, for example, the reference voltage Vgs0(Rd) or the reference voltage Vgs0(fd).

[0343] In process S4, the measurement system SYS determines the value of the correction coefficient β, that is, the specific coefficient value W, when the target physical quantity Gd is obtained at the second temperature T2 different from the first temperature T1 and the specific voltage value X of the reference voltage Vgs0. The correction coefficient β having the specific coefficient value W is the correction coefficient β(Gd). The correction coefficient β(Gd) is, for example, the correction coefficient β(Rd) or the correction coefficient β(fd).

[0344] As described above with reference to Figure 22 As described above, according to the correction coefficient determination method of the first embodiment, through process S3, the specific voltage value X of the reference voltage Vgs0 corresponding to the target physical quantity Gd is determined at the first temperature T1 where the correction voltage Vc is zero, and then through process S4, the specific coefficient value W of the correction coefficient β corresponding to the target physical quantity Gd is determined in the state where the reference voltage Vgs0 has the specific voltage value X at the second temperature T2. Therefore, it is not necessary to repeatedly obtain multiple G-β curves (for example, multiple R-β curves or multiple fc-β curves) for each different reference voltage Vgs0. As a result, it is possible to determine the combination of the target physical quantity Gd without temperature dependence and the specific coefficient value W of the correction coefficient β (correction coefficient β(Gd) (for example, correction coefficient β(Rd) or correction coefficient β(fd))) quickly and uniquely.

[0345] Further, according to the correction coefficient determination method of Embodiment 1, by changing the first current Ip and / or the second current Im in Process S2, the temperature T when the correction voltage Vc is zero, i.e., the first temperature T1, can be easily changed.

[0346] Figure 23 (a) is Figure 22 a flowchart of Process S3 in Figure 23 (a). As shown in Figure 22 Process S3 (determination process of the reference voltage Vgs0) of

[0347] In Process S31, the measuring device 400 of the measurement system SYS measures the field effect transistor-related physical quantity G while changing the voltage value of the reference voltage Vgs0 at the first temperature T1.

[0348] In Process S32, the computer 300 of the measurement system SYS determines the voltage value of the reference voltage Vgs0 when the physical quantity G is substantially the same as the target physical quantity Gd among the multiple physical quantities G measured while changing the voltage value of the reference voltage Vgs0 in Process S31 as the specific voltage value X of the reference voltage Vgs0. Therefore, according to Embodiment 1, the voltage value of the reference voltage Vgs0 when the target physical quantity Gd is obtained, i.e., the specific voltage value X, can be determined quickly and uniquely.

[0349] Figure 23 (b) is Figure 22 a flowchart of Process S4 in Figure 23 (b). As shown in Figure 22 Process S4 (determination process of the correction coefficient β) of

[0350] In Process S41, the measuring device 400 measures the field effect transistor TN-related physical quantity G while changing the value of the correction coefficient β at the second temperature T2 and the specific voltage value X of the reference voltage Vgs0.

[0351] In Process S42, the computer 300 determines the value of the correction coefficient β when the physical quantity G is substantially the same as the target physical quantity Gd among the multiple physical quantities G measured while changing the value of the correction coefficient β in Process S41 as the specific coefficient value W of the correction coefficient β. Therefore, according to Embodiment 1, the value of the correction coefficient β corresponding to the target physical quantity Gd, i.e., the specific coefficient value W, can be determined quickly and uniquely.

[0352] (Variant Example)

[0353] Refer to Figure 24 (a) to Figure 24 (d) to illustrate the first to fourth variant examples of Embodiment 1 of the present invention. Figure 24(a) to Figure 24 In (d), in order to simplify the drawings, the control voltage generation unit 10 and the temperature detection unit 13 are omitted. Hereinafter, the differences from the Figure 1 resistance device 100 in the first embodiment will be mainly described.

[0354] Figure 24 (a) shows a resistance device 100A according to the first modification of the first embodiment. As Figure 24 shown in (a), the resistance device 100A includes a voltage-controlled voltage source 19 and a plurality of transistors TN. The plurality of transistors TN are connected in series between the node n1 and the node n2. The voltage-controlled voltage source 19 is connected between the line LN to which the gate terminals of the plurality of transistors TN are connected and the node n1. Therefore, the voltage-controlled voltage source 19 applies a control voltage Vgs between the gate terminals of the plurality of transistors TN and the single source terminal connected to the node n1.

[0355] Further, the back gate terminals of the plurality of transistors TN are each connected to the source terminal. Therefore, compared with the case where the back gate terminal is not connected to the source terminal, the source-drain voltage Vds of each transistor TN becomes lower. As a result, the linearity of the characteristics of each transistor TN is improved. Also, the influence of the substrate deviation effect can be suppressed.

[0356] Figure 24 (b) shows a resistance device 100B according to the second modification of the first embodiment. As Figure 24 shown in (b), the resistance device 100B includes a plurality of voltage-controlled voltage sources 19 and a plurality of transistors TN. The plurality of transistors TN are connected in series between the node n1 and the node n2. The back gate terminals of the plurality of transistors TN are each connected to the source terminal. Therefore, similarly to the first modification, the linearity of the characteristics of each transistor TN is improved, and the influence of the substrate deviation effect can also be suppressed.

[0357] The plurality of voltage-controlled voltage sources 19 are arranged to correspond to the plurality of transistors TN respectively. Further, the voltage-controlled voltage source 19 is connected between the gate terminal and the source terminal of the corresponding transistor TN. Therefore, the voltage-controlled voltage source 19 applies a control voltage Vgs between the gate and source of the corresponding transistor TN. As a result, among the plurality of transistors TN, the voltage difference between the gate and source caused by the potential of the node n2 on the drain side of the transistor TN can be suppressed.

[0358] Figure 24 (c) shows a resistance device 100C according to the third modification of the first embodiment. As Figure 24As shown in (c), the resistance device 100C includes two voltage-controlled voltage sources 19 and two transistors TN. The two transistors TN are connected in series between node n1 and node n2. In such a case, the drain terminal of one transistor TN is connected to the drain terminal of the other transistor TN. The voltage-controlled voltage sources 19 are respectively arranged between the gate terminal and the source terminal of the corresponding transistor TN. Also, the back gate terminal of each transistor TN is connected to the source terminal.

[0359] Also, the combination PA1 of one voltage-controlled voltage source 19 and one transistor TN and the combination PA2 of the other voltage-controlled voltage source 19 and the other transistor TN are symmetrically arranged. As a result, it is possible to suppress the asymmetry of the potentials with respect to nodes n1 and n2 caused by the connection target of the back gate terminal and / or the arrangement of the voltage-controlled voltage source 19.

[0360] Figure 24 (d) shows the resistance device 100D according to the fourth modification of the first embodiment. As Figure 24 As shown in (d), the resistance device 100D includes one voltage-controlled voltage source 19 and two transistors TN. The two transistors TN are connected in series between node n1 and node n2. In such a case, the source terminal of one transistor TN is connected to the source terminal of the other transistor TN. The voltage-controlled voltage source 19 is arranged between the gate terminal and the source terminal of the transistor TN. Therefore, the voltage-controlled voltage source 19 applies a control voltage Vgs between the gate and source of the two transistors TN. Also, the back gate terminal of the transistor TN is connected to the source terminal.

[0361] Also, one transistor TN and the other transistor TN are symmetrically arranged with respect to the voltage-controlled voltage source 19. As a result, it is possible to suppress the asymmetry of the potentials with respect to nodes n1 and n2 caused by the connection target of the back gate terminal and / or the arrangement of the voltage-controlled voltage source 19.

[0362] In addition, Figure 24 (a) to Figure 24 In (d), the back gate terminals of the plurality of transistors TN are each connected to the source terminal, but they may also be connected to ground or earth (0 [V]). However, the linearity is deteriorated compared with the case where the back gate terminals of the plurality of transistors TN are each connected to the source terminal.

[0363] Also, it is possible to combine Figure 24 (a) to Figure 24 two or more of the first to fourth modifications in (d). For example, the first modification and the third modification can be combined, and each of the two transistors TN shown in Figure 24 (c) can be changed to Figure 24Multiple transistors TN shown in (a). For example, the first modification example and the second modification example can be combined, and each of the multiple transistors TN shown in Figure 24 (b) is changed to Figure 24 the multiple transistors TN shown in (a). For example, the resistance devices related to two or more modification examples among the first to fourth modification examples can also be connected in series or in parallel.

[0364] (Embodiment 2)

[0365] Refer to Figures 25 to 32 (d) to describe the resistance device 100Z in Embodiment 2 of the present invention. In Embodiment 2, the main difference from Embodiment 1 in which an NMOS transistor is used as a MOS resistor is that in the resistance device 100Z of Embodiment 2, a PMOS transistor is used as a MOS resistor. Hereinafter, the differences between Embodiment 2 and Embodiment 1 will be mainly described.

[0366] Figure 25 This is the resistance device 100Z in Embodiment 2. As shown in Figure 25 , the resistance device 100Z includes a field effect transistor TP and a voltage application circuit 1A. As shown in Figure 25 , in Embodiment 2, the field effect transistor TP is a P-type field effect transistor. Specifically, the field effect transistor TP is a p-type MOSFET (p-type Metal-Oxide-Semiconductor Field-Effect Transistor), that is, a PMOS transistor.

[0367] Hereinafter, the field effect transistor TP is sometimes referred to as "transistor TP".

[0368] In addition, the back gate terminal of the field effect transistor TP can be connected to the source terminal of the field effect transistor TN or to the power supply.

[0369] The field effect transistor TP functions as a resistance element. Specifically, the field effect transistor TP functions as a resistance element by using the resistance between the drain and source of the field effect transistor TP. That is, the field effect transistor TP functions as a MOS resistor. More specifically, the field effect transistor TP functions as a resistance element by using the resistance between the drain and source in the region where the voltage between the gate and source of the field effect transistor TP is greater than the threshold voltage (linear region and saturation region). Also, the field effect transistor TP functions as a resistance element by using the resistance between the drain and source in the region where the voltage between the gate and source of the field effect transistor TP is less than the threshold voltage (subthreshold region).

[0370] In Figure 25In the region AR, an equivalent circuit is shown when the field effect transistor TP functions as the MOS resistor MR.

[0371] Hereinafter, the resistance value R between the drain and source of the field effect transistor TP may sometimes be referred to as "the resistance value R of the field effect transistor TN".

[0372] The voltage application circuit 1A applies a control voltage Vgs corresponding to the temperature T between the gate and source of the field effect transistor TP to control the resistance value R between the drain and source of the field effect transistor TN. "Between the gate and source of the field effect transistor TP" means "between the gate terminal and the source terminal of the field effect transistor TP". The temperature T is the ambient temperature of the resistance device 100. The control voltage Vgs has a negative value. The control voltage Vgs refers to the voltage between the gate and source of the field effect transistor TP.

[0373] The control voltage Vgs is a voltage obtained by adding a correction voltage Vc to a reference voltage Vgs0. Specifically, the control voltage Vgs is represented by Equation (28).

[0374] Vgs = Vgs0 + Vc…(28)

[0375] The correction voltage Vc is a voltage added to the reference voltage Vgs0 in order to reduce the temperature dependence of the physical quantity related to the field effect transistor TP.

[0376] The physical quantity related to the field effect transistor TP can be measured from an electronic circuit including the field effect transistor TP, and is a physical quantity including the resistance value R of the field effect transistor TP. "The physical quantity including the resistance value R" means a physical quantity that depends on the resistance value R. For example, the physical quantity related to the field effect transistor TP is the resistance value R between the drain and source of the field effect transistor TP or the cut-off frequency fc of a filter circuit including the field effect transistor TP. In addition, similar to the physical quantity related to the field effect transistor TN in the first embodiment, the physical quantity related to the field effect transistor TP is not limited to the resistance value R and the cut-off frequency fc. For example, examples of the physical quantity related to the field effect transistor TP can be examples of the physical quantity related to the field effect transistor TN. Hereinafter, the physical quantity required for the field effect transistor TP may sometimes be referred to as the "target physical quantity". Therefore, the target physical quantity can be measured from an electronic circuit including the field effect transistor TP, is a physical quantity including the resistance value R of the field effect transistor TP, and is a physical quantity set as a target value.

[0377] Specifically, the correction voltage Vc is represented by Equation (29). In Equation (29), β represents a correction coefficient, T represents temperature, and T1 represents a first temperature. The correction coefficient β is a coefficient used to determine the correction voltage Vc. In the second embodiment, the correction coefficient β has a positive value. Therefore, the higher the temperature T, the larger the correction voltage Vc. Specifically, the correction coefficient β is a coefficient for correcting the control voltage Vgs, and the control voltage Vgs is applied between the gate and source of the field-effect transistor TP to reduce the temperature dependence of the physical quantity related to the field-effect transistor TP.

[0378] Vc = β(T - T1)…(29)

[0379] As shown in Equation (29), the correction voltage Vc depends on the temperature T and is set to be zero at the first temperature T1. In other words, the first temperature T1 is the temperature at which the correction voltage Vc is zero. Therefore, according to the second embodiment, there is no correction effect at the first temperature T1. By using the voltage application circuit 1A that has no correction effect at the first temperature T1, it is possible to efficiently determine the combination of the correction coefficient β for correcting the control voltage Vgs applied between the gate and source of the field-effect transistor TP and the physical quantity related to the field-effect transistor TP. Regarding this point, it is the same as in the first embodiment.

[0380] In addition, in the second embodiment as well as in the first embodiment, in any of the saturation region, linear region, and subthreshold region in the region of "|Vgs| > |Vth|", the control voltage Vgs can be corrected according to the linear function shown in Equation (29).

[0381] In the second embodiment, since the field-effect transistor TP is a PMOS transistor, the region of "|Vgs| > |Vth|" is the operating region of the PMOS transistor when the magnitude of the gate-source voltage Vgs is greater than the magnitude of the threshold voltage Vth. The region of "|Vgs| > |Vth|" corresponds to an example of "the first operating region of the field-effect transistor". The region of "|Vds| > |Vgs - Vth|" in the region of "|Vgs| > |Vth|" is the saturation region of the PMOS transistor. The region of "|Vds| < |Vgs - Vth|" in the region of "|Vgs| > |Vth|" is the linear region of the PMOS transistor. Also, the subthreshold region is the operating region of the PMOS transistor when the magnitude of the gate-source voltage Vgs is less than the magnitude of the threshold voltage Vth (|Vgs| < |Vth|). The subthreshold region corresponds to an example of "the second operating region of the field-effect transistor".

[0382] Also, the same applies to Embodiment 2. In any of the saturation region in the “|Vgs| > |Vth|” region, the linear region in the “|Vgs| > |Vth|” region, and the subthreshold region, the correction coefficient β can be determined by the same steps as those described in Figure 11 (a) to Figure 13 (c) and Figure 17 (a) to Figure 17 (c) of Reference Embodiment 1.

[0383] The reason is that the transistor TP as a PMOS transistor and the transistor TN as an NMOS transistor only have different polarities. As Figure 1 and Figure 25 shown, the different polarities of the transistor TP and the transistor TN mean that the polarities of the gate-source voltage Vgs, the drain-source voltage Vds, and the drain current Ids are different.

[0384] That is, the drain current Ids in the saturation region of the “|Vgs| > |Vth|” region of the transistor TP is represented by Equation (30). From the comparison between Equation (30) and Equation (10), it can be clearly seen that in the saturation region, the transistor TP and the transistor TN only have different polarities. The temperature coefficient αth represents a positive value.

[0385]

Equation 14

[0386]

[0387] Also, the drain current Ids in the linear region of the “|Vgs| > |Vth|” region of the transistor TP is represented by Equation (31). From the comparison between Equation (31) and Equation (9), it can be clearly seen that in the linear region, the transistor TP and the transistor TN only have different polarities. The temperature coefficient αth represents a positive value.

[0388]

Equation 15

[0389]

[0390] Also, the drain current Ids of the transistor TP in the subthreshold region (|Vgs| < |Vth|) is represented by Equation (32). From the comparison between Equation (32) and Equation (22), it can be clearly seen that in the subthreshold region, the transistor TP and the transistor TN only have different polarities.

[0391]

Equation 16

[0392]

[0393] Figure 26is a graph of the correction voltage Vc. The vertical axis represents the correction voltage Vc [V], and the horizontal axis represents the temperature T [K]. As Figure 26 shown, the correction voltage Vc varies linearly with respect to the temperature T. The slope of the straight line representing the correction voltage Vc is the correction coefficient β. Among them, since the transistor TP and the transistor TN have different polarities, therefore, compared with Figure 2 the graph of the correction voltage Vc of the first embodiment shown, the slope of the straight line representing the correction voltage Vc, that is, the correction coefficient β has a positive value.

[0394] Next, return to Figure 25 , and the voltage application circuit 1A will be described in detail. The voltage application circuit 1A is arranged between the gate terminal and the source terminal of the transistor TP. The potential of the source terminal (hereinafter, sometimes referred to as "source potential Vs") can take any value as in the first embodiment.

[0395] The voltage application circuit 1A includes a control voltage application unit 9A and a temperature detection unit 13A. The temperature detection unit 13A detects the temperature T and outputs a detection signal TM corresponding to the temperature T to the control voltage application unit 9A. In addition, the temperature detection unit 13A is the same as the Figure 1 temperature detection unit 13 in Figure 8 . For example, the temperature detection unit 13A can have the same structure as the temperature detection unit 13 in

[0396] The control voltage application unit 9A generates a control voltage Vgs according to the detection signal TM. The control voltage Vgs includes a correction voltage Vc that varies linearly with respect to the temperature T. Then, the control voltage application unit 9A applies the control voltage Vgs between the gate and source of the transistor TP. In addition, the control voltage application unit 9A is the same as the Figure 1 control voltage application unit 9 in

[0397] Specifically, the control voltage application unit 9A includes a control voltage generation unit 10A and a voltage-controlled voltage source 19A.

[0398] The control voltage generation unit 10A generates a control voltage Vgsa based on the detection signal TM, which includes a correction voltage Vc that varies linearly with respect to the temperature T. The control voltage Vgsa has a negative value. The control voltage Vgsa can also be referred to as the "reference control voltage Vgsa". The control voltage Vgsa is represented by Equation (33).

[0399] Vgsa = Vgs0 + Vc…(33)

[0400] As can be clearly seen from Equation (28) and Equation (33), the control voltage Vgsa and the control voltage Vgs have the same voltage components (reference voltage Vgs0 and correction voltage Vc) and the same voltage value. In Equation (33), the correction voltage Vc is also represented by Equation (29). In addition, the control voltage generation unit 10A is the same as the Figure 1 control voltage generation unit 10 in

[0401] Similar to the first embodiment, the control voltage Vgsa is a voltage with 0 [V] as the reference (that is, the potential difference with 0 [V] as the reference). Therefore, for the same reason as in the first embodiment, the control voltage application unit 9A has a voltage-controlled voltage source 19A.

[0402] The voltage-controlled voltage source 19A is connected between the gate terminal and the source terminal of the transistor TP. The voltage-controlled voltage source 19A is a voltage source that determines the potential difference between the output two terminals according to the potential difference between the input two terminals. In the voltage-controlled voltage source 19A, the control voltage generation unit 10A inputs the control voltage Vgsa with 0 [V] as the reference and the voltage 0 [V] as the reference, and the control voltage Vgsa is input as the potential difference. Then, by connecting the two terminals of the output of the voltage-controlled voltage source 19A to the gate terminal and the source terminal of the transistor TN respectively, a control voltage Vgs with the same voltage value as the control voltage Vgsa is applied between the gate and source of the transistor TP. At this time, even if the source potential Vs changes, the potential Vgs between the two terminals of the output of the voltage-controlled voltage source 19A does not change. In addition, the voltage-controlled voltage source 19A is the same as the Figure 1 voltage-controlled voltage source 19 in Figure 7 (a) and Figure 7 (b) has the same structure as the voltage-controlled voltage source 19.

[0403] Furthermore, in the voltage-controlled voltage source 19A, as long as the control voltage Vgsa is input as the potential difference, therefore, similar to the first embodiment, the voltage 0 [V] as the reference can be any value. In such a case, if the reference voltage is Vref and the output voltage from the control voltage generation unit 10A is "Vgsa + Vref", then the potential difference input to the voltage-controlled voltage source 19A is Vgsa of "Vgsa + Vref - Vref".

[0404] The control voltage generation unit 10A only needs to be able to generate the control voltage Vgsa shown in Equation (33), and the structure of the control voltage generation unit 10A is not particularly limited and can be composed of any control voltage generation circuit.

[0405] In Embodiment 2, the control voltage generation unit 10A includes a reference voltage generation unit 11A, a correction voltage generation unit 15A, and an addition calculation unit 17A.

[0406] The reference voltage generation unit 11A generates a reference voltage Vgs0 and outputs it to the addition calculation unit 17A. The reference voltage Vgs0 has a negative value. On the other hand, the correction voltage generation unit 15A generates a correction voltage Vc based on the detection signal TM of the temperature detection unit 13A and outputs it to the addition calculation unit 17A. The addition calculation unit 17A adds the correction voltage Vc to the reference voltage Vgs0 to generate a control voltage Vgsa as the addition calculation result. Then, the addition calculation unit 17A outputs the control voltage Vgsa to the voltage-controlled voltage source 19A.

[0407] In addition, the reference voltage generation unit 11A, the correction voltage generation unit 15A, and the addition calculation unit 17A are respectively the same as Figure 1 the reference voltage generation unit 11, the correction voltage generation unit 15, and the addition calculation unit 17 in Figure 8 . For example, the correction voltage generation unit 15A may have the same structure as Figure 25 the correction voltage generation unit 15 in Figure 1 . In addition,

[0408] Next, with reference to Figure 8 and Figure 27 (a) to Figure 27 (c), the temperature detection unit 13A and the correction voltage generation unit 15A will be described in detail. The temperature detection unit 13A and the correction voltage generation unit 15A are respectively the same as Figure 8 the temperature detection unit 13 and the correction voltage generation unit 15 in

[0409] Figure 27 (a) is a temperature dependence chart of the first current Ip and the second current Im. Figure 27 (b) is a temperature dependence chart of the difference current Io. Figure 27 (c) is a temperature dependence chart of the correction voltage Vc.

[0410] As shown in Figure 27 (a), similar to Embodiment 1, the temperature dependence of the first current Ip is different from that of the second current Im. The first current Ip and the second current Im each linearly change with respect to the temperature T.

[0411] Figure 27 In the example of (a), the temperature dependence of the first current Ip is higher than that of the second current Im. That is, the temperature dependence of the first current source circuit 131 is higher than that of the second current source circuit 133.

[0412] The temperature T when the first current Ip is equal to the second current Im is the first temperature T1. That is, the temperature T when the differential current Io is zero is the first temperature T1.

[0413] As Figure 27 shown in (b), the differential current Io (= Ip - Im) has a positive temperature characteristic. That is, the slope A of the straight line representing the differential current Io has a positive value. Therefore, the differential current Io is represented by Equation (34). When the temperature T is the first temperature T1, the differential current Io is zero.

[0414] Io = A × (T - T1) … (34)

[0415] Since the differential current Io has a positive temperature characteristic, as Figure 27 shown in (c), the correction voltage Vc also has a positive temperature characteristic. Figure 27 The graph in (c) is consistent with the graph of the temperature dependence of the correction voltage Vc Figure 26 shown. The correction voltage Vc is represented by Equation (35).

[0416] Vc = Ro × Io = Ro × A × (T - T1) = β(T - T1)

[0417] … (35)

[0418] As shown in Equation (35), “Ro × A” is the correction coefficient β. The correction coefficient β represents a positive value. Also, when the temperature T is the first temperature T1, the correction voltage Vc is zero.

[0419] In addition, Figure 27 in the example of (a), the slopes of the straight lines representing the first current Ip and the second current Im are both positive values, but they do not necessarily have to be positive. In Embodiment 2, in order to correct the temperature dependence of the transistor TP, the correction coefficient β of the correction voltage Vc is positive, and as long as the slope of the first current Ip is greater than the slope of the second current Im, the sign of the slope is irrelevant.

[0420] In addition, in Embodiment 2, similar to Embodiment 1, by changing the current value of the first current Ip and / or the current value of the second current Im, the first temperature T1 can be changed.

[0421] Among them, after the correction coefficient β for correcting the temperature dependence of the transistor TP is reflected in Equation (30), in the saturation region (|Vds| > |Vgs - Vth|) in the region of “|Vgs| > |Vth|”, the drain current Ids of the transistor TP is represented by Equation (36).

[0422]

Equation 17

[0423]

[0424] Further, after reflecting the correction coefficient β into Equation (31), in the linear region (|Vds| < |Vgs - Vth|) within the region of "|Vgs| > |Vth|", the drain current Ids of the transistor TP is represented by Equation (37). The temperature coefficient αth represents a positive value.

[0425]

Equation 18

[0426]

[0427] Further, after reflecting the correction coefficient β into Equation (32), in the subthreshold region, the drain current Ids of the transistor TP is represented by Equation (38). The temperature coefficient αth represents a positive value.

[0428]

Equation 19

[0429]

[0430] In any one of the saturation region, linear region, and subthreshold region, the resistance value R of the transistor TP is represented by Equation (39).

[0431] R = Vds / Ids…(39)

[0432] Next, with reference to Figure 28 (a), an example of a method for determining the correction coefficient β will be described. Figure 28 (a) is a graph showing the relationship between the correction coefficient β and the resistance value R of the transistor TP at the first temperature T1 and the relationship between the correction coefficient β and the resistance value R of the transistor TP at the second temperature T2. The first temperature T1 and the second temperature T2 are the same as the first temperature T1 and the second temperature T2 in the first embodiment, respectively.

[0433] As Figure 28 (a) shows, the R-β curve G110 represents the resistance value R at the first temperature T1. The R-β curve G120 represents the resistance value R at the second temperature T2. Then, by referring to Figure 12(b) illustrates the same steps as in the first embodiment to obtain the correction coefficient β(Rr) at the intersection point P of the R-β curve G110 and the R-β curve G120. In the second embodiment, in the same manner as in the first embodiment, the resistance value Rr corresponding to the correction coefficient β(Rr) having no temperature dependence must be consistent with the target resistance value Rd. At this time, the correction coefficient β(Rr) at the intersection point P is also consistent with the correction coefficient β(Rd) with respect to the target resistance value Rd. Therefore, according to the second embodiment, after setting the correction coefficient β(Rd) as the correction coefficient β of the correction voltage generation unit 15A of the resistance device 100A, the temperature dependence of the resistance value R of the transistor TP can be effectively reduced through the correction voltage Vc, and the resistance value R can be maintained at the target resistance value Rd.

[0434] Next, referring to Figure 28 (b) and Figure 28 (c), a further preferred example of the method for determining the correction coefficient β will be described. Figure 28 (b) is a relationship graph of the reference voltage Vgs0 and the resistance value R of the transistor TP at the first temperature T1. Figure 28 (c) is a relationship graph of the correction coefficient β and the resistance value R of the transistor TP at the second temperature T2. The correction coefficient β is determined through the following (Step 1) and (Step 2).

[0435] (Step 1) As shown in Figure 28 (b), at the first temperature T1 where the correction voltage Vc is zero, while changing the voltage value of the reference voltage Vgs0, the resistance value R of the transistor TP of the resistance device 100A is measured, and the reference voltage Vgs0(Rd) when the resistance value R represents the target resistance value Rd is determined. In addition, at the first temperature T1, since the correction voltage Vc is zero, the correction coefficient β can be any value, and the reference voltage Vgs0(Rd) is consistent with Figure 25 the control voltage Vgsa in Equation (33).

[0436] (Step 2) As shown in Figure 28 (c), in the resistance device 100A, the reference voltage Vgs0 is set to the reference voltage Vgs0(Rd) determined in (Step 1). Next, at the second temperature T2 different from the first temperature T1, while changing the value of the correction coefficient β, the resistance value R of the transistor TP is measured, and the correction coefficient β(Rd) when the resistance value R represents the target resistance value Rd is determined. The target resistance value Rd is consistent with the resistance value Rr having no temperature dependence.

[0437] As described above, referring to Figure 28 (a) and Figure 28(b) As described, according to the second embodiment, through (Step 1) and (Step 2), the correction coefficient β(Rd) is determined when the target resistance value Rd that is consistent with the temperature-independent resistance value Rr is obtained. Therefore, in the second embodiment, similar to the first embodiment, the combination of the temperature-independent target resistance value Rd and the correction coefficient β(Rd) can be determined quickly and uniquely.

[0438] In particular, in order to implement (Step 1) and (Step 2), the voltage application circuit 1 preferably generates a correction voltage Vc that has no correction effect at the first temperature T1.

[0439] In addition, the method for measuring the resistance value R of the transistor TP when determining the correction coefficient β is the same as that described in the first embodiment with reference to Figure 18 (a) and Figure 18 (b).

[0440] From the above, it can also be understood that in the second embodiment where the transistor TP of the PMOS transistor is used as a MOS resistor, temperature correction can be performed in the same manner as in the first embodiment.

[0441] So far, for the purpose of performing temperature correction on the resistance value R of the transistor TP of the resistance device 100Z, for example, the method for measuring the resistance value R of the target transistor TP has been described. However, similar to the first embodiment, temperature correction of the resistance value R of the transistor TN can also be performed using a physical quantity other than the resistance value R. In other words, by using a physical quantity (hereinafter referred to as "physical quantity G") that can be measured from an electronic circuit including the resistance device 100Z that performs temperature correction, the reference voltage Vgs0 and the correction coefficient β for temperature correction can be determined. Hereinafter, the physical quantity G that can be measured from the electronic circuit is sometimes referred to as the "physical quantity G of the electronic circuit".

[0442] The same is true for the second embodiment. When the transistor TP (MOS resistor MR) of the resistance device 100Z is built in as a circuit element in the Figure 19 (a) electronic circuit 3B, the physical quantity G that can be measured from the electronic circuit 3B, similar to Equation (24) of the first embodiment, can be generally expressed as a function of the resistance value R of the transistor TP, and can also be generally expressed as a function of the temperature T, the reference voltage Vgs0, and the correction coefficient β.

[0443] That is, similar to the first embodiment with reference to Figure 19 (a), Figure 19(b) Similar to the description of the sum formula (24), in the second embodiment, by measuring the physical quantity G of the resistance value R of the transistor TP including the resistance device 100Z from the electronic circuit 3B of the transistor TP including the resistance device 100Z, the reference voltage Vgs0(Gd) and the correction coefficient β(Gd) with respect to the target physical quantity Gd can be determined, and temperature correction can be implemented.

[0444] Also, similar to the reference in the first embodiment Figure 19 (a), Figure 20 and Figure 21 the description of, in the second embodiment, as Figure 19 an example in (a), the structure of the electronic circuit 3B is Figure 20 (a) or Figure 20 (b) the RC filter circuit 110X in or Figure 21 the active filter circuit 110C in, and the measured physical quantity G is the cut-off frequency fc. In such a case, temperature correction of the transistor TP of the resistance device 100Z included in the RC filter circuit 110X or the active filter circuit 110C can also be implemented.

[0445] Among them, in the second embodiment, with respect to the transistor TP of the resistance device 100Z, the correction coefficient β can also be determined by referring to the correction coefficient determination method in the first embodiment described in Figures 22 to 23 (b). In addition, in the second embodiment, the determination methods of the reference voltage Vgs0 and the correction coefficient β described in referring to Figure 11 (a) and Figure 12 (a) can also be used.

[0446] Also, in the second embodiment, it is also possible to arrange the transistor TP and the voltage control voltage source 19A in the same way as the transistor TN and the voltage control voltage source 19 in the first embodiment described in referring to Figure 24 (a) to Figure 24 (d).

[0447] Next, as a specific circuit example in the second embodiment, an example of the voltage application circuit 1A will be described with reference to Figure 29 (a) and Figure 29 (b). Figure 29 (a) is a circuit diagram of an example of the voltage application circuit 1A. As shown in Figure 29 (a), the voltage application circuit 1A includes a digital-to-analog converter (DAC: Digital to Analog Converter) 110, a resistance element Rg, a PTAT circuit 130, a variable resistor 150, a differential amplifier 170, and a voltage control voltage source 19.

[0448] The output terminal of the DAC 110 is connected to one terminal of the resistance element Rg and the inverting input terminal of the differential amplifier 170. The other terminal of the resistance element Rg is connected to the output terminal of the differential amplifier 170. One terminal of the variable resistor 150 is connected to the non-inverting input terminal of the differential amplifier 170. The reference voltage Vref from the reference voltage generation circuit is input to the other terminal of the variable resistor 150.

[0449] In the second embodiment, the DAC 110 is, for example, an m-bit R-2R ladder DAC. In the DAC 110, an input code d is input. The input code d is a digital code and is set in the range of "0 < d ≤ 2 m ". The DAC 110 outputs a current Ig0 from the output terminal according to the input code d. The current Ig0 is represented by Equation (40). In Equation (40), Ilsb is the minimum value of the current that the DAC 110 can output.

[0450] Ig0 = d × Ilsb…(40)

[0451] The variable resistor 150 is, for example, an n-bit resistive divider (n-bit digital potentiometer). Figure 29 In, both ends of the variable resistor 150 are respectively connected to the reference voltage Vref and the non-inverting input terminal of the differential amplifier 170. The intermediate node Nm located between both ends of the variable resistor 150 is connected to the PTAT circuit 130. The variable resistor 150 has a structure in which two resistors with the same resistance value are arranged in series, and the resistance value of one resistor is r. In the variable resistor 150, an input code s is input. The input code s is a digital code and is set in the range of "0 < s ≤ 2 n ". n Figure 29 In, the node between the s-th resistor and the (s + 1)-th resistor and the intermediate node Nm of the variable resistor are connected from the input side of the reference voltage Vref. Therefore, the resistance value from the input side of the reference voltage Vref to the intermediate node Nm is set to "s × r" according to the input code s.

[0452] ​The PTAT circuit 130 outputs a PTAT current Iptat corresponding to the temperature T, and inputs the PTAT current Iptat to the intermediate node Nm of the variable resistor 150. Among them, when no current flows through the non-inverting input terminal of the differential amplifier 170, no current flows from the PTAT circuit 130 to the differential amplifier 170. Therefore, all of the PTAT current Iptat flows to the reference voltage Vref side. At this time, no voltage drop occurs across the resistor between the intermediate node Nm of the variable resistor 150 and the non-inverting input terminal of the differential amplifier 170. As a result, the potential of the intermediate node Nm is equal to the potential Vp of the non-inverting input terminal of the differential amplifier 170. The potential of the intermediate node Nm is the potential obtained by adding the reference voltage Vref to the voltage of the resistance value "s×r" from the intermediate node Nm to the input side of the reference voltage Vref and the PTAT current Iptat. As can be seen from the above, the voltage Vp of the non-inverting input terminal of the differential amplifier 170 is represented by Equation (41). The PTAT current Iptat corresponds to the detection signal TM. Therefore, the PTAT circuit 130 corresponds to the temperature detection unit 13.

[0453] Vp = s×r×Iptat + Vref…(41)

[0454] Figure 29 (b) shows the temperature dependence of the PTAT current Iptat. The horizontal axis represents the temperature [K], and the vertical axis represents the current value [A]. As Figure 29 (b) shows, the PTAT current Iptat is proportional to the temperature T. The PTAT current Iptat is represented by Equation (42). p is a proportionality constant, and T1 is the first temperature. p has a positive value. Regarding the PTAT circuit 130, it is configured to include the first temperature T1 when the current value of the PTAT current Iptat is zero.

[0455] Iptat = p(T - T1)…(42)

[0456] Returning to Figure 29 (a), the differential amplifier 170 inputs the voltage determined by the resistance value Rg of the resistance element Rg and the current Ig0 and the voltage Vp, and outputs the control voltage Vgsa. The control voltage Vgsa is represented by Equation (43). Vgs0{d} in Equation (43) is represented by Equation (44), and β{s} is represented by Equation (45). Vgs0{d} corresponds to the reference voltage Vgs0, and β{s} corresponds to the correction coefficient β.

[0457] Vgsa = -Rg×Ig0 + Vp

[0458] = -d×Rg×Ilsb + Vref + s×r×p(T - T1)

[0459] = Vgs0{d} + β{s} × (T - T1)

[0460] …(43)

[0461] Vgs0{d} = -d × Rg × Ilsb + Vref…(44)

[0462] β{s} = s × r × p…(45)

[0463] As can be clearly seen from Equation (44), the value of the reference voltage Vgs0{d} can be changed by the input code d. Also, as can be clearly seen from Equation (45), the value of the correction coefficient β{s} can be changed by the input code s.

[0464] Refer to Figure 30 and Figure 31 to illustrate Figure 29 an example of the PTAT circuit 130 in Figure 30 is a circuit diagram of an example of the PTAT circuit 130. Figure 31 is a graph of the PTAT current Iptat. The horizontal axis represents the temperature [K], and the vertical axis represents the current value [A].

[0465] As Figure 30 shown, the PTAT circuit 130 includes a BGR (Band Gap Reference) circuit 140, PMOS transistors T5 to T12, NMOS transistors T14 to T17, a variable resistor R2, a resistor element R7, a capacitor C2, and an operational amplifier AP2. The BGR circuit 140 includes PMOS transistors T1, T2, bipolar transistors T3, T4, a variable resistor R1, resistor elements R3 to R6, a capacitor C1, and an operational amplifier AP1.

[0466] The PTAT circuit 130 outputs a PTAT current Iptat that is proportional to the temperature T.

[0467] Specifically, the current Ip1 flowing through the BGR circuit 140 has a positive linear temperature characteristic. The first current Ip flowing through the PMOS transistor T11 can be expressed as Ip = Ip1 / A according to the current mirror circuit. That is, A:1 = Ip1:Ip. Therefore, as Figure 31 shown, the first current Ip has a positive temperature characteristic that is proportional to the current Ip1.

[0468] The BGR circuit 140 outputs a voltage Vbgr with low temperature dependence. The voltage Vbgr with low temperature dependence is input to the non-inverting input terminal of the operational amplifier AP2. In such a case, due to the virtual short circuit effect of the operational amplifier AP2, a voltage Vz (≈Vbgr) with low temperature dependence is obtained at the inverting input terminal of the operational amplifier AP2. Also, by applying the voltage Vz to the variable resistor R2, a current Ip2 with low temperature dependence (=R2×Vz) flows through the variable resistor R2.

[0469] The second current Im flowing through the NMOS transistor T17 can be expressed as Im = Ip2a = Ip2 / B according to the current mirror circuit. That is, B:1 = Ip2:Ip2a. Therefore, as Figure 31 shown, the second current Im is proportional to the current Ip2 with low temperature dependence and has low temperature dependence. By adjusting the resistance value of the variable resistor R2, the current value of the second current Im can be changed.

[0470] The PTAT circuit 130 outputs the difference current between the first current Ip and the second current Im as the PTAT current Iptat (=Ip - Im) to Figure 29 (a)'s variable resistor 150. In such a case, as Figure 31 shown, at the first temperature T1 where the first current Ip and the second current Im are the same, the PTAT current Iptat is zero [A].

[0471] As Figure 31 shown, by adjusting the resistance value of the variable resistor R2 to change the current value of the second current Im, the first temperature T1 at which the PTAT current Iptat is zero [A] can be adjusted.

[0472] In addition, at the gate terminals of the transistors T2, T6, T8, T10, and T12, a voltage Vb that activates the PTAT circuit 130 is input.

[0473] (Modification example)

[0474] Refer to Figure 32 (a) to Figure 32 (d) to describe the first to fourth modification examples of the second embodiment of the present invention. Figure 32 (a) to Figure 32 (d), in order to simplify the drawings, the control voltage generation unit 10A and the temperature detection unit 13A are omitted. Hereinafter, mainly describe the points different from the Figure 25 shown resistance device 100Z in the second embodiment in the first to fourth modification examples.

[0475] Figure 32 (a) is the resistance device 100E related to the first modification example of the second embodiment. AsFigure 32 As shown in (a), the resistance device 100E includes a voltage-controlled voltage source 19A and a plurality of transistors TP. The plurality of transistors TP are connected in series between node n1 and node n2. The voltage-controlled voltage source 19A is connected between the line LN to which the gate terminals of the plurality of transistors TP are connected and node n1. Therefore, the voltage-controlled voltage source 19A applies a control voltage Vgs between the gate terminal of each of the plurality of transistors TP and the single source terminal connected to node n1.

[0476] Furthermore, the back gate terminal of each of the plurality of transistors TP is connected to the source terminal. Therefore, for the same reason as in the case of the resistance device 100A shown in Figure 24 (a), the linearity of the characteristics of each transistor TP is improved, and the influence of the substrate deviation effect can also be suppressed.

[0477] Figure 32 (b) shows a resistance device 100F according to a second modification of the second embodiment. As Figure 32 (b) shows, the resistance device 100F includes a plurality of voltage-controlled voltage sources 19A and a plurality of transistors TP. The plurality of transistors TP are connected in series between node n1 and node n2. The back gate terminal of each of the plurality of transistors TP is connected to the source terminal. Therefore, similarly to the first modification, the linearity of the characteristics of each transistor TP is improved, and the influence of the substrate deviation effect can also be suppressed.

[0478] The plurality of voltage-controlled voltage sources 19A are arranged corresponding to the plurality of transistors TP respectively. Furthermore, the voltage-controlled voltage source 19A is connected between the gate terminal and the source terminal of the corresponding transistor TP. Therefore, the voltage-controlled voltage source 19A applies a control voltage Vgs between the gate and source of the corresponding transistor TP. As a result, the voltage difference between the gate and source caused by the potential of node n2 on the drain side of the transistor TP among the plurality of transistors TP can be suppressed.

[0479] Figure 32 (c) shows a resistance device 100G according to a third modification of the second embodiment. As Figure 32 (c) shows, the resistance device 100G includes two voltage-controlled voltage sources 19A and two transistors TP. The two transistors TP are connected in series between node n1 and node n2. In such a case, the drain terminal of one transistor TP is connected to the drain terminal of the other transistor TP. The voltage-controlled voltage sources 19A are respectively arranged between the gate terminal and the source terminal of the corresponding transistor TP. Furthermore, the back gate terminal of each transistor TP is connected to the source terminal.

[0480] Further, a combination PB1 of a voltage-controlled voltage source 19A and a transistor TP and a combination PB2 of another voltage-controlled voltage source 19A and another transistor TP are symmetrically arranged. As a result, it is possible to suppress the asymmetry in the potentials with respect to nodes n1 and n2 caused by the connection target of the back gate terminal and / or the arrangement of the voltage-controlled voltage source 19.

[0481] Figure 32 (d) shows a resistor device 100H according to a fourth modification of the second embodiment. As Figure 32 (d) shows, the resistor device 100H includes one voltage-controlled voltage source 19A and two transistors TP. The two transistors TP are connected in series between node n1 and node n2. In such a case, the source terminal of one transistor TP is connected to the source terminal of the other transistor TP. The voltage-controlled voltage source 19A is arranged between the gate terminal and the source terminal of the transistor TP. Therefore, the voltage-controlled voltage source 19A applies a control voltage Vgs between the gate and source of the two transistors TP. Further, the back gate terminal of the transistor TP is connected to the source terminal.

[0482] Further, one transistor TP and the other transistor TP are symmetrically arranged with respect to the voltage-controlled voltage source 19A. As a result, it is possible to suppress the asymmetry in the potentials with respect to nodes n1 and n2 caused by the connection target of the back gate terminal and / or the arrangement of the voltage-controlled voltage source 19.

[0483] In addition, Figure 32 (a) to Figure 32 (d), the back gate terminals of the plurality of transistors TP are each connected to the source terminal, but may also be connected to the power supply voltage. However, the linearity is poorer than the case where the back gate terminals of the plurality of transistors TP are each connected to the source terminal.

[0484] Further, it is also possible to combine Figure 32 (a) to Figure 32 (d) of two or more of the first modification to the fourth modification. For example, the first modification and the third modification can be combined to change each of the two transistors TP shown in Figure 32 (c) to Figure 32 the plurality of transistors TP shown in Figure 32 (a). For example, the first modification and the second modification can be combined to change each of the plurality of transistors TP shown in Figure 32 (b) to

[0485] (Embodiment 3)

[0486] Refer toFigure 33 This describes the resistor device 100Q involved in Embodiment 3 of the present invention. The main difference between Embodiment 3 and Embodiment 1 and Embodiment 2 is that the resistor device 100Q involved in Embodiment 3 uses NMOS transistors and PMOS transistors with different polarities as MOS resistors. Hereinafter, the differences between Embodiment 3 and Embodiment 1 and Embodiment 2 will be mainly described.

[0487] Figure 33 This is the resistor device 100Q involved in Embodiment 3. As Figure 33 shown, the resistor device 100Q includes a voltage application circuit 1, a voltage-controlled voltage source 19, a field-effect transistor TN, a voltage application circuit 1A, a voltage-controlled voltage source 19A, and a field-effect transistor TP. The structures of the voltage application circuit 1, the voltage-controlled voltage source 19, and the field-effect transistor TN are the same as those of the voltage application circuit 1, the voltage-controlled voltage source 19, and the field-effect transistor TN described with reference to Figure 1 The structures of the voltage application circuit 1A, the voltage-controlled voltage source 19A, and the field-effect transistor TP are the same as those of the voltage application circuit 1A, the voltage-controlled voltage source 19A, and the field-effect transistor TP described with reference to Figure 25 respectively.

[0488] The voltage-controlled voltage source 19 is connected between the gate terminal and the source terminal of the transistor TN. The voltage-controlled voltage source 19A is connected between the gate terminal and the source terminal of the transistor TP.

[0489] The transistor TN and the transistor TP are connected in parallel between the node n1 and the node n2. Specifically, the source terminal of the transistor TN is connected to the node n1, and the drain terminal of the transistor TN is connected to the node n2. Also, the source terminal of the transistor TP is connected to the node n2, and the drain terminal of the transistor TP is connected to the node n1.

[0490] As described with reference to Figure 33 above, in Embodiment 3, by connecting the transistor TP as a PMOS resistor and the transistor TN as an NMOS resistor in parallel, the operating range of the resistor device 100Q can be enlarged, and the linearity of the resistor device 100Q can be further improved. The reason is that the operating range of the transistor TP as a PMOS transistor is on the power supply voltage side, and the operating range of the transistor TN as an NMOS transistor is on the ground side. The operating range of the resistor device 100Q refers to the operating range of one MOS resistor when the transistor TP and the transistor TN are regarded as one MOS resistor. Similarly, the linearity of the resistor device 100Q refers to the linearity of the combined resistance of the transistor TP and the transistor TN when the transistor TP and the transistor TN are regarded as one MOS resistor.

[0491] (Modification Example)

[0492] (a) to Figure 24 (d) illustrate the first to fourth modification examples of Embodiment 1 of the present invention, and Figure 24 (a) to Figure 32 (d) illustrate the first to fourth modification examples of Embodiment 2. Figure 32 And Figure 33 , the modification example of Embodiment 3 will be described. In Embodiment 3, by connecting the transistor TP and the transistor TN in parallel, it is regarded as one MOS resistor. That is, Figure 33 In, a single transistor TN and a single transistor TP form one MOS resistor. However, Figure 24 (a) to Figure 24 (d) and Figure 32 (a) to Figure 32 (d) show that a plurality of transistors TN and a plurality of transistors TP can be combined to form one MOS resistor.

[0493] For example, in the resistance device 100Q, replacing Figure 33 the voltage control voltage source 19 and the transistor TN, it is possible to have Figure 24 (a) to Figure 24 (d) any one of the first to fourth modification examples of the voltage control voltage source 19 and a plurality of transistors TN, and it is also possible to have Figure 24 (a) to Figure 24 (d) a combined structure of two or more modification examples among the first to fourth modification examples. Also, for example, in the resistance device 100Q, replacing Figure 33 the voltage control voltage source 19A and the transistor TP, it is possible to have Figure 32 (a) to Figure 32 (d) any one of the first to fourth modification examples of the voltage control voltage source 19A and a plurality of transistors TP, and it is also possible to have Figure 32 (a) to Figure 32 (d) a combined structure of two or more modification examples among the first to fourth modification examples.

[0494] (Embodiment 4)

[0495] Refer to Figure 34 and Figure 35, the brain-computer interface device BMI involved in the fourth embodiment of the present invention will be described. In the brain-computer interface device BMI involved in the fourth embodiment, the resistance devices 100, 100A, 100B, 100C, 100D of the first embodiment (including modified examples), the resistance devices 100Z, 100E, 100F, 100G, 100H of the second embodiment (including modified examples), or the resistance device 100Q of the third embodiment (including modified examples) are installed. The brain-computer interface device BMI is an example of a "living body interface device". The living body interface device detects living body signals or sends stimulation signals to living body tissues and is a device that connects the living body and the computer.

[0496] Figure 34 This is the brain-computer interface device BMI involved in the fourth embodiment. As Figure 34 shown, the brain-computer interface device BMI detects brain wave signals or sends stimulation signals to the brain and is a device that connects the brain and the computer.

[0497] The brain-computer interface device BMI includes an in-vivo device 6 and an ex-vivo device 7. The in-vivo device 6 is implanted in the head HD. The in-vivo device 6 is an example of an "in-vivo implanted device implanted in the body". The head HD is an example of a "body". The in-vivo device 6 detects brain wave signals and sends brain information corresponding to the brain wave signals to the ex-vivo device 7. Also, the in-vivo device 6 sends a stimulation signal to the brain according to an instruction from the ex-vivo device 7. The ex-vivo device 7 performs an action corresponding to the brain information received from the in-vivo device 6. The ex-vivo device 7 and the in-vivo device 6 communicate wirelessly with each other.

[0498] The in-vivo device 6 includes at least one of a measuring device 63 and a stimulating device 67. In the fourth embodiment, the in-vivo device 6 includes a measuring device 63 and a stimulating device 67. Also, the in-vivo device 6 further includes a plurality of measuring electrodes 65, a plurality of stimulating electrodes 69, and a control device 61.

[0499] The control device 61 controls the measuring device 63 and the stimulating device 67. The control device 61 has a communication machine (not shown) for wireless communication with the ex-vivo device 7. The control device 61 includes a processor such as a CPU (Central Processing Unit) and a storage device such as a semiconductor memory. For example, the control device 61 is a microcomputer.

[0500] The measuring device 63 measures the electroencephalogram signal through a plurality of measuring electrodes 65. Specifically, each of the plurality of measuring electrodes 65 is disposed on the brain, detects the electroencephalogram signal, and outputs the electroencephalogram signal to the measuring device 63. The measuring device 63 includes an integrated circuit device (IC: Integrated Circuit) 631. The integrated circuit device 631 amplifies the electroencephalogram signal and outputs the amplified electroencephalogram signal to the control device 61. The control device 61 sends the brain information representing the electroencephalogram signal to the external device 7. The measuring device 63 is an example of a "measuring device disposed in the body and measuring a living body signal". The electroencephalogram signal is an example of a "living body signal".

[0501] The stimulating device 67 sends a stimulating signal to the brain through a plurality of stimulating electrodes 69. Specifically, the stimulating device 67 generates a stimulating signal and outputs the stimulating signal to the plurality of stimulating electrodes 69. Then, each of the plurality of stimulating electrodes 69 is disposed on the brain and sends the stimulating signal to the brain. The stimulating device 67 includes an integrated circuit device (IC) 671. The integrated circuit device 671 eliminates noise and amplifies the stimulating signal, and outputs the amplified stimulating signal to the plurality of stimulating electrodes 69. The stimulating device 67 is an example of a "stimulating device disposed in the body and sending a stimulating signal to a living tissue". The brain is an example of a "living tissue".

[0502] In each of the integrated circuit device 631 and the integrated circuit device 671, the voltage application circuit 1 and the field effect transistor TN of the resistance devices 100, 100A to 100D of the first embodiment (including modified examples) are integrated, or the voltage application circuit 1A and the field effect transistor TP of the resistance devices 100Z, 100E to 100H of the second embodiment (including modified examples) are integrated, or the voltage application circuits 1, 1A and the field effect transistors TN, TP of the resistance device 100Q of the third embodiment (including modified examples) are integrated. In particular, in the in-vivo device 6 implanted in the head HD, the power supply voltage cannot be too high, and electroencephalogram signals from the plurality of measuring electrodes 65 and stimulating signals to the plurality of stimulating electrodes 69 are required. Therefore, the integration of the resistance devices 100, 100A to 100H, 100Z, 100Q into integrated circuits is particularly effective.

[0503] Next, the reason why it is particularly effective to provide the integrated circuit device 631 and the integrated circuit device 671 in the in-vivo device 6 will be described. As an example, the measurement of the electroencephalogram signal will be focused on.

[0504] That is to say, in the measurement of brain wave signals, the signals after the brain wave signals in the frequency band of 1 Hz to several hundred Hz are superimposed on the slow fluctuations of the low-frequency baseline level below 1 Hz are input into the measurement device 63 through the measurement electrode 65. Generally speaking, the brain wave signals are at the level of several μV to several hundred μV. In order to avoid being affected by external noise from the surrounding environment or internal noise caused by the internal device 6, for example, the brain wave signals need to be amplified by several hundred times to several thousand times.

[0505] On the other hand, generally speaking, the fluctuations of the baseline level are larger than the brain wave signals. Therefore, if the signals obtained by the measurement electrode 65 are simply amplified, they will exceed the input range of the measurement device 63, and the brain wave signals cannot be obtained. Therefore, a filter circuit that can amplify only the brain wave signals in the required frequency band according to the signals from the measurement electrode 65 is needed. The filter circuit is installed in the integrated circuit device 631.

[0506] It is possible to use an Figure 20 RC filter circuit 110X composed of a combination of a resistor R and a capacitor C as shown or Figure 21 an active filter circuit 110C as shown as the filter circuit. In such a case, the cut-off frequency fc for frequency band limitation is expressed as fc = 1 / (2π×RC). Also, in the integrated circuit device 631, in order to constitute a filter that blocks signals in the frequency band below 1 Hz, a combination of a resistor R and a capacitor C with RC≥1 / 2π is required. However, it is difficult to achieve ordinary high-resistance polysilicon and MIM (Metal-Insulator-Metal) capacitors in a realistic size.

[0507] Then, in the resistor element constituting the filter circuit, by adopting the MOS resistors of the transistors TN and TP using the resistor devices 100, 100A to 100H, 100Z, 100Q of Embodiment 1 to Embodiment 3, a size similar to that of ordinary high-resistance polysilicon can achieve a resistance value 10 times to 10 8 times that of ordinary high-resistance polysilicon, and in a realistic size, a combination of a resistor R and a capacitor C with RC≥1 / (2π×1) can be achieved.

[0508] Moreover, by integrating the resistor devices 100, 100A to 100H, 100Z, 100Q of Embodiment 1 to Embodiment 3 into the integrated circuit device 631, according to the temperature correction of the voltage application circuits 1 and 1A, the temperature dependence of the resistance value R of the transistors TN and TP is reduced. Furthermore, the temperature dependence of the cut-off frequency fc is reduced. Therefore, by using a filter circuit (for example, Figure 21Integrating the active filter circuit 110C) into the integrated circuit device 631 can suppress the influence of temperature changes in the head HD (brain) as much as possible. According to the signal from the measurement electrode 65, it can accurately extract only the brain wave signals in the required frequency band and amplify only the brain wave signals in the required frequency band. Similarly, it is also effective to set the integrated circuit device 671 in the in-vivo device 6.

[0509] Figure 35 This is an example circuit diagram of the integrated circuit device 631. As Figure 35 shown, the integrated circuit device 631 includes a plurality of amplifiers 81, a plurality of sample-and-hold circuits 82, a multiplexer 83, an analog-to-digital converter (ADC: Analog to Digital Converter) 84, and the voltage application circuit 1 of Embodiment 1.

[0510] Each of the plurality of amplifiers 81 receives the detection signals detected by the plurality of measurement electrodes 65. Each of the plurality of amplifiers 81 eliminates noise from the detection signals of the corresponding measurement electrodes 65, extracts the brain wave signals, amplifies the brain wave signals, and outputs the amplified brain wave signals to the corresponding sample-and-hold circuits 82.

[0511] Specifically, each of the plurality of amplifiers 81 has a filter circuit including a transistor TN that is a MOS resistor of Embodiment 1, for example Figure 21 the active filter circuit 110C shown. Also, the voltage application circuit 1 applies a control voltage Vgs between the gate and source of the transistor TN (MOS resistor MR) of each of the plurality of amplifiers 81. In such a case, the voltage application circuit 1 and the plurality of transistors TN constitute a resistance device 100. Additionally, a voltage control voltage source 19 can also be provided on each transistor TN. Also, for example, each of the plurality of amplifiers 81 can also be constituted by Figure 20 the RC filter circuit 110X shown.

[0512] In addition, each of the plurality of amplifiers 81 can also have a filter circuit including a transistor TP that is a MOS resistor of Embodiment 2, and can also have a filter circuit including transistors TN and TP that are MOS resistors of Embodiment 3.

[0513] Each of the plurality of sample-and-hold circuits 82 holds the brain wave signals output by the corresponding amplifier 81 and outputs the held brain wave signals to the multiplexer 83. The multiplexer 83 outputs the serial signal after multiplexing the brain wave signals output by the plurality of sample-and-hold circuits 82 to the ADC 84. The ADC 84 converts the analog serial signal into a digital signal and outputs the digital serial signal to the control device 61.

[0514] So far, in the fourth embodiment, the brain-machine interface device BMI that measures and stimulates the brain of a living body has been described as an application example of the resistance devices 100, 100A to 100H, 100Z, and 100Q. However, the application of the resistance devices 100, 100A to 100H, 100Z, and 100Q is not limited to the brain-machine interface device BMI. For example, it can also be applied to a living body interface device. Specifically, in the fourth embodiment, the in-vivo device 6 has been described. However, the application of the resistance devices 100, 100A to 100H, 100Z, and 100Q is not limited to the in-vivo device 6 implanted in the head HD. For example, it can also be applied to an in-vivo implanted device that is implanted in the body of an animal such as a human and has at least one of a stimulation device that emits a stimulation signal to living tissue and a measurement device that measures a living body signal. For example, the in-vivo implanted device is a cardiac pacemaker or a cochlear implant.

[0515] As described above, the embodiments (including modified examples) of the present invention have been described with reference to the accompanying drawings. However, the present invention is not limited to the above-described embodiments and can be implemented in various ways without departing from its gist. Further, a plurality of structural elements disclosed in the above embodiments can be appropriately changed. For example, some of the structural elements in all the structural elements shown in a certain embodiment can be added to the structural elements of other embodiments, or some of the structural elements in all the structural elements shown in a certain embodiment can be deleted.

[0516] For ease of understanding, the structural elements are mainly shown schematically in the drawings. For convenience of drawing, the thickness, length, number, interval, etc. of the illustrated structural elements may deviate from the actual ones. Further, the structure of each structural element shown in the above embodiments is an example and is not particularly limited, and various changes can be made without substantially departing from the effects of the present invention.

[0517] 〔Industrial Applicability〕

[0518] The present invention provides a resistance device, an integrated circuit device, an in-vivo implanted device, and a correction coefficient determination method, which have industrial applicability.

[0519] 〔Explanation of Reference Numerals〕

[0520] 1, 1A Voltage application circuit

[0521] 6 In-vivo device (in-vivo implanted device)

[0522] 9, 9A Control voltage application unit

[0523] 13, 13A Temperature detection unit

[0524] 63 Measurement device

[0525] 67 Stimulation device

[0526] 100, 100A to 100H, 100Q, 100Z resistor devices

[0527] 131 First current source circuit

[0528] 133 Second current source circuit

[0529] 200, 200A, 200B electronic circuit devices (integrated circuit devices)

[0530] 631, 671 integrated circuit devices

[0531] TN, TP field effect transistors

Claims

1. A resistance device, comprising a field-effect transistor and a voltage application circuit, wherein the voltage application circuit controls the resistance value between the drain and source of the field-effect transistor by applying a control voltage corresponding to the temperature between the gate and source of the field-effect transistor, the control voltage being a voltage obtained by adding a correction voltage to a reference voltage, the correction voltage depending on the temperature and being set to zero at a first temperature, the correction voltage being represented by the following formula, Vc = β(T - T1), where "Vc" represents the correction voltage, "β" represents a correction coefficient, "T" represents the temperature as a variable, and "T1" represents the first temperature.

2. The resistance device according to claim 1, characterized in that the voltage application circuit includes a temperature detection unit and a control voltage application unit, the temperature detection unit outputting a detection signal corresponding to the temperature, the control voltage application unit generating the control voltage based on the detection signal, causing the control voltage to include the correction voltage that linearly varies with respect to the temperature, and applying the control voltage between the gate and source of the field-effect transistor.

3. The resistance device according to claim 1 or 2, characterized in that the first temperature represents the temperature at which a physical quantity related to the field-effect transistor is constant with respect to a change in the correction coefficient, the correction coefficient being a coefficient used to determine the correction voltage, the physical quantity being measurable from an electronic circuit including the field-effect transistor and including the resistance value of the field-effect transistor, and being a physical quantity set as a target value.

4. The resistance device according to claim 1 or 2, characterized in that the correction coefficient is a coefficient used to determine the correction voltage, and regarding the value of the correction coefficient, based on the reference voltage when obtaining the target physical quantity related to the field-effect transistor at the first temperature, the value when obtaining the target physical quantity at a second temperature different from the first temperature is the value of the correction coefficient, the target physical quantity being measurable from an electronic circuit including the field-effect transistor and including the resistance value of the field-effect transistor, and being a physical quantity set as a target value.

5. The resistance device according to claim 2, characterized in that the temperature detection unit includes a first current source circuit that generates a first current and a second current source circuit that generates a second current, the temperature dependence of the first current source circuit being different from the temperature dependence of the second current source circuit, the first current source circuit and the second current source circuit being connected in series, the difference current between the first current and the second current being the detection signal, changing the current value of the first current by the first current source circuit and / or changing the current value of the second current by the second current source circuit to change the first temperature.

6. The resistance device according to claim 1 or 2, characterized in that The voltage application circuit applies the control voltage between the gate and source of the field effect transistor to control the resistance value between the drain and source in the first operating region of the field effect transistor. The first operating region refers to the region where the magnitude of the voltage between the gate and source of the field effect transistor is greater than the magnitude of the threshold voltage.

7. The resistance device according to claim 1 or 2, characterized in that the voltage application circuit applies the control voltage between the gate and source of the field effect transistor to control the resistance value between the drain and source in the second operating region of the field effect transistor. The second operating region refers to the region where the magnitude of the voltage between the gate and source of the field effect transistor is smaller than the magnitude of the threshold voltage.

8. An integrated circuit device, integrating the field effect transistor and the voltage application circuit of the resistance device according to any one of claims 1 to 7.

9. An in-vivo implant device implanted in the body, characterized in that it includes at least one of a stimulation device and a measurement device, the stimulation device is used to send a stimulation signal to living tissue, and the measurement device is used to measure a living signal. At least one of the stimulation device and the measurement device includes the integrated circuit device according to claim 8.

10. A correction coefficient determination method for determining a correction coefficient when correcting a control voltage applied between the gate and source of a field effect transistor. Characterized in that In the following formula, "Vgs" represents the control voltage, "Vgs0" represents a reference voltage, "Vc" represents a correction voltage, "β" represents the correction coefficient, "T" represents a variable temperature, and "T1" represents a first temperature, that is, the temperature when the correction voltage Vc is zero. The correction coefficient determination method includes a specific voltage value determination process and a specific coefficient value determination process. The specific voltage value refers to the voltage value of the reference voltage Vgs0 when obtaining a target physical quantity related to the field effect transistor at the first temperature T1. The specific coefficient value refers to the value of the correction coefficient β when obtaining the target physical quantity under the conditions of a second temperature different from the first temperature T1 and the specific voltage value of the reference voltage Vgs0, Vgs = Vgs0 + Vc = Vgs0 + β(T - T1).

11. The correction coefficient determination method according to claim 10, characterized in that The determination process of the specific voltage value of the reference voltage Vgs0 includes: A physical quantity measurement process, measuring the physical quantity related to the field effect transistor while changing the voltage value of the reference voltage Vgs0 at the first temperature T1; And A voltage value determination process, in the process of measuring and obtaining a plurality of the physical quantities while changing the voltage value of the reference voltage Vgs0, determining the voltage value of the reference voltage Vgs0 when the measured physical quantity is consistent with the target physical quantity as the specific voltage value of the reference voltage Vgs0. The determination process of the specific coefficient value of the correction coefficient β includes: Physical quantity measurement step: Under the conditions of the second temperature and the specific voltage value of the reference voltage Vgs0, measure the physical quantity related to the field-effect transistor while changing the value of the correction coefficient β; And Coefficient value determination step: In the process of measuring a plurality of the physical quantities while changing the value of the correction coefficient β, determine the value of the correction coefficient β when the measured physical quantity is consistent with the target physical quantity as the specific coefficient value of the correction coefficient β.

12. The correction coefficient determination method according to claim 10 or 11, Characterized in that The first temperature T1 refers to the temperature at which the physical quantity related to the field-effect transistor is constant with respect to the change of the correction coefficient β, The physical quantity can be measured from an electronic circuit including the field-effect transistor and includes the resistance value of the field-effect transistor, and is a physical quantity set as a target value.

13. The correction coefficient determination method according to claim 10 or 11, Characterized in that The correction voltage Vc has a value based on the difference current between the first current and the second current, The first current represents a current that changes linearly with respect to the change in temperature, The second current represents a current that changes linearly with respect to the change in temperature, The temperature dependence of the first current is different from the temperature dependence of the second current, The correction coefficient determination method further includes a step of changing the first temperature T1 by changing at least one of the current values of the first current and the second current.

Images