JP2808195B2 - Time constant automatic adjustment circuit - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention detects an error in the resistance value of a resistor and the capacitance of a capacitor (these are simply referred to as R and C) caused by manufacturing process variations of a monolithic integrated circuit (hereinafter referred to as an IC), The present invention relates to a circuit in which a time constant can be changed by a given voltage, and which automatically adjusts a time constant error caused by an RC error based on IC process variation using the RC error detection output.

[0002]

2. Description of the Related Art In an IC, a resistor is formed by diffusion of an impurity (diffusion resistance), and a capacitor is formed by a technique such as forming a thin oxide film on a substrate and attaching a metal electrode thereon. Due to variations in diffusion and the like, variations in oxide film thickness and the like, variations in R and C occur for each semiconductor wafer.

In a time constant circuit having a resistor and a capacitor (for example, a filter circuit or a delay circuit), ± [(Δ
R) ² + (ΔC) ² ] ^1/2 error occurs.

In order to correct the time constant error, a time constant variable circuit capable of controlling the time constant by an external voltage is used. As shown in FIG.
In the above, the time constant was adjusted by adjusting the voltage applied externally at the time of factory shipment or the like using the variable resistor 59 such as a volume.

FIG. 2 shows an active filter as an example of a time constant variable circuit.

The active filter has operational amplifiers 51, 52, 52 having transfer conductances g _m1 and g _m2 , respectively.
Capacitors (capacitors) C ₁ and C ₂ for determining the time constant (cutoff frequency) together with the transfer conductances g _m1 and g _m2 , and variable gain amplifiers GCA1 and GCA2 (hereinafter simply GCA1 and GCA2) for adjusting the time constant
), And buffer circuits 53 and 54.
It is manufactured by the C process. GCA1 and GC
The gain K of A2 is adjusted by the voltage supplied from the external variable resistor 59.

The active filter has three input terminals (their inputs are denoted by v _A , v _B and v _C respectively) and one output terminal (the output is denoted by v _O ). The output v _O is given by the following equation. Where K = 1
And

[0008]

[Number 1] ^{_{v O = (s 2 C 1}} C 2 v C + sC 1 g m2 v B + g m1 g m2 v A) / (s 2 C 1 C 2 + sC 1 g m2 + g m1 g m2) ‥‥ Formula 1

By appropriately setting the inputs v _A , V _B and v _C , a low-pass filter, a high-pass filter and a band-pass filter (hereinafter referred to as LPF, HPF and BPF, respectively) can be constructed.

For example, the HPF is constructed by grounding the terminals of the inputs v _A and v _B. v _C is the input.
If v _A = v _B = 0, the following equation is obtained from Equation 1.

[0011]

## EQU2 ## v _O / v _C = (C ₁ C ₂ s ² ) / (s ² C ₁ C ₂ + sC ₁ g _m2 + g _m1 g _m2 ) = ω ₀ ² / [s ² + (ω ₀ / Q) s + ω ₀ ² ] ‥‥ Equation 2

[0012]

Ω ₀ = (g _m1 g _m2 / C ₁ C ₂ ) ^1/2 ‥‥ Equation 3

[0013]

Q = (C ₂ g _m1 / C ₁ g _m2 ) ^1/2 Equation 4

The LPF is constructed by grounding the terminals of the inputs v _B and v _C. v _A is an input. v _B
= V _C = 0, the following equation is obtained.

[0015]

V _O / v _A = (g _m1 g _m2 ) / (s ² C ₁ C ₂ + sC ₁ g _m2 + g _m1 g _m2 ) = ω ₀ ² / [s ² + (ω ₀ / Q) s + ω ₀ ² ] ‥‥ Equation 5

[0016]

Ω ₀ = (g _m1 g _m2 / C ₁ C ₂ ) ^1/2 ‥‥ Equation 6

[0017]

Q = (C ₂ g _m1 / C ₁ g _m2 ) ^1/2 Equation 7

Here, ω ₀ is the cutoff angular frequency, and Q is the sharpness.

By setting the constants g _m1 , g _m2 , C ₁ , and C _2, a filter having an arbitrary frequency characteristic can be obtained. As an example, a Butterworth filter has Q = (1/1 /
2) It is obtained by determining each constant as ^1/2 .

However, R (ie, g _m1 , g _m2 ),
If there is a process variation in C, a desired frequency characteristic cannot be obtained.

Therefore, the output currents of the operational amplifiers 51 and 52 are
By making adjustments using CA1 and GCA2, respectively, the transmission conductances g _m1 and g _m2 are substantially changed to correct errors based on process variations.

The transfer conductances corrected by the gains K of GCA1 and GCA2 are respectively g ₁ = K
g _m1 , g ₂ = Kg _m2 .

Therefore, Equation 3 or Equation 6 is modified as follows.

[0024]

Ω ₀ = (g ₁ g ₂ / C ₁ C ₂ ) ^1/2 = (Kg _m1 Kg _m2 / C ₁ C ₂ ) ^1/2 = K (g _m1 g _m2 / C ₁ C ₂ ) ^{1 / 2} Equation 8

[0025]

As described above, the conventional time constant variable circuit requires an adjusting step of changing the externally set voltage to an appropriate value so that a desired time constant can be obtained at the stage of manufacture or shipment. This led to an increase in cost. Further, there is a problem that the relationship between the externally set voltage and the time constant is not uniform due to the variability of RC almost inevitably generated in the IC process, and this IC is difficult to use. This is an important problem especially in an IC that not only corrects an error of a time constant by an externally set voltage but also sets and uses a time constant (frequency characteristic) itself to a desired value by an externally set voltage.

According to the present invention, there is provided a time constant variable circuit in which a time constant is controlled in accordance with an externally set voltage.
A method for automatically adjusting a time constant error based on an RC error by using a detection output of an RC error detection circuit so that a time constant can be uniquely controlled only by a set voltage irrespective of the presence / absence of an RC error and its magnitude. Aim.

[0027]

An automatic time constant adjusting circuit according to the present invention is a time constant variable circuit formed by an IC process on a semiconductor substrate, wherein the time constant is variably controlled by an applied control voltage. Formed on
An RC error detecting circuit for detecting an RC error caused by a variation in the C process and outputting a correction voltage representing the RC error, and setting an externally applied voltage for setting a time constant of the time constant variable circuit to the RC error A setting voltage correction circuit that corrects with a correction voltage output from the detection circuit and outputs the corrected setting voltage, and an RC error when the setting voltage is fixed based on the correction voltage output from the RC error detection circuit. A first control circuit for outputting a first control voltage for correcting a change in a time constant based on the control signal, and the correction setting voltage and the first control voltage as inputs, and the time constant of the time constant variable circuit is A second control circuit for generating a second control voltage as the control voltage and applying the second control voltage to the time constant variable circuit so as to be determined by the set voltage regardless of an RC error; Characterized in that it comprises.

Examples of the time constant variable circuit include a filter circuit and a delay circuit.

[0029]

The automatic time constant adjusting circuit according to the present invention is applied to a time constant variable circuit in which the time constant is variably controlled by a given control voltage. According to the present invention, a plurality of resistors R formed on one semiconductor wafer have the same direction of variation and the same magnitude, and a plurality of capacitors similarly formed on one semiconductor wafer. It is assumed that the directions of the variation in the capacitance of the (capacitor) are the same and their sizes are also about the same. Therefore, on the same semiconductor substrate as the time constant variable circuit, the RC
An error detection circuit is provided. The RC error detection circuit detects an RC error caused by a variation in the IC process of the semiconductor wafer, and outputs a correction voltage representing the RC error.

The variation of the time constant based on the variation of the RC in the above-mentioned variable time constant circuit includes the variation affecting the amount of change of the set voltage given from outside in order to set the desired time constant, and the variation in the fixed voltage. It can be considered separately from variations affecting the time constant.

The set voltage correction circuit generates and outputs a correction set voltage for correcting a variation affecting the amount of change of the set voltage based on the correction voltage.

The first control circuit creates and outputs a first control voltage that corrects a variation that affects the time constant when the set voltage is fixed.

Then, based on the correction set voltage and the first control voltage, the second control circuit makes R
A second control voltage to be applied to the time constant variable circuit is created so that a time constant defined by an externally applied set voltage is obtained irrespective of the variation in C.

[0034]

As described above, according to the automatic time constant adjusting circuit according to the present invention, the time constant error of the time constant variable circuit based on the RC error caused by the variation in the IC process is automatically compensated. The time constant of the variable circuit can be controlled only by the set voltage. Therefore, the setting voltage can be set without considering the time constant error based on the RC error, so that it is easy to use.

Further, since the temperature change of the time constant based on the temperature change of R and C is compensated at the same time as the compensation of the time constant error based on the R and C error, a circuit with less time constant change caused by the ambient temperature change can be realized. it can.

[0036]

FIG. 3 shows an embodiment of the present invention.

A time constant variable circuit 50 and a time constant variable circuit are provided in a monolithic IC manufactured on one semiconductor substrate.
A circuit 10 for automatically adjusting the time constant of 50 is provided. The time constant automatic adjustment circuit 10 is provided with the RC error detection circuit 1. It can be considered that the direction of variation (R increases or decreases) of a plurality of resistances formed on one semiconductor wafer and the amount of variation are substantially the same, and the plurality of resistances similarly formed on one semiconductor wafer are similar. It can be considered that the direction and the amount of variation of the capacitance of the capacitor are almost the same. The RC error detection circuit 1 includes a resistor and a capacitor (capacitor), and detects the RC error. Using the detected output (corrected voltage output) ΔV representing the detected RC error, the time constant
The C error is automatically corrected.

As an example of the time constant variable circuit 50, consider a frequency variable active filter shown in FIG.

The active filter can constitute the HPF, LPF and BPF by appropriately setting the three input terminals (v _A , v _B and v _C ) as described above. This variable frequency active filter can be arbitrarily set by a set voltage V _S that externally supplies a cutoff frequency ω ₀ in these filters. Setting voltage V _S is variable resistor 6 provided outside of the IC circuit is set by the other of the voltage generating circuit.

However, since there is an RC error based on the IC manufacturing process variation, the cutoff frequency ω ₀ is not uniquely determined with respect to the set voltage V _{S as} calculated. The change in the cutoff frequency ω ₀ due to the RC error is automatically corrected by the automatic time constant adjusting circuit 10 as described above.

Assume that the RC error coefficient in the variable frequency active filter is γ. The actual resistance value of the resistor in the active filter is R _r , and the capacitance of the capacitor (capacitor) is C _r . If the calculated values (reference values) of these resistors and capacitors are R _r0 and C _r0 , respectively,
The RC error coefficient γ is given by the following equation.

[0042]

R _r = αR _r0 ‥‥ Equation 9 C _r = βC _r0 ‥‥ Equation 10 γ = αβ ‥‥ Equation 11

FIG. 4 shows the relationship between the set voltage V _S and the cutoff frequency ω ₀ of the variable frequency active filter.

A straight line of γ = 1 represents an ideal characteristic. This characteristic line shifts right and left and changes its slope due to the RC error based on the process variation generated in the IC manufacturing process.

The correction of such a variation of the characteristic line based on the RC error is performed by shifting the varied characteristic line to a position intersecting the straight line of γ = 1 at a predetermined point (indicated by a broken line) and correcting the inclination thereof. It is performed by

In the automatic time constant adjusting circuit 10 shown in FIG. 3, the control circuit 5 includes a control voltage V _C and a correction set voltage V _SC.
And is given.

The correction voltage ΔV obtained from the RC error detection circuit 1 is amplified by the amplifier circuit 2 to become the control voltage V _C. Assuming that the gain of the amplifier circuit 2 is B, the control voltage V _C is given by the following equation.

[0048]

V _C = BΔV Equation 12

On the other hand, the set voltage V _S is applied to a variable gain amplifier (GCA) 3. The gain (gain) G of the variable gain amplifying circuit 3 is equal to the correction voltage ΔV output from the RC error detecting circuit 1 through the GCA control circuit (amplifying circuit) 4.
Is controlled by Therefore, the corrected set voltage V _SC output from the variable gain amplifier circuit 3 is given by the following equation.

[0050]

## EQU11 ## V _SC = GV _S = G ₀ (1 + DΔV) V _S ‥
‥ Equation 13

Here, G ₀ and D are the center gain and sensitivity of the variable gain amplifier circuit 3, respectively.

The control circuit 5 is a differential circuit which receives the above-mentioned control voltage V _C and correction set voltage V _SC as inputs, and includes two transistors 11 and 12 and a current source 14 for driving these transistors 11 and 12 respectively. , 15, these transistors 11 and
It comprises a resistor 13 connected between the emitters 12 of the transistor 12, diodes 16 and 17 for preventing backflow connected to the collectors of the transistors 11 and 12, and a current limiting resistor 18.
The control voltage V _C is applied to the base of the transistor 11, and the correction setting voltage V _SC is applied to the base of the transistor 12. The output voltage of the differential circuit is supplied to GCA1 and GCA2 of the variable frequency active filter shown in FIG. 2 as control voltages for controlling their gains. The transmission conductances g _m1 and g _{m2 of the} operational amplifiers 51 and 52 are adjusted by these GCA1 and GCA2, respectively.

These transfer conductances are collectively expressed as g
_Expressed by _m . The transfer conductance g _m is determined by the error coefficient γ and G
This is a function with the control voltage applied to CA1 or GCA2, and is given by the following equation.

[0054]

G _m = (g _m0 / γ) [1 + A (V _SC −V _C )]

Where g _m0 is the reference transconductance,
A is the control sensitivity of GCA1 or GCA2 by the control circuit 5.

As will be described in detail later, the RC error detection circuit 1
Is given by the following equation.

[0057]

ΔV = K (γ−1) ‥‥ 15 where K is a constant.

By substituting equation (15) into equations (12) and (13), the following equations are obtained.

[0059]

V _C = BK (γ-1) {Equation 16 V _SC = γ G ₀ V _S } Equation 17 where D = 1 / K.

By substituting Equations 16 and 17 into Equation 14, the following equation is obtained.

[0061]

G _m = g _m0 (1 + AG ₀ V _S ) {Equation 18 where B = −1 / AK.

In this way, the error coefficient γ is eliminated from the equation (18). That is, Equation 18 gives the variable frequency active
The transfer conductance g _{m of the} filter (that is, the cutoff frequency ω ₀ ) is controlled only by the set voltage V _S irrespective of the RC error in this circuit.

Referring to the relationship with the graph of FIG. 4, the characteristic straight line is shifted by the correction set voltage V _SC (Equation 13 or 17) corrected by the correction voltage ΔV, and the control voltage V _SC
C (equation 12 or equation 16) corrects the slope of the characteristic line.

In this way, a filter having a cutoff frequency defined by the set voltage V _S is obtained irrespective of the variation in RC occurring in the IC process, so that the usability is very good.

Next, a specific configuration example of the RC error detection circuit 1 will be described with reference to FIGS.

In FIG. 5, the RC error detection circuit 1
Charge / discharge circuits 20, 30; differential amplifier 7;
Capacitor C_HAnd These charge and discharge circuits 2
0 and 30 include a resistor and a capacitor, respectively.
Charge / discharge of one of the two charge / discharge circuits 20, 30
The resistors and capacitors included in the circuit are formed in the IC,
The resistor and capacitor included in the other charge / discharge circuit
It is provided outside. This embodiment will be described in detail later.
As shown in FIG._BAnd capacitor C _BBut others
Formed by IC process in IC together with other circuit components
And the resistance R of the charge / discharge circuit 20_AAnd capacitor C
_AIs provided outside the IC, and is connected to the IC externally.
Have been.

The charge / discharge circuit 20 includes a reset power supply 23 (power supply voltage = V _R ), a capacitor C _A , a constant current source 22 for supplying a current for charging the capacitor C _A , and a capacitor C _A.
It comprises a switching element 21 for controlling the charging and discharging of _A.

An example of the constant current source 22 is shown in FIG.
The constant current source 22 includes transistors 24 and 25, resistors R _R and R _A which also serve as load resistance and bias resistance of the transistors 24 and 25, and a current mirror 26. Resistor _RA is coupled with capacitor C _A as described above.
It is external to the C circuit. By two resistors R _R and transistor 24 emitter potential of the transistor 25 is determined to V _CC / 2. V _CC is a power supply voltage. Accordingly, the current of V _CC / 2R _A flows in the transistor 25, this current is to be output from the current mirror 26. Output current I _A of the constant current source 22,

## EQU16 ## I _A = V _CC / 2R _Aら れ る given by equation (19). It goes without saying that the direction of the output current is reversed by exchanging the types of transistors (pnp and npn) used in the constant current source 22.

FIG. 9 shows the charge / discharge timing and charge / discharge voltage waveform of this charge / discharge circuit 20.

The switching element 21 is turned on by the clock pulse CLK. Thereby, the capacitor C _A
Negative charge accumulated in the rapidly discharged, the terminal voltage V _A of the capacitor C _A is the reset voltage V _R of the reset voltage source 23.

[0071] When a clock pulse CLK and the switching element 21 is turned off the fall, the current I _A which is output from the current source 22 flows into the capacitor C _A, the capacitor C _A
Is negatively charged. This is the terminal voltage V _A of the capacitor C _A descends.

When the switching element 21 is turned on by the input of the next clock pulse CLK, the capacitor C _A
There is discharged, the terminal voltage V _A of the capacitor C _A is returned to the reset voltage V _R. The above operation is performed using the clock pulse CLK.
It is repeated in the cycle of. The terminal voltage V of the capacitor C _A
A is supplied to the positive input terminal of the differential amplifier 7.

[0073] charge-discharge circuit 30, a reset power supply 33 (power supply voltage V _R), the capacitor C _B, from the current source 32 and a switching element 31 for controlling the charging and discharging of the capacitor C _B for charging the capacitor C _B It is configured. The switching element 31 is turned on and off by the same clock pulse CLK that controls the switching element 21. Charging and discharging operations of the charge and discharge circuit 30 is basically the same as the charging and discharging operation of the charging and discharging circuit 20, how the change in the terminal voltage V _B of the capacitor C _B is shown in FIG.

[0074] the terminal voltage V _B of the capacitor C _B is supplied to the negative input terminal of the differential amplifier 7. The output voltage V _O of the differential amplifier 7 is held in the capacitor C _H, it is fed back to the current source 32 of the charge and discharge circuit 30. Thus, the difference between the output voltage V _A and V _B of the two charging and discharging circuit 20 and 30 is that the voltage V _O such that zero is obtained.

FIG. 7 shows an example of the configuration of the current source 32.
This current source 32 is a constant current source 40 for generating a current I _b of the two same values, and the differential circuit is driven by these currents I _b, and a current mirror 43. Differential circuit includes transistors 41 in which one of the current I _b flows, the transistor 42 which is the other current I _b flows, is composed of these transistors 41 and 42 connected to a resistor R _C between the emitters of I have. The reference voltage _VE is applied to the base of one transistor 41, and the output voltage V _O of the differential amplifier 7 is applied to the base of the other transistor 42. Resistance R _C
Is the difference between these voltages,

ΔV = V _E −V _O電流 A current ΔV / _RC according to the expression 20 flows.

The current mirror 43 includes one transistor 43a connected to the transistor 41 and the other transistor 43b.
It is composed of Since the current flowing through the transistor 43a becomes I _b -ΔV / R _C, the current of the output current I _B is also equal to the value of the current mirror 43.

[0077]

Equation 18] _{I B = I b -ΔV / R} C ‥‥ formula 21

FIG. 8 shows a configuration example of the constant current source 40. This constant current source 40 is basically a constant current source 22 shown in FIG.
And two bias resistors R _R , two transistors 34 and 35, and a resistor R that determines an output current value.
_B and a two-output current mirror 36. The two-output current mirror 36 outputs two currents of the same value. The output current I _b of the constant current source 40 is given by the following equation.

[0079]

Equation 19] _{I b = V CC / 2R B} ‥‥ formula 22

[0080] As described above, the charge and discharge circuit 30 resistance R _B and capacitor C _B in are produced by IC process the IC. The resistor R _A and the capacitor C _A in the charge / discharge circuit 20 are external and are not affected by the IC process. Therefore, when the resistance R _A and the capacitor C _A are used as references, the resistance R
R, C error in _B and capacitor C _B, appears as the difference in time constants in these charging and discharging circuit 20 and 30, it is the terminal voltage V of the capacitor C _A and C _B as shown in FIG. 9
It appears as the difference between the _A and V _B.

[0081] caused by the IC process variations in the resistance R _B and capacitor C _B R, then quantitatively describing the C error.

[0082] In the one period of the clock pulse CLK, the capacitor C in the charge-discharge circuit 20, 30 _A, C _B of the terminal voltage V _A, the time variation of V _B is expressed by the following equation. t
Represents time.

[0083]

V _A = V _R − (I _A / C _A ) t (23)

[0084]

V _B = V _R- (I _B / C _B ) t Equation 24

When the transfer conductance of the differential amplifier 7 is g, the output voltage V _O of the differential amplifier 7 is given by the following equation.

[0086]

V _O = [g (V _A −V _B ) / C _H ] t ‥‥
Equation 25

Substituting Equations 23 and 24 into Equation 25,
23 and 24 of I _A, Formula 19 as the value of I _B, the use of expressions 21 and 22 as follows.

[0088]

Equation 23] _{V O = gV CC [(1} / 2R B C B C H) - (1 / 2R A C A C H)] t 2 -gΔV (1 / R C C B C H) t 2 ‥‥ Equation 26

By substituting equation 20 into equation 26 and deriving ΔV, the following equation is obtained.

[0090]

Equation 24] _{ΔV = V E / {1- [} g / (C H C B R C)] t 2} - (gV CC / 2C H) [(1 / C B R B) - (1 / C A ^{_{R A)] t 2 / [}} 1- (g / C H C B R C) t 2] ‥‥ formula 27

If t is ∞, finally

Equation 25] [Delta] V = the _{(V CC / 2) [(} R C / R B) -C B R C / C A R A)] ‥‥ formula 28.

On the basis of R _A and C _A ,

Equation 26] puts the R _A = R _O ‥‥ formula 29 C _A = C _O ‥‥ formula 30. If R and C errors generated in the IC manufacturing process are α and β, respectively, R _B ,
C _B and R _C are given by the following equations.

[0093]

Equation 27] R _B = αR _O ‥‥ formula 31 C _B = βC _O ‥‥ formula 32 R _C = αR _C0 ‥‥ formula 33

Using equations 29 to 33, equation 28 becomes as follows.

[0095]

ΔV = − (V _CC / 2) (R _C0 / R _O ) (γ−1) Equation 34 where αβ = γ

Here,

K = − (V _CC / 2) (R _C0 / R _O ) ‥‥ Equation 35 gives Equation 15.

In FIG. 5, ΔV
= V _C -V ₀ , and the correction voltage ΔV is output. Here, the gain of the differential amplifier 8 is set to 1.

[Brief description of the drawings]

FIG. 1 is a circuit diagram showing a conventional time constant adjusting method in a time constant variable circuit.

FIG. 2 is a circuit diagram showing a conventional method for adjusting a cutoff frequency in an active filter.

FIG. 3 is a circuit diagram showing a configuration of a time constant automatic adjustment circuit according to the present invention.

FIG. 4 is a graph showing a relationship between a set voltage and a cutoff frequency.

FIG. 5 is a circuit diagram illustrating a configuration example of an RC error detection circuit.

FIG. 6 is a circuit diagram showing a specific configuration example of a constant current source in one charge / discharge circuit.

FIG. 7 is a circuit diagram showing a specific configuration example of a current source in the other charge / discharge circuit.

8 is a circuit diagram showing a specific configuration of a constant current source used in the current source in FIG.

FIG. 9 is a waveform diagram showing clock pulses and charge / discharge voltage waveforms.

[Explanation of symbols]

 Reference Signs List 1 RC error detection circuit 2 Amplifier circuit 3 Variable gain amplifier circuit 4 GCA control circuit 5 Control circuit 10 Time constant automatic adjustment circuit 50 Time constant variable circuit

Claims (2)

(57) [Claims] 1. A time constant variable circuit formed on a semiconductor substrate by an IC process, wherein a time constant is variably controlled by an applied control voltage.
An RC error detection circuit for detecting an RC error caused by a variation in the C process and outputting a correction voltage representing the RC error, and an externally set voltage for setting a time constant of the time constant variable circuit is determined by the RC error A setting voltage correction circuit that corrects with a correction voltage output from the detection circuit and outputs the corrected setting voltage, and an RC error when the setting voltage is fixed based on the correction voltage output from the RC error detection circuit. A first control circuit for outputting a first control voltage for correcting a change in a time constant based on the first control voltage and the correction setting voltage and the first control voltage as inputs; A second control circuit for generating a second control voltage as the control voltage and applying the second control voltage to the time constant variable circuit so as to be determined by the set voltage regardless of an RC error; Constant automatic adjusting circuit which includes a. 2. The automatic time-constant adjustment circuit according to claim 1, wherein said time-constant variable circuit is a frequency-variable active filter, and said cut-off frequency is controlled by said set voltage.