JP2006014111A - TRANSCONDUCTANCE AMPLIFIER AND Gm-C TYPE INTEGRATOR - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a transconductor amplifier which is made into integrated circuit so as to ensure an integrator gain of about 60dB at a DC level. <P>SOLUTION: P-channel type MOSFETs M7A, M8A are connected to drains of MOSFETs M7, M8 of the same P-channel type, gates of these MOSFETs M7A, M8A are connected in common, and a drain of the MOSFET M7A is connected with a drain of an MOSFET M5 via an N-channel type MOSFET M5A. A drain of the MOSFET M8A is connected to an output terminal 12 and connected with a drain of an MOSFET M6 via an N-channel type MOSFET M6A. Furthermore, capacitors C2, C3 are connected for preventing oscillation by increasing capacitance between gates and sources of MOSFETs M3, M4. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a transconductance amplifier having good linearity and a Gm-C type integrator using the transconductance amplifier, and more particularly to an integrated circuit-conductance amplifier and a Gm-C type integrator having a high output resistance.

FIG. 8 is a diagram showing a circuit configuration of a conventional transconductance amplifier.
In FIG. 8, in a differential circuit composed of P-channel MOSFETs (MOS field effect transistors) M1 and M2 constituting an input differential pair, the gates of MOSFETs M1 and M2 are connected to input terminals 10 and 11, respectively. . Differential inputs IN (-) and IN (+) are applied to the input terminals 10 and 11, respectively, and constant current sources I1 and I2 connected to the respective drains of these MOSFETs M1 and M2 are used as constant currents. Driven. Also, constant current sources I3 and I4 are connected to the sources of the MOSFETs M1 and M2, respectively.

  Among these, the drain of the MOSFET M1 is provided with a first current mirror circuit composed of N-channel MOSFETs M3 and M5, and is connected to the gates of the commonly connected MOSFETs M3 and M5. The drain of the MOSFET M3 of the first current mirror circuit is connected to the constant current source I3 and one end of the resistor R1 via a node (node) X.

  The drain of the MOSFET M2 is provided with a second current mirror circuit composed of N-channel MOSFETs M4 and M6, and is connected to the gates of the commonly connected MOSFETs M4 and M6. The drain of the MOSFET M4 of the second current mirror circuit is connected to the other end of the constant current source I4 and the resistor R1 through a node (node) Y.

  The P-channel MOSFETs M7 and M8 have a third current mirror circuit in which the gates are connected in common and the sources are connected to the power supply terminal VDD. In the third current mirror circuit, the drain of one of the diode-connected MOSFETs M7 is connected to the drain of the MOSFET M5 of the first current mirror circuit, the drain of the other MOSFET M8 is connected to the output terminal 12, and The current mirror circuit 2 is connected to the drain of the MOSFET M6. From the output terminal 12, the output current OUT can be extracted with a magnitude corresponding to the difference between the differential inputs IN (-) and IN (+).

  The input stage of such a transconductance amplifier is shown, for example, in “Supervised by Katsuharu Kimura, Analog Circuit Design Technology for CMOSization of Portable Wireless Terminals, pages 114 and 115”. The output stage is a normal current mirror circuit, and a similar circuit is described in Patent Document 1 described later. Moreover, as for the above-described input stage, Patent Document 2 discloses a similar circuit configured with a bipolar transistor.

  The circuit operation of this transconductance amplifier will be described in detail. Here, for the sake of simplicity, M3 and M5, M4 and M6, and M7 and M8 are MOSFETs each having the same size, and constant current sources I1 and I2, I3 and I4 are respectively It is assumed that currents having the same current values i0 and i1 flow.

  Assuming that no current flows through the gates of the MOSFETs M3 to M6 constituting the first and second current mirror circuits, a constant current value i0 always flows through the MOSFETs M1 and M2 of the differential circuit. At this time, the MOSFETs M1 and M2 are not affected by the substrate effect because the source and the substrate (bulk) are short-circuited. Therefore, the gate-source voltage Vs of the MOSFETs M1 and M2 is always a constant value determined by the device size and the current value i0.

  The potentials at both ends of the resistor R1, that is, the potentials Vx and Vy of the nodes X and Y are the gate-source voltage Vs and the voltage value Vinm of the differential inputs IN (−) and IN (+) at the input terminals 10 and 11, Using Vinp, the following expressions (1a) and (1b) can be expressed.

  In this transconductance amplifier, the MOSFETs M3 and M4 function to maintain the current value flowing through the MOSFETs M1 and M2 at i0. For example, when the current value flowing through the MOSFET M1 decreases, the source-drain voltage of the MOSFET M1 increases to return to the current value i0, the gate potential of the MOSFET M3 decreases, the drain current decreases, and as a result, the current flowing through the MOSFET M1 increases. Thus, the original current value is maintained.

  The drain currents Idm3 and Idm4 of the MOSFETs M3 and M4 can be expressed as the following equations (2a) and (2b) by applying Kirchhoff's first law (continuity of current) at the nodes X and Y. Here, R is the resistance value of the resistor R1.

  Here, drain currents Idm3 and Idm4 flowing in MOSFETs M3 and M4 are equal to current values flowing in MOSFET M5 and MOSFET M6, respectively. Therefore, when the difference is taken out from the output terminal 12 as the output current Iout using the third current mirror circuit composed of the MOSFET M7 and the MOSFET M8, the magnitude thereof is expressed by the following equation (3).

  That is, in this transconductance amplifier, an output current Iout proportional to the difference between the two input voltages Vinm and Vinp is obtained. From equation (3), it can be seen that the transconductance value of this amplifier is Gm = 2 / R, and that Gm value is determined by the size of the resistor R1. For this reason, this transconductance amplifier has high linearity.

Next, a Gm-C type integrator using this will be described.
FIG. 9 is a circuit diagram illustrating an example of a Gm-C type integrator. Here, including the one in which the resistor R2 is inserted in series with the capacitor C1 on the output side of the transconductance amplifier 1, it is called a Gm-C type integrator.

  The transfer function T (s) from the input IN to the output OUT of this Gm-C integrator is the transconductance Gm of the transconductance amplifier 1, the values of the resistor R2 and the capacitor C1 connected to the output (R2, C1 respectively) Can be expressed as the following equation (4).

In this equation (4), s is a Laplace operator, and the gain becomes infinite in the case of DC input (s = 0).
As shown in FIG. 10, in the transfer function equivalent to the equation (4), a resistor R3 having a resistance value of 1 / Gm is provided on the input side, and a series circuit of the resistor R2 and the capacitor C1 forms a feedback circuit. It can also be realized by using the operational amplifier 2. However, the input impedance can be increased by using the transconductance amplifier 1 shown in FIG. Further, when the transconductance amplifier 1 is integrated into an integrated circuit and the resistor R2 and the capacitor C1 are connected to the external terminal of the integrated circuit, there is an advantage that only one external terminal is required.
Japanese Patent Laid-Open No. 10-150332 Japanese Patent Application Laid-Open No. 08-222969

  However, in the Gm-C type integrator using the transconductance amplifier 1, as shown in FIG. 11, the output resistance Ro of the output stage MOSFET is actually inserted in parallel with R2 and C1 connected to the output side. . For this reason, when the output resistance Ro in the saturation region of the MOSFET constituting the transconductance amplifier 1 is small, the maximum gain of the Gm-C integrator is limited by Gm · Ro. There was a problem that could not be obtained. Such a problem may occur in the entire output stage of the current output amplifier, and is not limited to the transconductance amplifier having the configuration shown in FIG.

  The present invention has been made in view of these points, and an object of the present invention is to provide an integrated circuit transconductance amplifier capable of securing an integrator gain of about 60 dB at a DC level.

  Another object of the present invention is to provide a Gm-C type integrator that operates stably with respect to a signal source having a high output resistance.

  In the present invention, in order to solve the above problem, first and second differential inputs are applied to the respective gates, the first and second constant current sources are connected to the drains as loads, and the first and second constant current drives are performed. And a third circuit having a gate connected to the drain of the first MOSFET in common, and the drain of the third MOSFET serving as the first node. Via a third constant current source and a first current mirror circuit connected to the source of the first MOSFET, and fourth and sixth gates connected in common to the drain of the second MOSFET. And a drain of the fourth MOSFET is connected to a fourth constant current source and a source of the second MOSFET via a second node. A mirror circuit, a first resistor connecting the first node and the second node, and seventh and eighth MOSFETs each having a gate connected in common. A third current mirror circuit in which the drain of the MOSFET is connected to the drain of the fifth MOSFET, and the gate of the eighth MOSFET is connected to the drain of the sixth MOSFET via an output terminal; A transconductance amplifier configured to obtain a current output corresponding to a difference between a current flowing from the output terminal to the fifth MOSFET and a current flowing to the sixth MOSFET, the sixth and eighth MOSFETs being The drains of the sixth and eighth MOSFETs are connected via the ninth and tenth MOSFETs that are cascode-connected, respectively. Transconductance amplifier, characterized in that it is connected to the force terminal is provided.

  According to the present invention, it is possible to configure a Gm-C type integrator that has good linearity, operates stably with respect to a signal source having a high output resistance, and can obtain a sufficient DC gain.

FIG. 1 is a circuit diagram showing a transconductance amplifier according to an embodiment. Here, parts corresponding to those of the conventional circuit of FIG.
The difference from the conventional circuit of FIG. 8 is that the third current mirror circuit constituting the output stage is cascoded. That is, the same P-channel type MOSFETs M7A and M8A are connected to the drains of the P-channel type MOSFETs M7 and M8, the gates of these MOSFETs M7A and M8A are connected in common, and the drain of the MOSFET M7A is the N-channel type It is connected to the drain of MOSFET M5 through MOSFET M5A. The drain of the MOSFET M8A is connected to the output terminal 12 and is connected to the drain of the MOSFET M6 via the N-channel type MOSFET M6A.

  Further, the substrates of the MOSFETs M7A and M8A are respectively connected to the power supply terminal VDD similarly to the MOSFETs M7 and M8. Further, MOSFETs M3A and M4A of the same size as the N-channel MOSFETs M5A and M6A are connected to the drain side of the N-channel MOSFETs M3 and M4 constituting the first and second current mirror circuits, respectively. The gates of these MOSFETs M3A, M4A, M5A and M6A are connected to a DC voltage source Vbias suitable for operating them in the saturation region.

  Here, the cascode-connected current mirror circuit has an effect of minimizing the influence of the current change due to the channel length modulation effect even when the output voltage at the output terminal 12 changes. The channel length modulation effect means that a drain voltage change based on the source potential causes an effective channel length change and a drain current in the saturation region changes. As a result, when the output potential of the transconductance amplifier changes, the output current generated by the difference between the drain currents of MOSFET M6 and MOSFET M8 changes. However, by configuring a cascode-connected current mirror circuit as shown in FIG. 1, it is possible to suppress changes in the drain voltage of the MOSFET M6 and MOSFET M8 due to voltage changes at the output terminal 12, and to prevent changes in the drain current of both MOSFETs. A sufficient DC gain can be obtained when the output terminal 12 is open.

  Further, in the transconductance amplifier of FIG. 1, capacitors C2 and C3 for preventing oscillation are added between the gates of MOSFET M3 and MOSFET M4 and the drains of MOSFET M3A and MOSFET M4A, respectively. As will be described later, this point is different from the conventional circuit of FIG.

  FIG. 2 is a diagram illustrating an example of a bias circuit that generates the constant current sources I1 and I2, I3 and I4, and the DC voltage source Vbias. This bias circuit is composed of P-channel MOSFETs MB1 to MB9 and a constant current source Ib between a power supply terminal VDD and a ground potential (GND), and generates a bias voltage necessary for the transconductance amplifier of FIG. is doing. A DC voltage source Vbias is output from the output terminal 12.

  FIG. 3 is a diagram illustrating frequency characteristics of the integrator gain. This shows a simulation of the frequency characteristics when the integrator is configured using the circuit (FIG. 8) excluding the transconductance amplifier of FIG. 1 and the cascode-connected devices (MOSFETs M3A to M8A). Yes. A sufficient DC gain cannot be obtained with an integrator (indicated by a broken line in the figure) whose output stage is not cascode-connected. On the other hand, in the integrator using the transconductance amplifier of FIG. 1, a gain of about 60 dB that can be regarded as practically infinite is secured.

  As described above, the transconductance amplifier of the present invention can increase the output resistance because the change in the output voltage is not directly transmitted to the drain voltage of the MOSFET that determines the current value by cascoding the output stage. . Therefore, a necessary DC gain can be obtained in the Gm-C type integrator.

  FIG. 4 is a circuit diagram showing another example of cascode connection. In this cascode circuit, the output resistance can be increased by applying feedback using the auxiliary amplifier 14 so that the drain voltage of the MOSFET M5 is always constant.

  Here, the oscillation preventing capacitors C2 and C3 added to the transconductance amplifier of FIG. 8 as described above will be described. The problem with the circuit of FIG. 8 is that the input stage may oscillate when connected to a signal source with high output resistance. This is because the transconductance amplifier is often used when inputting a signal source having a high output resistance, including the case where the transconductance amplifier generates a reference voltage by resistance division.

Therefore, this oscillation phenomenon will be described.
FIG. 5 is a circuit diagram showing only the input stage of the transconductance amplifier of FIG. A signal source (voltage source) Vx is connected to the input terminal of this transconductance amplifier via its output resistor Rs. FIG. 6 is an equivalent circuit diagram showing a small signal equivalent circuit of this input stage.

  In FIG. 6, in order to simplify the description, the output resistance Ro of the MOSFETs M1 and M3 is infinite. Here, gm1 and gm3 are transconductances near the bias points of M1 and M3, and Cgs1 and Cgs3 are gate-source capacitances of M1 and M3. Cp is the total value of parasitic capacitance such as the drain-bulk (substrate) capacitance of the MOSFET that constitutes M3 and the bias current source I3. Other parasitic capacitances are not considered for the sake of simplicity.

  When the input impedance Zx at the node (node) X of the equivalent circuit shown in FIG. 6 is calculated, the following equation (5) is obtained.

  Here, the oscillation condition in the equivalent circuit of FIG. 6 is determined. That is, the Laplace operator s in equation (5) is replaced with jω and returned to the frequency domain, and then the denominator in equation (5) is obtained in order to obtain a condition where both the real part and imaginary part of the denominator are zero To summarize, the following equation (6) is obtained.

Since the circuit oscillates when the input impedance Zx becomes infinite, gm1 · gm3 = ω ² · Cgs3 (Cgs1 + Cp) is derived by setting the real part of the equation (6) to 0. Here, if the gate potential of the MOSFET M1 is increased, the MOSFETs M1 and M3 are cut off, and if the gate potential of the M1 is decreased, the M1 is not saturated. As a result, when it is assumed that the value of gm1 · gm3 becomes small, the oscillation condition can be rewritten as gm1 · gm3 ≧ ω ² · Cgs3 (Cgs1 + Cp).

Further, when the imaginary part of the equation (6) is 0, ω ² = gm1 / Cgs1 · Cp · Rs is derived. Therefore, when the inequality indicating the oscillation condition is arranged for the gate-source capacitance Cgs3 of the MOSFET M3, the following equation (7) is obtained.

  From this expression (7), the following expression (8) is derived as a conditional expression for the parasitic capacitance Cp for preventing the transconductance amplifier from oscillating.

  In this equation (8), gm3 (Cgs1‖Cp) is a positive value. Therefore, in order to design the transconductance amplifier so as not to oscillate even when the resistance Rs becomes a large value, care is taken for the parasitic capacitance. It can be seen that not only circuit design / layout is required, but also the usable device size is limited. That is, it is necessary to prevent the minute parasitic capacitance Cp from increasing. In order to avoid such a problem by interposing a buffer circuit between the signal source Vx and the input terminal of the transconductance amplifier, another problem arises that the scale of the amplifier circuit increases.

  According to the present invention, the gate-source capacitance Cgs3 of the MOSFET M3 is made sufficiently large by the capacitor C2, thereby satisfying the condition (8) and preventing the oscillation. Note that it is better to connect the capacitor C2 between the gate and drain of the MOSFET M3 (via the MOSFET M3A when the MOSFET M3A is provided as in the circuit shown in FIG. 1) than to connect the capacitor C2 directly between the gate and source of the MOSFET M3. By using the mirror effect, the same effect can be obtained with a smaller capacity. The above discussion can be applied to the capacitor C3 connected to the MOSFET M4 in exactly the same manner, and has the same effect. The transconductance amplifier of FIG. 1 is an example in which the gate-source capacitance Cgs3 is made sufficiently large by utilizing the Miller effect.

FIG. 7 is a diagram illustrating a transient response characteristic of the transconductance amplifier.
Here, for each of the transconductance amplifier shown in FIG. 1 and the circuit from which the capacitors C2 and C3 are removed, a power supply is applied with 1 V (= Vx) applied to both inputs via a 1 MΩ resistor (= Rs). The simulation shows the time response of the potential at the non-inverting input (input terminal 11) when the terminal is raised to 2.5V. In the circuit in which the capacitors C2 and C3 are not provided, the oscillation state continues as shown by the broken line in the figure. On the other hand, the transient response of the transconductance amplifier of FIG. 1 confirms that it is operating stably with respect to a signal source having a high output resistance.

FIG. 12 is a table showing an example of the size of each element such as a MOSFET required for configuring the transconductance amplifier of FIG. 1 with a conductance value of 1 [μS] (10 ⁻⁶ Siemens). In addition, the element value of MOSFET in the table | surface is shown by ratio (a unit is micrometer) of W (gate width) / L (gate length).

It is a circuit diagram which shows the transconductance amplifier which concerns on embodiment. It is a figure which shows an example of a bias circuit. It is a figure which shows the frequency characteristic of an integrator gain. It is a circuit diagram which shows the example of another cascode connection. FIG. 9 is a circuit diagram showing only the input stage of the transconductance amplifier of FIG. 8. FIG. 6 is an equivalent circuit diagram showing a small signal equivalent circuit of the input stage of FIG. 5. It is a figure which shows the transient response characteristic of transconductance amplifier. It is a figure which shows the circuit structure of the conventional transconductance amplifier. It is a circuit diagram which shows an example of a Gm-C type | mold integrator. It is a circuit diagram which shows an example of the Gm-C type | mold integrator using an operational amplifier. It is a circuit diagram which shows a Gm-C type | mold integrator when the output impedance of amplifier is finite. It is a table | surface which shows an example of magnitude | sizes of MOSFET etc. regarding the transconductance amplifier of FIG.

Explanation of symbols

10, 11 input terminal 12 output terminal C2, C3 capacitor M1-M8, M3A-M8A MOSFET (MOS type field effect transistor)
I1 to I4 constant current source X, Y node R1 resistance Vbias DC voltage source

Claims (7)

  A transconductance amplifier that receives a voltage signal and outputs a current signal corresponding to the voltage signal from an output terminal, connected between the output MOSFET and the output terminal, an output MOSFET that determines the magnitude of the output current A transconductance amplifier comprising a channel length modulation effect preventing MOSFET. A differential circuit composed of first and second MOSFETs, each of which has a differential input applied to each gate, the first and second constant current sources are connected to the drain as a load, and are driven by a constant current;
The drain of the first MOSFET is composed of third and fifth MOSFETs each having a gate connected in common, and the drain of the third MOSFET is connected to the third constant current source via the first node; and A first current mirror circuit connected to a source of the first MOSFET;
The fourth MOSFET is composed of fourth and sixth MOSFETs each having a gate commonly connected to the drain of the second MOSFET, and the drain of the fourth MOSFET is connected to the fourth constant current source via the second node, and A second current mirror circuit connected to the source of the second MOSFET;
A first resistor connecting between the first node and the second node;
Each of the gates is composed of seventh and eighth MOSFETs connected in common, the drain of the seventh MOSFET is connected to the drain of the fifth MOSFET, and the gate of the eighth MOSFET has an output terminal. A third current mirror circuit connected to the drain of the sixth MOSFET via
A transconductance amplifier configured to obtain a current output corresponding to a difference between a current flowing from the output terminal to the fifth MOSFET and a current flowing to the sixth MOSFET,
The drains of the sixth and eighth MOSFETs are connected to the output terminal via the ninth and tenth MOSFETs that are cascode-connected to the sixth and eighth MOSFETs, respectively. Transconductance amplifier.   The eleventh MOSFET cascode-connected to the seventh MOSFET and the twelfth MOSFET cascode-connected to the fifth MOSFET via the drain of the seventh MOSFET and the drain of the fifth MOSFET The transconductance amplifier according to claim 2, wherein:   The drain of the third MOSFET is connected to the first node via a thirteenth MOSFET that is cascode-connected to the third MOSFET, and the fourteenth MOSFET is cascode-connected to the fourth MOSFET. 4. The transconductance amplifier according to claim 3, wherein a drain of the fourth MOSFET is connected to the second node. The transconductance amplifier according to any one of claims 1 to 4,
A series circuit composed of a capacitor and a second resistor having one end connected to the output terminal of the transconductance amplifier and the other end grounded;
A Gm-C type integrator comprising:   The transconductance amplifier is configured by providing a capacitor for connecting a gate and a source of the third and fourth MOSFETs to the first current mirror circuit and the second current mirror circuit, respectively. The Gm-C type integrator according to claim 5, wherein the Gm-C type integrator is provided. The transconductance amplifier includes a capacitor for connecting the gate and drain of the third and fourth MOSFETs to the first current mirror circuit and the second current mirror circuit, respectively. The Gm-C type integrator according to claim 5, wherein the Gm-C type integrator is provided.

Images