JPH0128526B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an operational amplifier circuit that is capable of highly accurate gain control, is suitable for programmable gain amplifiers, multiplexers, demultiplexers, etc., and is easily monolithic.

An operational amplifier (hereinafter referred to as an operational amplifier) can control the gain arbitrarily by adding an appropriate feedback circuit, so by combining it with a switch, it can be used as a programmable gain amplifier.
It can be used as a multiplexer, demultiplexer, etc., and has therefore been widely used in various fields.

An example of such a programmable gain amplifier is shown in FIG.

In the figure, 1 is an operational amplifier, 2 is a decoder,
3 to 6 are switch elements consisting of N-channel MOSFETs, 7 is an analog input signal terminal, 8 is an output signal terminal, 9 to 12 are decode output lines, 13,
14 is decode input line, 15 is operational amplifier 1
16 is the same non-inverting input, 17-2
0 is an input resistance, and 21 is a feedback resistance.

In the amplifier circuit in which the input resistors 17 to 20 and the feedback resistor 4 are connected in this way, as long as the conditions that the input impedance and gain of the operational amplifier 1 are sufficiently large and its output impedance is sufficiently small are satisfied, It has the characteristic that its gain is uniquely determined only by the resistance values of the input resistance and feedback resistance.

So, now, the resistance value of resistor 17 is 10kΩ, resistor 18 is 50kΩ, resistor 19 is 100kΩ, and resistor 20 is
500kΩ and resistor 21 is 1MΩ. If "0" is supplied to the decode input lines 13 and 14 as gain setting signals, only the Q0 output of the decoder 2 becomes "1" and the other output Q1
~Q3 remains at "0", so only switch element 3 is turned on.

Therefore, only the resistor 17 is inserted between the input terminal 7 and the inverting input 15, and it operates as an operational amplifier circuit with an input resistance of 10 kΩ and a feedback resistance of 1 MΩ, so the gain at this time is A _NF is A _NF = -R _f /R _s = 1 [MΩ] / 10 [kΩ] = 100, where R _f : feedback resistance value R _s : input resistance value. That is, the input signal supplied to the input terminal 7 is multiplied by 100 and output to the output terminal 8.

In this way, the gain setting signals for the decode inputs 13 and 14 are set to [0, 0] in addition to the above [0, 0].
1], [1, 0], [1, 1], each time at [0, 1], only the decode output Q1 becomes "1", the switch element 4 is turned on, and the input resistance R _s becomes 50 kΩ, and a gain A _NF =50 is obtained, and subsequently the gain can be set to 10 times or 2 times in the same manner, making it possible to easily configure a programmable gain amplifier circuit. In addition, related devices of this type include, for example, Japanese Patent Application Laid-Open No.
The disclosure of Publication No.-25209 can be mentioned.

Therefore, conventionally, such operational amplifier circuits have been widely employed in programmable gain amplifiers and the like.

By the way, as operational amplifiers used in such circuits, MOS monolithic ones are mainly used, and it is becoming possible to obtain ones with sufficient performance at low cost. Therefore, the influence of variations in the characteristics of the operational amplifier on the gain A _NF when used as an operational amplifier circuit can be almost ignored, and the accuracy of such an operational amplifier circuit is mostly determined by the accuracy of the input resistance R _s and the feedback resistance R _f It's starting to be done. For example, open loop gain A is 80bB
When using an operational amplifier with a gain A _NF = 10, the error caused by the finite gain A of the operational amplifier is only 0.1%, and even if this gain A changes by ±20%, the gain A _NF The resulting fluctuation is only 0.02%.

Therefore, increasing the accuracy of the gain setting of such an operational amplifier circuit depends mostly on increasing the accuracy of the resistors used as input resistors and feedback resistors and improving their tracking characteristics.

However, increasing the precision of discrete resistors leads to a significant increase in cost, and tracking characteristics cannot be improved even if high-cost metal film resistors are used.
Only about 100 to 50 ppm/℃ can be expected. Also,
In recent years, ICs have been known that have multiple thin film resistors on-chip to improve their tracking characteristics, but their tracking characteristics are only 10 ppm/°C at most, which further increases costs.

Apart from these, in the operational amplifier circuit shown in Fig. 1, a switch element 3 consisting of a MOSFET is
Since an on-resistance of ~6 is inserted in series with the input resistors 17-20, this on-resistance causes an error and makes it difficult to achieve high accuracy. In order to solve this problem, the on-resistance of the switch elements 17 to 20 must be made low, or the resistors 17 to 2 must be made low.
It is necessary to design a large resistance value at 0, but in order to reduce the on-resistance of switch elements 17 to 20, the channel ratio W/L of the MOSFET must be increased, which increases the chip size and makes it monolithic. is in difficulty. For example, in a practical monolithic LSI, the on-resistance of a MOSFET switch element reaches several kΩ, and it is extremely difficult to reduce this. On the other hand, when attempting to increase the resistance of a resistor, it becomes difficult to maintain high precision with a monolithic resistor. For example, known methods for obtaining high resistance in monolithic devices include using ion-implanted polysilicon and using a diffusion layer, but the former method using polysilicon results in extremely large variations in resistance and high resistance. In the case of the latter diffusion layer, the resistance value is dependent on voltage and it is difficult to obtain high resistance.

Normally, a diffusion layer is used as a resistor in a monolithic operational amplifier circuit, but due to this voltage dependence, linearity of only about 8 bits can be obtained, and sufficient precision cannot be obtained. A countermeasure for this has been proposed (Lecture No. 394, National Conference of the Institute of Electrical and Communications Engineers of Japan, 1974), but it is complicated and inevitably increases costs.

SUMMARY OF THE INVENTION An object of the present invention is to provide an operational amplifier circuit which eliminates the above-mentioned drawbacks of the prior art, allows gain setting with high precision, is easily monolithic, and is low cost.

In order to achieve this object, the present invention is characterized in that at least one of the input resistance and the feedback resistance of the operational amplifier circuit is constituted by the on-resistance of a switch element consisting of a MOSFET or a CMOSFET.

Embodiments of the operational amplifier circuit according to the present invention will be described below with reference to the drawings.

FIG. 2 shows an embodiment of the present invention, in which the same or equivalent parts as in the conventional example shown in FIG. 1 are given the same reference numerals.

In the figure, 20 to 29 are switch elements consisting of MOSFETs, 30 and 31 are decoders, and 32 to 3
9 is a decode output line of the decoders 30 and 31.

Switch elements 20 to 29 are decoders 30 and 3
It is turned on and off by the selection switching signal appearing on the output lines 32 to 39 of No. 1, and the resistance values of the input resistance and feedback resistance are set according to the value of the conduction resistance, that is, the on-resistance when the switch is in the on state. It works like this.

The gain _ANF at this time is as shown in the following equation.

A _NF = − _o 〓 ⁱ⁼¹ G _si / _n 〓 ^j=1 G _fj …(1) However, G _si = 1/R _si , G _fj = 1/R _fj R _si : Resistance value of input resistance (switch element 2
(equivalent to on-resistance of 2 to 25) R _fj : Resistance value of feedback resistor (switch element 2
6 to 29) On the other hand, switch elements 22 to 29 consisting of MOSFETs
The on-resistance R _po of No. 29 is expressed by the following equation.

R _po =1/δI _DS /δV _DS | (V _D →0) = 1/β ₀ (W/L) (V _G −V _T )...(2) Here, β ₀ : Channel conductance W/L: MOSFET channel V _G : Gate-source voltage V _T : Threshold voltage This on-resistance R _po can be relatively easily obtained with fairly high accuracy, but as is clear from the above equation, the gate-source voltage V T : Threshold voltage It exhibits nonlinear characteristics with respect to the source-to-source voltage V _G.

However, as is clear from FIG. 2, in this embodiment, the non-inverting input 16 of the operational amplifier 1 is grounded, so that the inverting input 15 becomes a virtual ground point, which is equivalent to being grounded. Therefore, the gates and gates for these switch elements 22 to 29 are
All source voltages V _G have the same value, and regardless of changes in this voltage V _G , the switch elements 22 to 2
Since the on-resistance ratio of 9 is always almost constant, there is no error in gain setting, and high accuracy can be maintained.

As a result of the above, according to this embodiment, the switch element 22 consisting of MOSFET can be used without using independent resistors as input resistance or feedback resistance.
Since it is possible to set the gain by using .~29, problems caused by the use of resistors are eliminated, and high precision can be easily achieved. The chip area is reduced and the chip can be made smaller.

Further, if an N-channel MOSIC is used, the substrate is grounded, and the source electrode is kept at the ground potential, which has the advantage of reducing the substrate effect.

Note that in order to control the on-resistance of the switch elements 22 to 29, a configuration may be adopted in which the channel ratio W/L of the MOSFETs is varied. In other words, as is clear from equation (2), the on-resistance of a switch element made of a MOSFET can be set to any value over a wide range by changing its channel ratio W/L, and the setting accuracy is quite high. Therefore, it is possible to perform highly accurate gain control over a wide range.

FIG. 3 shows the switch element 22 in the above embodiment.
-24 is a schematic diagram showing an example of the layout on the semiconductor substrate, 7' is an aluminum strip layer forming an extension from the input terminal 7, 1
5' is also an aluminum strip layer that serves as an extension from the inverting input 15 of the operational amplifier 1, 32' to 3.
4' is a polysilicon strip layer which becomes the gate electrode of the MOSFET switch elements 22 to 24, and a decoder 30.
40 is an N ⁺ diffusion layer, and 41 is a contact hole between the diffusion layer 40 and the aluminum strip layers 7' and 15'.

A thin alignment film is formed on the substrate in the hatched areas of the polysilicon strip layers 32' to 34', and the polysilicon strip layers 32' to 34' are formed thereon. This portion will work as a gate electrode, and the channel ratio will be W/L in the figure.

Now, the potential of the polysilicon strip layer 32' is set to high level "H" by a signal from the decoder 30 etc.
If this happens, a channel is formed in the shaded area of the switch element 22, the switch element 22 is turned on, and the aluminum strip layers 7' and 15' are electrically connected via the on-resistance of the switch element 22, and a predetermined level is reached. It operates as an input resistor with a resistance value to set the gain.

As in this embodiment, each switch element 22-2
5, the area on the chip can be significantly reduced.

FIG. 4 is a schematic diagram showing another embodiment of the layout of the switch elements 22 to 25, in which the same or equivalent parts as in the embodiment of FIG. 3 are given the same reference numerals.

In this embodiment, the aluminum strip layers 7', 15' and the polysilicon strip layers 32'-35' are elongated in the lateral direction, and the diffusion layer 40 is correspondingly extended.
Contact holes are also provided in a matrix in the horizontal and vertical directions, so that each of the polysilicon strip layers 32' to 35' commonly forms the gate electrode of five switch elements. .
For example, five switch elements 22a to 22e are formed in the polysilicon strip layer 32', so that when the potential of the polysilicon strip layer 32' reaches a high level "H", these switch elements 22a to 2
2e will be on and their on-resistances will all be inserted in parallel between the aluminum strip layers 7' and 15'.

Therefore, according to this embodiment, the resistors for setting the gain are connected to the switch elements 22a to 22, for example.
The resistance value is set by the parallel on-resistance of the five switch elements e, making it possible to set the resistance value with even higher accuracy. In other words, for things whose characteristics generally exhibit random variation distribution, such as resistors,
It is known that as a large number of them are collected and used, their variations are statistically reduced.

Therefore, if a switch element group consisting of a plurality of switch elements 22a to 22a is used as shown in FIG. 4, n and m in equation (1) will increase, making it easier to increase the precision of the gain.

Note that the threshold voltage of a switch element consisting of a MOSFET has a temperature dependence of about -2 mV/°C, but according to the above embodiment, each switch element 22
29 can be provided close to each other on the same substrate, so that variations in characteristics due to temperature changes appear in the same way for each element, making it possible to obtain excellent tracking characteristics.

Next, FIG. 5 shows another embodiment of the present invention having a multiplexer function, where 60 to 63 are a group of switch elements consisting of MOSFETs that also serve as input resistors and multiplexer switches, and 64 is a feedback resistor.
MOSFET switch element group, 60' to 64' are gate drive signal lines, 65 to 68 are analog signal input lines, and 69 is an analog signal output line. Note that 1 is an operational amplifier, and 15 is its inverting input.

Each analog signal provided on input lines 65-68 is inverted through one of switch elements 60-63 turned on by a gate drive signal selectively provided on lines 60'-63'. Selected and supplied to input 15, switch element group 6
The signal is amplified by the operational amplifier 1 under a predetermined gain set by the on-resistance of 0 to 63 and the on-resistance of the switch element group 64, and is taken out to the output line 69. Therefore, if the on-resistances of the switch element groups 60 to 64 are all set to the same value, they will operate as a multiplexer with a gain of 1.

FIG. 6 is a schematic diagram showing one embodiment of the layout of the switch element groups 60 to 63, in which 15' is an aluminum strip layer connected to the inverting input 15;
63'' is a polysilicon strip layer forming an extension of gate drive signal lines 60' to 63', and 65' to 68' are aluminum strip layers forming an extension of input lines 65 to 68. The diffusion layer 41 is a contact hole.

Polysilicon strip layers 60''-63'' form gate electrodes and switch elements 60a-60e, 61a.
~61e... is configured. Therefore, the switch element group 60 includes five switch elements 60a connected in parallel.
.about.60e, it becomes easier to achieve higher precision as in the embodiment shown in FIG. 4, and the switch element group can be accommodated within a small area.

In the above embodiments, the gate electrode of the MOSFET switch element is made of a polysilicon strip layer, but similar effects can be expected even if these are replaced with an aluminum strip layer.

Furthermore, in the above embodiments, all the switch elements were configured with MOSFETs, but the present invention is not limited to this, and can also be implemented with, for example, CMOS switch elements, and similar effects can be obtained. Can be done.

As explained above, according to the present invention, one or both of the input resistance and the feedback resistance of the operational amplifier circuit is replaced with the on-resistance of the switch element. In addition to being able to set or program the gain, it has excellent tracking characteristics and is resistant to fluctuations in power supply voltage. It is easy to make monolithic, small and low cost, and the source electrode of the NMOSFET switch element is connected to the virtual ground point of the operational amplifier. To provide an operational amplifier circuit which can reduce variations in on-resistance of switch elements and substrate effects, and can obtain high accuracy of gain by using a statistical variation reduction effect by using a group of switch elements. be able to.

[Brief explanation of drawings]

FIG. 1 is a circuit diagram showing an example of a programmable gain amplifier, FIG. 2 is a circuit diagram showing an example of an operational amplifier circuit according to the present invention, and FIG. 3 is a schematic diagram showing an example of the layout of a switch element. , Fig. 4 is a schematic diagram showing another embodiment of the layout, Fig. 5 is a circuit diagram showing an embodiment of the present invention having a multiplexer function, and Fig. 6 is a schematic diagram showing an embodiment of the layout of the switch element. It is a diagram. 1... operational amplifier, 7... input terminal, 8... output terminal, 15... inverting input, 16... non-inverting input, 22-29... switch element consisting of MOSFET, 30, 31... decoder.

Claims (1)

[Claims] 1. Using an operational amplifier whose non-inverting input is grounded,
In an operational amplifier circuit that performs gain control by switching the resistance value of at least one of an input resistor and a feedback resistor connected to its inverting input, a plurality of MOSFETs and CMOSFETs having the same channel ratio are connected to the same semiconductor substrate. a switch element group consisting of at least one parallel connected element;
A decoder that supplies a switching signal to the gate of each of these switch element groups is provided, and the conduction resistance between the source and drain that occurs when the plurality of switch element groups are switched to a conductive state is determined by the input resistance and the feedback resistance. 1. An operational amplifier circuit characterized in that the operational amplifier circuit is configured to be used as at least one resistor, and to perform gain control by selecting the switching state of the switching element group by the decoder. 2. In claim 1, the switch element group includes a plurality of diffusion layers arranged in a matrix on a semiconductor substrate and formed independently, and diffusion layers located in the row direction of the matrix of the plurality of diffusion layers. A plurality of parallel aluminum strip layers connecting the layers in parallel and a polysilicon strip layer formed on the semiconductor substrate surface between these aluminum strip layers are provided, whereby the polysilicon strip layer is formed on the semiconductor substrate surface between these aluminum strip layers. MOSFETs are formed in a matrix, with the layer serving as a gate electrode, the diffusion layer located on one side thereof serving as a source electrode, and the diffusion layer located on the other side serving as a drain electrode, with one of the aluminum strip layers serving as a common MOSFET. 1. An operational amplifier circuit comprising a group of MOSFET switch elements each having a source electrode lead line on the other side, a common drain electrode lead line on the other side, and a common gate electrode using the polysilicon strip layer. 3. The operational amplifier circuit according to claim 2, wherein the common gate electrode is composed of an aluminum strip layer.