JPH0454012A - Low pass filter circuit - Google Patents

Abstract

PURPOSE:To form a resistance deciding a filter constant in the inside of an LSI by providing a load resistor comprising plural resistive elements to an output terminal of an amplifier and forming a low pass filter circuit with a switch element and a capacitor connecting to a power supply. CONSTITUTION:With a switch S1 closed, a load circuit L consists of a parallel circuit comprising a resistive element R1 and a capacitor C and its cut-off frequency f1 is expressed in f1=1/2piR1C. Moreover, suppose that the resistance of resistive elements R2-Rn is equal to the resistance R1, the cut-off frequency f2 with the switch S1 opened and a switch S2 closed is expressed in f2=1/4piR0C, which is a half of the cut-off frequency when the S1 is closed. The cut-off frequency of the filter circuit is changed by having only to select a resistive element in use through the switch. When the filter circuit is applied to an LSI filter circuit, it is not required to employ an externally mounted resistive element.

Description

[Detailed Description of the Invention] (Summary) This invention relates to a low-pass filter circuit whose cut-off frequency can be adjusted, particularly a low-pass filter circuit suitable for application to an analog LSI. The purpose is to provide a filter circuit that makes it possible to configure a resistor that determines a constant inside an LSI, and has a load resistor whose l end is connected to the output end of an amplifier and which is divided into multiple resistor elements. A low-pass filter circuit was constituted by a switch element that selectively connects the connection point between these resistance elements to a power supply, and a capacitor connected between the output terminal of the amplifier and the power supply.

[Industrial application field]

The present invention relates to a low-pass filter circuit whose cut-off frequency can be adjusted, and particularly to a low-pass filter circuit suitable for application to analog LSI.

[Conventional technology]

In order to reduce the size of computer peripheral devices, the analog circuits used are also becoming LSI, but when configuring the analog circuit as an LSI, the absolute accuracy of the resistor configured on the semiconductor substrate is low, but the relative accuracy is low. is relatively good, so the circuit is designed such that the performance of the constructed circuit depends only on the relative accuracy of this resistor.

[Problem to be solved by the invention]

However, in an RC filter circuit, its characteristics are determined by the absolute accuracy of the resistor and capacitor that determine the filter constant, so in LSIs where it is difficult to obtain absolute accuracy of the resistance value, this resistor cannot be configured inside the LSI. First, because this resistor is mounted externally, it has not been possible to sufficiently meet the demand for miniaturization.

SUMMARY OF THE INVENTION An object of the present invention is to provide a filter circuit that allows the cutoff frequency of a filter to be adjusted and a resistor that determines a filter constant to be configured inside an LSI.

[Means to solve the problem]

Referring to the principle diagram in FIG. 1, one end is connected to the output end to the amplifier, and a plurality of resistive elements R
a switch element S,, S2, which has a load resistance divided into 2,---R- and selectively connects the connection point between these resistance elements to the supply voltage V; -S.

A low-pass filter circuit was constructed by providing a capacitor C connected between the output terminal of the amplifier A and the power supply.

[For production]

The cutoff frequency f of amplifier A is determined by the resistance element R1 in the load circuit.
, Rt, ......If the values of the combined resistance of R1 and the capacitor C are R and C, respectively, and the internal resistance of the transistor T of amplifier A and the resistance of its emitter circuit are ignored, the following is obtained. It becomes as shown in Eq.

f = 1/2πRC Therefore, when switch S in Fig. 1 is closed, the parallel circuit of resistor R and capacitor C constitutes a load circuit, and its cutoff frequency f is f. , = 1/2πR+C.

Also, assuming that the resistance value of resistance element Rt,...R1 is equal to the value of this resistance element R1, the cutoff frequency when switch S1 is open and switch S2 is closed. f2 is f2=1/2π(Ro + R+) C=l/4πR
oC, which is 1/2 of the cutoff frequency when the switch S1 is closed.

The filter constant determined by the values of the resistor R3 and capacitor C is set to 1/(1+
If you select a value that gives the maximum resistance error rate), the LS
When the error of the resistance element R1 configured in I is the maximum in the positive direction, by closing the switch S1 as described above, the resistance element R1 is used as the resistance that determines the filter constant.
1 is used to obtain the required cutoff frequency.

Note that this maximum resistance error rate indicates the ratio of the maximum error expected when generating a resistance element configured in an LSI, and the error in the direction where the resistance value increases is called the positive direction error. An error in the direction in which the value decreases is called a negative error.

On the other hand, if all switches S, , S, . , L
The combined resistance of these resistance elements R9 to R1, that is, RI
Make sure to select a value of Rt+...10R.

That is, when the resistance element R configured in the LSI has a maximum error in the positive direction, only this resistance element R1 is used, and when the resistance element R1 configured in the LSI has the maximum error in the negative direction, the resistance element R1 is used. If an error occurs in the resistance element R3 configured in the LSI by using the combined resistance of ''-R-, that is, R, +R2+...+R, as the resistance value that determines the filter constant. However, the required filter constant can be obtained.

Note that, as is clear from the above, the resistance element R
2 to R1 are resistance elements for compensating for errors occurring in resistance element R3, and the number by which the combined resistance R2+ +K of these resistance elements R2 to R1 is divided, that is, the number of resistance elements R2 to R1. The number can be increased to increase the accuracy of determining the filter constant, and the number can be decreased if lower accuracy is sufficient.

〔Example〕

FIG. 2 is a diagram showing an embodiment of a filter circuit in which the present invention is applied to a differential amplifier configured on an LSI, as shown by the two-dot chain line.

This filter circuit includes a constant current circuit unit F, an amplifier unit A having an input terminal and an output terminal, a gain setting resistor unit G for determining the gain of this amplifier A, and a filter constant circuit unit including a capacitor for determining a filter constant. C1 filter constant correction resistance units R, , R, , including a pair of resistance elements each having an equal value;
-1...R1-1R1 and switch unit S,
, S2. ...S, -2S9.

Two transistors Ta of amplifier unit A.

T a 2 constitutes a differential amplifier including the resistor Rg of the gain setting resistor unit G and the constant current circuit unit F, and these transistors T a + and T a 2
The above filter constant correction resistance unit R+ serves as a load.
, R2,--...R,-, ,R, a pair of resistive elements, for example, the resistors rl of units R4 and R2,
r2 and rl, , r22 are the resistors R3 in FIG.
Each corresponds to R7, ......,
Combined resistance value of these resistors and filter constant circuit unit C
The cutoff frequency, that is, the filter constant, is determined by the capacitor.

Specifically, when the required cutoff frequency of the filter circuit is F, find the resistance value R such that F = 1/ (2πX2CxR), and when the maximum resistance error rate is K, a pair of load resistance elements rl+,
The value r of each r2+ is set so that r + =R/ (1+K).

In addition, filter constant correction resistance units Rl, R2--
-R, resistance element r 1,, r l t, -r
l or r 2 +, r 2 L ---r
2, the combined resistance value is ', that is, K' = r + + r 2-1--+ r =
is set so that K' = R/ (1-K).

When the resistance error is maximum in the positive direction, if a low potential "L" is applied to the select input terminal SWl of the LSI and a high potential "H" is applied to the other select input terminals SW, ~SW, two of the switch units S, two transistors TsL, Ts2
+ conducts and transfers the current from the power supply Vcc to the unit R,
resistance r]1. r2+ to the collectors of the two transistors T a , Ta, of the amplifier unit A, and the other switch units St, . . .
-L s , each tiger in the register Ts1. ~Ts1.
．． Ts2+~Ts2. is in a cut-off state, so other filter constant correction resistance units Rt, ”-・-・-R, −+
, R- are separated.

As a result, the amplifier unit A uses the resistance elements rl+, r2 in the filter constant correction resistance unit R0 as load resistances at the input terminals 1. , I, and performs differential amplification of the signals from output terminals 0. ．． A differential signal is output to 02, and the cutoff frequency f1 of amplifier A at this time is the same as this filter constant correction resistance unit R.

L = 1/ (2πX2CXr+ ) = 1/ (2π
x2CxR)=F, and the required cutoff frequency of the filter circuit can be obtained.

In addition, a low potential “L” is applied to the select input terminal SW of the LSI.
” to the other select input terminals SW, ~SW,.

When a high potential "H" is applied to the switch unit S,
The two transistors Ts1. ．． T s2. conducts to transfer the current from the power supply Vcc to the resistive elements rl, ~rl, and r2 . of the units R1 to R1. ~r2, the two transistors Ta,,Ta, of amplifier unit A;
Since the amplifier unit) A is supplied to the collector of the filter constant correction resistor units R5 to R1, the resistive elements rl,
~r1. and r2, ~r2. as the load resistance at input terminal 1. , I2 are differentially amplified and the signals from output terminals 0 . ．．
A differential signal is output to 02.

At this time, the cutoff frequency f of the amplifier A is determined by the resistance value r of the above equation (2), and the filter constant correction resistance units R1 to R.
1, the resistance elements r12 to r1. Or r22~r2.
The combined resistance value r2+-...-+r.

r + + r 2 + −−−−+ r a
Therefore, the cutoff frequency f5 is f. ==1/(2πx2CX (r+ +rt +-+
r, )) = 1/ (2πX2CXR) = F, and the required cutoff frequency of the filter circuit can be obtained.

As is clear from the above, the selection input terminals SW2 to SW of the intermediate switch units 82 to S and -4
When "L' potential is applied to one SW of , , , the resistance element rl+ in the filter constant correction resistance units R3 to R1
~rl, and r2. It goes without saying that ~r2° operates as a resistance value that determines the filter constant.

In this way, the filter constant correction resistance unit R.

When the resistance value of resistance element r1 has the maximum error in the positive direction, only the resistance element of this unit R1 is used by applying an "L" potential only to the select input terminal SW1, and the resistance value of this resistance element r has an error in the positive or negative direction, the filter constant correction resistor unit R
3 to an appropriate number of units of resistance elements are the unit R1.
The select input terminal SW is connected in series with the resistive element.

A desired cutoff frequency can be obtained by applying an "L" potential to any one of SW.

〔Effect of the invention〕

According to the present invention, the cut-off frequency of the filter circuit can be changed simply by selecting the resistor element to be used with a switch, and especially when applied to an LSI filter circuit, the required frequency can be changed without using an external resistor element. The particular effect achieved is that it is possible to obtain a cut-off frequency of .

[Brief explanation of drawings]

FIG. 1 is a diagram showing the principle of the present invention, and FIG. 2 is a diagram showing an embodiment in which the present invention is applied to a differential amplifier configured as an LSI. Patent applicant Fujitsu Limited Principle diagram Figure 1

Claims (1)

[Claims] One end is connected to the output end of amplifier A, and a plurality of resistive elements R_0, R_1, R_2, . . .
switch elements S_1, S_2, ......S_m having load resistances divided into R_m and selectively connecting the connection points between these resistance elements to the supply voltage V; and the output of the amplifier A; A low-pass filter circuit comprising a capacitor C connected between the end and the power supply.