JP2001308683A - Gm-C FILTER - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a filter having desired characteristics which has no fluctuation over the entire temperature range as well as at the temperature at which frequency characteristic adjustment is made. SOLUTION: Mismatching ΔVth between the threshold of MOSFET M1 of a mater circuit 50 and that of MOSFET M2 of a slave circuit 51 are eliminated with a variable voltage generating circuit E. This enables adjustment to be made so that the filter cut-out frequency fc is set to an ideal value thus realizing characteristics of a filter of which the fc does not fluctuate over the entire temperature ranges as well as at the temperature for adjustment.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a Gm-C filter.

[0002]

2. Description of the Related Art FIG. 1 is a circuit diagram of a conventional Gm-C filter, wherein 1 is a master circuit constituting a PLL circuit including a Gm-C filter, and 2 is a Gm-C filter having the same configuration as the Gm-C filter. This is a slave circuit composed of filters. The master circuit 1 outputs a control signal (DC signal) phase-locked based on the input reference clock, and
Gm-C filter and Gm-
It is supplied to the C filter as a signal for controlling the filter characteristics.

As shown in FIG. 2, the filter characteristic (actual characteristic) of the slave circuit is such that an error from a target value (ideal value) of the filter characteristic caused by a relative error between the Gm elements of the master circuit and the slave circuit. There is a disadvantage that it is large.

FIG. 3 is a circuit diagram of another conventional Gm-C filter. In order to eliminate the drawback of the filter shown in FIG. 1, a control signal generated from the master circuit 1 is adjusted by an adjusting circuit 3 (FIG. (Fine) Adjustment to bring actual characteristics closer to ideal characteristics. As a result, it is possible to realize a filter which is not affected by variations between Gm elements of the master circuit and the slave circuit. However, even if the variation between the Gm elements can be eliminated by the adjustment circuit 3, the filter characteristics may change depending on the temperature. Then, there is a problem that required performance cannot be obtained.

FIG. 4 is a circuit diagram of another conventional Gm-C filter for solving the drawback of the filter shown in FIG. 3, and 10 is a PLL circuit (including a Gm amplifier and a filter circuit including a capacitor). Master circuit), 11 is Gm
Gm-C filter circuit composed of amplifier and capacitor, 12
Is a current source for temperature correction, 13 is an adder, 14 is a DC current source, and an output line of a reference current generated by a 16 PLL circuit (master circuit) 10.

FIG. 5 is a block diagram of the adder 13 shown in FIG.
2 shows a specific circuit example of a Gm amplifier included in a −C filter circuit 11. In this figure, reference numerals 17 to 21 denote MOSFETs, which constitute the Gm amplifier 26. 27 and 28 are input terminals of the Gm amplifier 26, and 29 and 30 are output terminals of the Gm amplifier 26. Reference numeral 22 denotes a MOSFET, which operates as an adder 13 for adding output currents from the master circuit 10, the temperature correction current source 12, and the DC current source 14, and forms a current mirror circuit together with the MOSFET 21. Then, the addition current is supplied to the MOSFET 21.

The Gm value of the Gm amplifier 26 is determined by the input MOSFE
It is determined by the Gm value of T19 and T20. Therefore, if the current is increased, the Gm value is increased, and if the current is decreased, the Gm value is increased.
The value will decrease.

In FIG. 4, if the temperature correction current source 12, the adder 13, and the DC current source 14 are not provided, the cut-off frequency of the Gm-C filter is changed as shown in FIG. As shown by the curve 40, it increases on the high temperature side and decreases on the low temperature side. Therefore, a current generated from the temperature correction current source 12 such that the output current value decreases on the high temperature side and increases on the low temperature side, and a current generated from the PLL circuit (master circuit) 10 are added by the adder 13. By supplying the generated current to the Gm-C filter circuit 11, the cutoff frequency of the Gm-C filter is higher at the low temperature side than at the high temperature side, so that the cutoff frequency is as shown by the curve 41 in FIG. On the low temperature side, the cutoff frequency is corrected so as to be higher, and as a result, a filter having characteristics with small fluctuations with respect to temperature is obtained. Therefore, the current generated by the DC current source 14 is added to the adder 1
The addition via 3 makes the temperature characteristic of the cut-off frequency approach the desired characteristic as shown by the curve 42.

As described above, the Gm-C filter (FIG. 4)
Can have nearly the desired properties over the entire temperature range.

In the case where the temperature drift amount is not predictable, the value of the above-mentioned correction current is determined after manufacturing the LSI.
It can be adjusted by a technique such as trimming or register writing.

FIG. 7 shows the current source 1 for temperature correction shown in FIG.
2 shows a circuit example. In this figure, 33 is an operational amplifier,
34 is a resistor having a resistance value of R, and 35 to 37 are MOSFE
T and 38 are output current terminals for outputting an output current i2, and 39 is a reference voltage input terminal for inputting a reference voltage _Vref . Here, the output current i2 of the temperature correction current source 12 is represented by the following equation.

I 2 = V _ref / R (1) Here, the resistor 34 is usually formed of polysilicon or a diffusion layer in an LSI, but in any case, the resistor 34 is basically made of a silicon material. These resistors exhibit such a behavior that the resistance value increases on the high temperature side. Therefore, the output current value i2 decreases on the high temperature side and increases on the low temperature side.

By using a current source having such a temperature characteristic, the Gm value of the Gm amplifier 26 shown in FIG. 5 can be relatively reduced on the high temperature side as compared with the master circuit 10.

[0014]

The control lines relating to the master circuit 1 and the slave circuit (Gm-C filter circuit) 2 shown in FIG. 1 are specifically input to the respective circuits as shown in FIG. . That is, the control signal (Vgs) generated by the master circuit is applied to the input M which is a part of the master circuit 1.
It is applied to the gate of the OSFET (M1) and the gate of the input MOSFET (M2) which is a part of the slave circuit 2, respectively. When the control voltage Vgs is two MOSFETs (M1,
M2), currents I1 and I2 responsive to the control voltage Vgs flow through M1 and M2, respectively. These currents I1 and I2 are used as bias currents for gm amplifiers in the master circuit 10 and the slave circuit 11.

The relation between the bias current and the Gm value of the gm amplifier is obtained by the following equation.

[0016]

(Equation 1)

[0017]

(Equation 2)

W and L are channel widths of the input MOSFET,
Channel length, μ is carrier mobility, _cox is gate capacitance per unit area, I is input MOS of Gm amplifier
This is the input current of the FET. I = K (Vgs−Vth) ² (4) (Vth is the threshold voltage of the input MOSFET) When (4) is substituted into (2), Gm = 2K (Vgs−Vth) (5) That is, the Gm value is It can be seen that it depends on (Vgs-Vth).

Further, in order to make the following calculation easy to understand, the sizes of the MOSFETs (M1, M2) are assumed to be the same. As a result, the currents I1 and I2 also have the same value.

Here, there is a mismatch of Vth in the MOSFET M2 on the slave side (Vth of MOSFET (M1, M2)).
The difference between Δth and ΔVth is calculated, and the current values of M1 and M2 are calculated.

I1 = K (Vgs−Vth) ² (6) I2 = K (Vgs−Vth + ΔVth) ² = K (Vgs−Vth) ² + 2KΔVth (Vgs−Vth) + KΔVth ² (7) When there is no Vth mismatch, The current ratio I2 / I1 is 1, but when there is a mismatch, the following is obtained.

[0022]

(Equation 3)

From equation (2), the Gm value ratio, that is, the ratio of Gm1 of the Gm amplifier in the master circuit to that of the slave circuit (Gm2) is

[0024]

(Equation 4)

The Gm-C filter is controlled so that the Gm value is always constant. Since the K value in equation (5) has a temperature change, Vgs-Vth also has a temperature change,
In the range of -10 ° C to 80 ° C, it fluctuates about twice. As a result, from equation (9), the Gm ratio changes with temperature.

Since the cutoff frequency of the slave circuit (Gm-C filter) is given by gm / c, when there is no threshold mismatch ΔVth as shown in FIG.
(As is clear from equation (9), the Gm ratio is constant with respect to temperature fluctuation.) Although the cutoff frequency remains at the ideal value f ₀ , it does not fluctuate with temperature as shown by the dotted line A. If there is a mismatch, the cutoff frequency varies with temperature as shown by the solid line B.

Also, in the circuit of FIG.
However, even if the temperature characteristic is adapted to the ideal characteristic, a temperature characteristic occurs. The reason is as follows.

The control lines of the master circuit and the slave circuit are as shown in FIG. 8, and M is a part of the circuit on the slave circuit side.
The currents I1 and I2 flowing through the MOSFETs M1 and M2 when the threshold mismatch ΔVth is given to the OSFET are as follows.

I1 = K (Vgs−Vth) ² (10) I2 = K (Vgs−Vth + ΔVth) ² (11) When the current mismatch is adjusted by the adjustment circuit 3, the original current value is constant. Is multiplied by the coefficient A.

This is expressed by the following equation.

I2 ′ = A {K (Vgs−Vth + ΔVth) ² } (12) The ratio of the current I1 on the master circuit side to the current I2 on the slave side after adjustment is as follows.

[0032]

(Equation 5)

That is, even if A is adjusted so that I2 'coincides with I1, the temperature fluctuates (Vgs-Vt).
When h) changes, I2 '/ I1 changes. As a result, the problem that the frequency characteristic of the filter circuit shifts depending on the temperature cannot be solved.

On the other hand, the conventional Gm-C filter shown in FIG. 4 is effective when the average value of the temperature drift amount for each solid is not zero and the variation is small.

However, the amount of temperature drift varies greatly depending on the solid, and the conventional Gm-C filter shown in FIG. 4 has the following problems.

That is, specifically, the calculation is performed for the case where Vgs-Vth = 400 mV (at room temperature, 25 ° C.) ΔVth = 10 mV.

When the MOSFET is heated at a high temperature, the mobility of the carrier is deteriorated, and the current for compensating for the deterioration is increased. For this reason, Vgs−Vth = about 550 mV (80 ° C.). As a result, the ratio of the gm values at 25 ° C. and 80 ° C. (Equation 9) is: gm2 / gm1 = 1 + 10/400 = 1.0 at 25 ° C.
25 gm2 / gm1 = 1 + 10/550 = 1.0 at 80 ° C.
The difference between 18.25 ° C. and 80 ° C. is 0.007, and a temperature drift amount of about 0.7% is expected. Therefore, when the cutoff frequency accuracy of the filter needs to be higher than this, there is a problem that this method (FIG. 4) cannot cope with it.

Therefore, an object of the present invention is to provide a Gm-C filter which has solved the above problems.

[0039]

According to the first aspect of the present invention, the Gm
A master circuit including a first Gm-C filter circuit including an amplifier and a capacitor, generating and outputting a filter adjustment reference signal for self-adjusting the frequency characteristics of the filter circuit, and a second circuit including a Gm amplifier and a capacitor Gm-
A slave circuit having a C filter circuit and exhibiting filter characteristics according to the filter adjustment reference signal; and a first and second Gm by adding a DC signal to the filter adjustment reference signal output from the master circuit. -C filter circuit to provide the second G
variable voltage generating means for controlling the frequency characteristic of the mC filter circuit.

According to a second aspect of the present invention, in the first aspect, the variable voltage generating means includes the first and second Gm-C
A DC signal is added to the filter adjustment reference signal so as to eliminate a mismatch between input means of the filter adjustment reference signal of the filter circuit.

According to a third aspect of the present invention, in the first or second aspect, the variable voltage generating means generates a temperature-compensated DC signal.

[0042]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 10 is a circuit diagram of a Gm-C filter, 50 is a master circuit constituting a PLL circuit including a Gm-C filter, and 51 is a Gm-C filter having the same configuration as the above-described Gm-C filter. This is a slave circuit composed of filters. Although details are shown in FIG. 13, the master circuit 50 outputs a control signal (DC signal) phase-locked based on the input reference clock, and a Gm-C filter in the same circuit 50 and a slave circuit 51 in the slave circuit 51. It is supplied to the Gm-C filter as a signal for controlling filter characteristics. E denotes a variable voltage generating circuit for adjusting the frequency characteristic of the filter to an ideal value (desired value), whereby the control voltage generated by the master circuit 50 is changed by the variable voltage generating circuit E.
Is added to the slave circuit 51 as a new control voltage. Assuming that the Gm value of the Gm amplifier in the slave circuit 51 is set to be increased by increasing the control voltage, the cutoff frequency of the filter (hereinafter, referred to as the filter voltage is increased by increasing the voltage of the variable voltage generation circuit E. fc) is adjusted in the higher direction, and the fc of the filter is adjusted in the lower direction by lowering the voltage of the variable voltage generation circuit E.

As a specific adjustment method, fc of the filter
Is smaller than the ideal value, the value of the current flowing through the Gm amplifier on the slave side is smaller than the value of the current flowing through the Gm amplifier on the master side. The voltage is adjusted so that the fc of the filter finally approaches the ideal value. In this way, not only a desired fc value can be obtained at the temperature at the time of adjustment, but also a filter in which the fc value does not fluctuate over the entire temperature range (the reason will be described later).

FIG. 11 shows another embodiment of the present invention. FIG.
10 is the same as FIG. 10 except that the variable voltage generation circuit E is provided on the master circuit unit side, and the operation and effects are the same as those in FIG.

A more specific circuit will be described with reference to FIG. 151 enclosed by a dotted line is a Gm-C filter, 152 is a phase comparator, 153 is a low-pass filter (LPF), and 154 and 155 are comparators.
A master circuit (PLL) from 1 to 155 as a whole is configured. The Gm-C filter 151 has a Gm of 161-164.
It consists of an amplifier and 165 and 166 capacitors. FIG.
The circuit diagram of the Gm amplifier is shown in FIG.

In FIG. 13, Gm amplifiers 161 to 16
The Gm-C filter 151 having 4 and capacitances 165 and 166 has a low-pass filter characteristic when the input terminal is 167 and the output terminal is 168, and has a phase shift of 0 ° in a low band and a phase shift in a high band. It has a phase characteristic in which the shift is 180 ° and the phase shift is 90 ° at the cutoff frequency fc.

That is, when the frequency of the input signal matches the cutoff frequency fc, the filter input signal and the filter output signal are output from the comparators 155 and 1
54, and further through an exclusive OR circuit (EXOR) functioning as a phase comparator 152, the frequency is twice that of the input signal, and the periods of the high level logic and the low level logic become equal. The output signal has a so-called duty ratio of 50%. At this time, the signal output from the phase comparator 152 is converted to a low-pass filter (LP
F), the integrator 1 also passes through the integrator 153.
There is no change in the DC output level of 53, and a phase locked state can be realized.

When the cutoff frequency of the Gm-C filter is smaller than the design value fc, the phase delay becomes larger than the design value (90 °). As a result, the output signal of the phase comparator 152 operates in the direction of lowering the output level of the integrator 153 because the high-level logic period is shorter than the low-level logic period. The bias voltage generated when the output level of the integrator 153 decreases is
It works so that the Gm values of all the Gm amplifiers 161 to 164 increase. In particular, since the Gm values of the Gm amplifiers 162 and 163 determine the cutoff frequency of the Gm-C filter 151, the cutoff frequency increases as the Gm value increases.

Thus, the output level of the integrator 153 shifts in the direction in which the cutoff frequency of the filter 151 becomes equal to the design value, and when the duty ratio of the output signal of the phase comparator 152 finally becomes 50%, That is, when the cutoff frequency of the filter becomes equal to the design value,
The output of the integrator 153 falls to a certain level. G
When the cutoff frequency of the m-C filter 151 is higher than the design value, the same operation is performed, and finally, the Gm-C
The cutoff frequency of the filter 151 becomes equal to the design value, and the output of the integrator 153 falls to a certain level.

On the other hand, the circuit configuration of the Gm-C filter constituting the slave circuit is exactly the same as that of the low-pass Gm-C filter 151 used in the PLL circuit 156, and is used there. If the Gm value and the capacitance value of the used Gm amplifier are also the same, the characteristics of both filters become the same.

On the other hand, the control voltage generated by the master circuit is passed through a control line 22 to a MOSFET which functions as a current source.
Apply to the gate of T21. A Gm amplifier having the same configuration is used as the Gm amplifier in the slave circuit. Similarly, a control voltage generated by the master circuit is applied to the gate of the MOSFET functioning as a current source. FIG. 12 shows a circuit in which only MOSFETs functioning as these current sources are described.

In FIG. 12, MOSFETs M1 and M2
12 are the same as those in FIG. 10 when they are considered to be part of the master circuit and the slave circuit, respectively. As described above, the MOSFETs M1, M2
If there is a threshold mismatch ΔVth between them, a current mismatch occurs, resulting in a cutoff frequency fc shift. Here, the variable voltage generation circuit E can adjust the fc of the filter to be an ideal value. That is, the mismatch ΔVt is generated by the variable voltage generation circuit E.
h can be eliminated. As a result, the filter fc can be adjusted to an ideal value, and a filter characteristic in which fc does not fluctuate not only over the temperature at the time of adjustment but also over the entire temperature range can be obtained.

When the master circuit and the slave circuit each have only one Gm amplifier, it can be understood that the circuit shown in FIG. 12 can be applied. 4 for the Gm-C filter of the master circuit
Gm amplifiers are used. In general, the order of the Gm-C filters on the slave side is arbitrary, and the number of Gm amplifiers is generally four or more. FIG. 15 is a circuit diagram in which the Gm value of each Gm amplifier can be adjusted in that case. In the circuit of FIG. 15, the MOSFETs (M1-1, M2-2,... M2-
N) independently applies the control voltage from each of the variable voltage generation circuits E1 to EN, so that accurate adjustment can be performed. In this case as well, the characteristic does not change due to the temperature change. However, in this case, there are some omission methods because the circuit scale becomes large.

For example, when the cut-off frequency is important and there is no problem even if the characteristic curve slightly deviates from ideal, the adjustment may be made to one place as shown in FIG. 16 (that is, one variable voltage generating circuit E), or It is also possible to group several variable voltage generation circuits by grouping several Gm amplifiers such as off frequency and Q value control. Of course, the variable voltage generation circuit of FIG. 15 can be controlled by being divided into several subgroups.

Here, a case where the cutoff frequency is important will be described as an example. For example, in the case of a second-order filter as in the Gm-C filter 151 in FIG. 13, the cutoff frequency of the filter is proportional to (Gm1 · Gm2) ^0.5 . However, Gm1 and Gm2 are Gm amplifiers 162 and 1
63 are the respective Gm values. In other words, it can be said that the cutoff frequency is proportional to the average value of the specific Gm value.

Therefore, there is a method of adjusting with only one variable voltage generating circuit as in the circuit of FIG. In FIG. 16, when there is an average mismatch error between the current source A of the master circuit and the current source B of the slave circuit, the Gm value of the Gm amplifier also shifts, resulting in the shift of the cutoff frequency.

By adjusting the amount of this deviation by the variable voltage generation circuit E in FIG. 16, the average deviation can be corrected, and the cutoff frequency can be adjusted to that of the ideal filter.

As long as the variation of the current sources is not extremely large, a desired characteristic can be sufficiently achieved by correcting the average value using a circuit as shown in FIG.

Next, another example of the variable voltage generating circuit will be described. FIG. 17 is a circuit diagram of another variable voltage generation circuit E ′. As shown in FIG. 17, the variable voltage generation circuit E ′
Is a resistor R which is a resistive element connected between both gates of the MOSFETs (M1, M2), and current sources 60, which are respectively connected to both ends of the resistor R and flow currents in opposite directions to the resistor R. The voltage generated across the resistor R by the current from the current sources 60 and 61 does not fluctuate with temperature.

In this other variable voltage generating circuit E ′, the magnitude of the current value of the current source 60 and the current source 61 is changed, or the direction of each current is reversed, so that the potential difference between both ends of the resistor R is changed. Can be changed, whereby the potential between the MOSFETs (M1, M2) can be appropriately adjusted to cancel the mismatch voltage ΔVth.

Next, a specific configuration of the current sources 60 and 61 shown in FIG. 17 will be described with reference to FIG. As shown in FIG. 18A, the current source 61
An operational amplifier 62 forming a follower; a resistor R0, which is a resistor element and made of the same material as the resistor R;
T (Q1, Q2).

The operational amplifier 62 has a negative input terminal to which the reference voltage Vref is applied, and an output terminal connected to each gate of the MOSFET (Q1, Q2) to form a current mirror. MOSFET Q1
Is connected to a voltage source _VDD , and its drain is grounded via a resistor R0. Its drain and resistor R0
Are connected to the + input terminal of the operational amplifier 62. MOSFET Q2 has a voltage source V
_DD , the drain of which is the output terminal and the resistor R
And the gate of the MOSFET (M2).

As shown in FIG. 18B, the current source 60 includes an operational amplifier 63 constituting a voltage follower,
A resistor R0 which is a resistor element and is made of the same material as the resistor R;
P-type MOSFET (Q3, Q4) and N-type MOSFET
ET (Q5, Q6).

The operational amplifier 63 has a negative input terminal to which the reference voltage Vref is applied, and an output terminal connected to each gate of the MOSFETs (Q3, Q4) to form a current mirror. MOSFET Q3
Is connected to a voltage source _VDD , and its drain is grounded via a resistor R0. Its drain and resistor R0
Are connected to the + input terminal of the operational amplifier 63. MOSFET Q4 has a voltage source V
Is connected to the _DD, the drain is connected to the drain of the NMOSFETQ5. NMOSFETs Q5 and Q6 constitute a current mirror, and NMOSFET Q6
The drain of (60) becomes the output terminal, and the resistance R and MOS
It is connected to the gate of the FET (M1).

Here, the operational amplifiers 62 and 63 and the resistor R0
Constitute voltage-current conversion means, and MOSFETs Q2, Q
6 and the like constitute current generating means. In the current sources 60 and 61 having such a configuration, the voltage applied to the resistor R0 becomes the reference voltage Vref applied to the minus input terminals of the operational amplifiers 62 and 63 by the operation of the operational amplifiers 62 and 63 formed of voltage followers. Is controlled as follows. For this reason,
The current iref flowing through the resistor R0 is expressed by the following equation (14).

Iref = Vref / R0 (14) The current iref1 flowing through the MOSFETs Q2 and Q6
Becomes equal to the current iref flowing through the resistor R0. Therefore, the current iref1 can be controlled by controlling the reference voltage Vref, and the current iref1 is controlled by the current source 6 shown in FIG.
When used as the current IB of 0,61, the current sources 60,6
1 can be controlled by the reference power supply Vref.

When the resistor R0 is made of the same material as the resistor R in FIG. 17, its absolute value varies depending on the forming conditions at the time of manufacture, but its resistance ratio R / R0 is constant irrespective of the manufacturing process and temperature. Becomes Also, the current iref
1 is iref1 = iref when the sizes of the MOSFETs (Q1, Q2) are the same, and the MOSFET Q2
Is larger than the transistor size of MOSFET Q1, Aref1 = A × iref
Becomes

Therefore, when the current iref1 is used for the current IB, the voltage VB generated across the resistor R by the current IB is expressed by the following equation (15) with reference to the equation (14).

VB = IB × R = A × (R / R0) × Vref (14) The current IB is determined by the MOSFET (Q1, Q2) or the MOSFET
Since the transistor size ratio of TQ3, Q4 and Q5, Q6 can be changed, the size can be changed by changing the size ratio. As a result, the voltage VB can be changed. As can be seen from equation (14), since the resistance ratio R / R0 is constant irrespective of the temperature, a DC signal independent of temperature can be achieved by the circuits of FIGS.

The configuration of the current variable circuit for varying the current IB will be described with reference to FIG. In order to change the size of the MOSFET Q2 in FIG. 18, the current variable circuit includes a plurality of MOSFETs Q2 and Q6 having different sizes as shown in FIGS.
SFETs Q7 to Q10, Q11 to Q14, and these MOSFETs are connected to switches SW1 to SW4, SW5 to SW5.
The current IB is varied by selecting the switch SW8 and varying the current iref1.

More specifically, MOSFETs Q7 to Q7
10 is a P-type, and MOSFETs Q11 to Q14 are an N-type. For example, each size has a power-of-two relationship such as 1, 2, 4, and 8. In addition, MOSFETQ
7 to Q10 are connected to the output terminal of the operational amplifier 62, and their sources are connected to the voltage source _VDD . Further, the drains of the MOSFETs Q3 to Q6
It is connected to the resistor R via the switches SW1 to SW4.
Similarly, the sources of the MOSFETs Q11 to Q14 are grounded, and the drains are connected to resistors R through switches SW5 to SW8.
, And the gate is connected to the gate of MOSFET Q5.

In the current variable circuit having such a configuration, the switches SW1 to SW connected to the MOSFET
4, SW5 to SW8, the current IB becomes 0 to 1
Up to five 16-step adjustments are possible. Switch S
Switching of W1 to SW4, SW5 to SW8 is performed by the switch S
The on / off states of W1 to SW4, SW5 to SW8 are stored in advance in, for example, a register, and the switches SW1 to SW4, SW5 to SW8 are turned on and off, thereby adjusting the voltage VB by varying the current IB. it can.

If the current sources 60 and 61 are switched and connected, the direction of the generated voltage can be changed.

[0074]

As described above, according to the present invention, a DC signal is added to the filter adjustment reference signal output from the master circuit and supplied to one of the first and second Gm-C filter circuits. Then, by controlling the frequency characteristic of the second Gm-C filter circuit, it is possible to provide a filter having desired characteristics without temperature fluctuation not only in the temperature at the time of adjustment but also in all temperature ranges. In addition, the present invention can adjust the temperature performance only by performing the adjustment at an arbitrary temperature, so that the present invention has not only a time and economic effect, but also is extremely effective for, for example, an LSI shipping inspection.

[0075]

[Brief description of the drawings]

FIG. 1 is a circuit diagram of a conventional Gm-C filter.

FIG. 2 is a diagram showing filter characteristics of the Gm-C filter.

FIG. 3 is a circuit diagram of another conventional Gm-C filter.

FIG. 4 is a circuit diagram of still another conventional Gm-C filter.

FIG. 5 is a diagram showing some details of the circuit of FIG. 4;

FIG. 6 is a diagram showing a cut-off frequency characteristic with respect to a temperature of a Gm-C filter.

FIG. 7 is a circuit diagram of a current source for temperature correction.

FIG. 8 is a diagram showing some details of the circuit of FIG. 1;

FIG. 9 is a diagram illustrating cutoff frequency characteristics.

FIG. 10 is a circuit diagram of an embodiment of the present invention.

FIG. 11 is a circuit diagram of another embodiment of the present invention.

FIG. 12 is a diagram showing a part of FIG. 10 in detail.

FIG. 13 is a diagram showing details of a master circuit.

FIG. 14 is a circuit diagram of a Gm amplifier.

FIG. 15 is a circuit diagram of still another embodiment of the present invention.

FIG. 16 is a circuit diagram of still another embodiment of the present invention.

FIG. 17 is a diagram illustrating a variable voltage generation circuit.

FIGS. 18A and 18B are circuit diagrams each showing a specific example of a current source.

FIGS. 19A and 19B are diagrams showing current variable circuits for varying a current IB, respectively.

[Explanation of symbols]

 50 Master circuit 51 Slave circuit E Variable voltage generation circuit

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5J098 AA03 AA11 AB02 AB03 AB07 AB08 AB11 AB15 AB16 AC02 AC21 AD06 AD18 CA02 CB01 CB08

Claims (3)

[Claims] 1. A first G comprising a Gm amplifier and a capacitor.
a master circuit for generating and outputting a filter adjustment reference signal for self-adjusting the frequency characteristic of the filter circuit; and a second Gm-C filter circuit including a Gm amplifier and a capacitor. A slave circuit exhibiting a filter characteristic according to the filter adjustment reference signal; and a first and second Gm-C by adding a DC signal to the filter adjustment reference signal output from the master circuit.
A variable voltage generating means for controlling a frequency characteristic of the second Gm-C filter circuit by supplying the voltage to one of the filter circuits. 2. The method according to claim 1, wherein said variable voltage generating means includes said first and second Gm-
A DC signal is added to the filter adjustment reference signal so as to eliminate a mismatch between input means of the filter adjustment reference signal of the C filter circuit.
C filter. 3. The Gm-C filter according to claim 1, wherein the variable voltage generator generates a temperature-compensated DC signal.