This application is based upon and claims the benefit of priority from the prior Japanese Patent Application No. 2007-137233, filed on May 23, 2007, the entire contents of which are incorporated herein by reference.
1. Field of the Invention
The present invention relates to a variable impedance circuit using an internal resistance of a transistor; and a variable impedance system, a filter circuit, an amplifier, and a communication system using the same.
2. Description of the Related Art
A variable impedance circuit is used for varying an amount of signals passing through a switch, an amount of signals attenuating in an attenuator, and a gain in a variable gain amplifier, or the like.
The present inventors have found out a variable impedance circuit having a better linearity over a broad frequency band than the variable impedance circuit shown in
A variable impedance circuit of an embodiment of the present invention is a variable impedance circuit in which a first transistor and a second transistor are connected in series. The first transistor includes a first terminal that receives one of differential input signals, and a second terminal that receives a control signal for varying an impedance, a third terminal connected to the second transistor, and a fourth terminal that supplies a potential to a substrate (a semiconductor substrate), and the second transistor includes a fifth terminal that receives the other of the differential input signals, a sixth terminal that receives the control signal, a seventh terminal connected to the first transistor, and an eighth terminal that supplies a potential to the substrate (the semiconductor substrate). The third terminal, the fourth terminal, the seventh terminal, and the eighth terminal are connected together. Specifically, when a NMOSFET is adopted, the NMOSFET has a well structure in which a semiconductor region having a polarity electrically opposite to the substrate, in other words, an n-type semiconductor region, is formed in the p-type semiconductor substrate; and the back gate terminal thereof is connected to the well; and the well may or may not be common between the above first transistor and second transistor. On the other hand, when a PMOSFET is adopted, the back gate terminal thereof is directly connected to the substrate.
The invention will now be described by reference to the preferred embodiments. This does not intend to limit the scope of the present invention, but to exemplify the invention.
The first transistor M1 and the second transistor M2 are constituted by the N channel type MOSFETs (Metal-Oxide-Semiconductor Field Effect Transistors), respectively. The first transistor M1 and the second transistor M2 are respectively provided with four terminals, which are the source terminal, the gate terminal, the drain terminal, and the back gate terminal.
The first terminal of the first transistor M1 is connected to the first I/O terminal 10 via the first resistance R1. The second terminal of the first transistor M1 is the gate terminal connected to the control terminal 14. The control signal CTRL for varying an impedance of the variable impedance circuit 100 is inputted to the control terminal 14. The second terminal of the first transistor M1 is connected to the seventh terminal of the second transistor M2 mentioned later. Either of the second terminal and the third terminal, which has a higher potential, serves as a drain terminal, while the other as a source terminal. The fourth terminal of the first transistor M1 is the back gate terminal (also referred to as the substrate terminal), which is a terminal for supplying a voltage to the substrate, not to the gate side of a channel to be formed.
The fifth terminal of the second transistor M2 is connected to the second I/O terminal 12 via the second resistance R2. The sixth terminal of the second transistor M2 is the gate terminal connected to the control terminal 14. The seventh terminal of the second transistor M2 is connected to the second terminal of the first transistor M1. Either of the sixth terminal and the seventh terminal, which has a higher potential, serves as a drain terminal, while the other as a source terminal. The eighth terminal of the second transistor M2 is the back gate terminal.
All of the third terminal, the fourth terminal, the seventh terminal, and the eighth terminal are short-circuited to have the same potential. When integrating the first transistor M1 and the second transistor M2, a triple well structure is used in order to separate the substrates terminals of the first transistor M1 and the second transistor M2 from the substrate terminal of a peripheral circuit.
As shown in the following equation (1), a drain-source current Ids flows through the first transistor M1 and the second transistor M2 in accordance with a gate-source voltage Vgs and a drain-source voltage Vds, in the non-saturation region, i.e., in the linear region:
I
ds=2K{(Vgs−VT)−Vds/2}Vds (1)
wherein K represents a transconductance parameter; and VT represents a threshold voltage.
The threshold voltage VT is represented by the following equation (2):
V
T
=V
T0+γ{√{square root over ( )}(2φgs+Vsb)−√{square root over ( )}(2φgs)} (2)
wherein γ represents a device constant; φ represents a surface potential in a strong inversion layer; and Vsb represents a source-substrate voltage.
When a high drain-source voltage is applied to the first transistor M1 or the second transistor M2, either of the two falls into the diode operation region, thereby the characteristic as a variable impedance circuit is not obtained. Herein, the diode operation region refers to a region where a drain-source voltage becomes higher in a negative direction to exceed a negative threshold. When exceeding the negative threshold, a drain-source current starts flowing like in a diode.
To prevent this from happening, the first resistance R1 and the second resistance R2 are inserted between the first I/O terminal 10 and the first transistor M1, and between the second I/O terminal 12 and the second transistor M2, respectively. The first resistance R1 and the second resistance R2 divide the voltages applied to the first I/O terminal 10 and the second I/O terminal 12 to reduce the drain-source voltage to be applied to the first transistor M1 and the second transistor M2. At the time, the first transistor M1 and the second transistor M2 operate in a deep linear region.
Herein, the characteristic of the variable impedance circuit 100 according to the embodiment illustrated in
The circuit according to a conventional technology has still a large distortion in a low frequency region, while having an improved one in a high frequency region. This indicates that a circuit structure according to a conventional technology has an advantage only in a high frequency region. This is because a circuit according to a conventional technology has a frequency dependence due to a time constant specified by a resistance and a capacitance, making the circuit not suitable for the use in a low frequency region. By contrast, it can be understood that the circuit according to the embodiment has a smaller distortion than the circuit according to a conventional technology and has better linearity ranging from a low frequency region to a high frequency region.
As mentioned above, better linearity can be obtained over a wide band according to the embodiment. Since the source terminal and the substrate terminal of a transistor are always short-circuited, deterioration of the linearity caused by the substrate effect can be suppressed. That is, a source-substrate voltage Vsb in the above equation (2) can always be zero. When the sauce terminal and the substrate terminal are not short-circuited, resistance values of the first transistor M1 and the second transistor M2 greatly vary each other when differential signals are inputted, therefore the source-substrate voltage Vsb also varies. Thereby, a distortion occurs in the threshold voltage VT of the first transistor M1 or the second transistor M2, causing a distortion in the drain-source current Ids. The structure according to the present embodiment can prevent this from happening. In addition, since the structure of the embodiment does not use an element, such as a capacitance, a better characteristic can be obtained over a wide band ranging from a direct current to a high frequency region.
According to the modification example 2, the contribution rate of the resistance values in the first transistor M1 and the second transistor M2 to the total impedance value, is reduced in comparison with the variable impedance circuit 100 illustrated in
An application example of the variable impedance circuit 100 according to the present embodiment will be described below. An example will be at first explained in which the circuit 100 is applied to an active filter circuit. As a premise for that, an example will be explained in which a plurality of the variable impedance circuits 100 according to the embodiment are provided, and each impedance of each variable impedance circuit 100a, 100b, and 100c is controlled by a single control signal CTRL.
R1a: R1b: R1c=α1: α2: α3 (3)
R2a: R2b: R2c=α1: α2: α3 (4)
(L1a/W1a): (L1b/W1b): (L1b/W1b)=α1: α2: α3 (5)
(L2a/W2a): (L2b/W2b): (L2b/W2b)=α1: α2: α3 (6)
wherein L1 represents the gate length of the first transistor M1; W1 represents the gate width of the first transistor M1; L2 represents the gate length of the second transistor M2; and W2 represents the gate width of the second transistor M2:
Z1: Z2: Z3=α1: α2: α3 (7)
wherein Z represents the value of the impedance of the variable impedance circuit 100.
The second-order low-pass filter 20 in the first stage includes three fully-differential type OTAs (Operational transconductance amplifiers) 21, 22, and 23, and 2 sets of respective pairs of capacitances: C21a and C21b; and C22a and C22b. To the non-inversion input terminal and the inversion input terminal of the first OTA 21, the differential input voltages Vin± are inputted. The inversion output terminal of the first OTA 21 is connected to one end of the first capacitance C21a, to the non-inversion input terminal of the second OTA 22, and to the non-inversion output terminal of the third OTA 23. The non-inversion output terminal of the first OTA 21 is connected to one end of the second capacitance C21b, to the non-inversion input terminal of the second OTA 22, and to the non-inversion output terminal of the third OTA 23. The other ends of the first capacitance C21a and the second capacitance C21b are connected to a predetermined fixed potential, herein, the ground potential. It is noted that the nodes of the first capacitance C21a and the second capacitance C21b, which are opposite to the ground, are denoted by the first node N1 and the second node N2, respectively.
The inversion input terminal of the second OTA 22 is connected to the first node N1, and the non-inversion input terminal of the second OTA 22 to the second node N2. The non-inversion output terminal of the second OTA 22 is connected to one end of the third capacitance C22a, to the non-inversion input terminal of the third OTA 23, and to the non-inversion input terminal of the second-order low-pass filter circuit in the next stage. The inversion output terminal of the second OTA 22 is connected to one end of the fourth capacitance C22b, to the inversion input terminal of the third OTA 23, and to the inversion input terminal of the second-order low-pass filter circuits in the next stage. The non-inversion input terminal of the third OTA 23 is connected to the first node N1, and the inversion input terminal of the second OTA 22 to the second node N2.
The first OTA 21 converts the differential input voltages Vin± to currents in accordance with the transconductance Gm. The first capacitance C21a and the second capacitance C21b integrate the outputs of the first OTA 21. The second OTA 22 converts the voltages of the first node N1 and the second node N2 to currents in accordance with the transconductance Gm. The third capacitance C22a and the fourth capacitance C22b integrate the outputs of the second OTA 22. The outputs of the third OTA 23 are feedbacked to the outputs of the first OTA 21 and the inputs of the second OTA 22. With this structure, a second-order transfer function can be realized. In addition to this structure, the first variable impedance circuit 100a is connected between the first node N1 and the second node N2.
The second-order low-pass filter circuit 30 in the second stage includes three fully-differential type OTAs 31, 32, and 33, and 2 sets of respective pairs of capacitances: C31a and C31b; and C32a and C32b. The second-order low-pass filter circuit 40 in the third stage includes three fully-differential type OTAs 41, 42, and 43, and two sets of respective pairs of capacitances: C41a and C41b; and C42a and C42b. The structures of the second-order low-pass filter circuits 30 and 40 in the second stage and the third stage are the same as that of the second-order low-pass filter circuit 20 in the first stage, therefore description with respect thereto will be omitted.
The transconductance values Gm of the first OTA 21, the second OTA 22, the third OTA 23, the fourth OTA 31, the fifth OTA 32, the sixth OTA 33, the seventh OTA 41, the eighth OTA 42, and ninth OTA 43, are controlled by the same control signal Vc. The impedance values of the first variable impedance circuit 100a, the second variable impedance circuit 100b, and the third variable impedance circuit 100c, are controlled by the same control signal Vq.
The Q factor of each stage of the sixth-order low-pass filter circuit 200 is represented by the following equations (8), (9), and (10) from the first stage:
Q
1(Vc)=Gm•Z1•√{square root over ( )}(C21/C22) (8)
Q
2(Vc)=Gm•Z2•√{square root over ( )}(C31/C32) (9)
Q
3(Vc)=Gm•Z3•√{square root over ( )}(C41/C42) (10).
In the sixth-order low-pass filter circuit 200, when the transconductance value Gm varies by a variation of an element among transistors, the Q factor of each stage of the circuit 200 varies accordingly. In order to correct this, it is necessary to return the Q factor of each stage to a desired preset value by controlling the first variable impedance circuits 100a-100c and adjusting the each impedance value. When each impedance of the variable impedance circuits 100a-100c is controlled by a single control terminal, it becomes impossible to return all of the Q factor of each stage to desired preset values simultaneously, if the ratio of each impedance varies depending on a control voltage. In this case, it is necessary to control each of a plurality of variable impedances independently, making the control difficult or complicated.
By contrast, when using the structure illustrated in
An example where the variable impedance circuit 100 according to the embodiment is applied to a differential amplifier circuit will be described below. A differential amplifier circuit is widely used for amplification in a high frequency band.
Differential input signals Vin± are inputted to the gate terminals of the pair of transistors M11 and M12. The source terminals of the pair of transistors M11 and M12 are commonly connected to the constant current source 50. The other end of the constant current source 50 is connected to a fixed potential, herein the ground potential. The drain terminals of the pair of transistors M11 and M12 are connected to the pair of loads L11 and L12, respectively. The other end of the pair of loads L11 and L12 are connected to a predetermined fixed potential, herein the power supply potential Vdd. The variable impedance circuit 100 according to the present embodiment is connected between the drain terminals of the pair of transistors M11 and of M12. Alternatively, the variable impedance circuits 110 and 120 according to the modification examples 1 and 2 may be connected.
The pair of transistors M11 and M12 generate differential output signals Vout± by applying the drain-source current corresponding to the differential input signals Vin± to the pair of loads L11 and L12, respectively. According to the present application example, the linearity of an amplifier operating at low gain can be improved better than the case where a conventional variable impedance circuit is used.
An example where the variable impedance circuit 100 according to the present embodiment is applied to a passive filter will be described below.
An example where the variable impedance circuit 100 according to the present embodiment is applied to an amplifier using an op-amp will be described below. An amplifier using an op-amp is widely used for amplification in a low frequency band.
The amplifier 500 inverts and amplifies an input voltage in accordance with the impedance ratio of the two variable impedance circuits 100. A voltage gain can be controlled by controlling the impedance value of at least one of the variable impedance circuits 100. The linearity can be improved better than the case where a conventional variable impedance is used, over a wide band by applying the variable impedance circuit 100 according to the embodiment.
The communication system 600 includes an antenna 52, a band-pass filter 54, an LNA (Low Noise Amplifier) 56, a local oscillator 58, a phase shifter 60, mixers 62 and 68, low-pass filter circuits 64 and 70, variable gain amplifiers 65 and 71, and AD conversion circuits 66 and 72.
An RF signal received by the antenna 52 is inputted to the LNA 56 via the band-pass filter 54. The LNA 56 amplifies the RF signal with a low noise and outputs it to the I signal mixer 62 and the Q signal mixer 68, wherein the I signal and the Q signal are quadrature baseband signals. The local oscillator 58 outputs a local (Lo) signal having a local frequency. The phase shifter 60 outputs the local signal to the I signal mixer 62 without shifting its phase, while outputs the local signal to the Q signal mixer 68 with its phase shifted 90 degrees.
The mixers 62 and 68 mix the RF signals and the Lo signals to output signals having a difference frequency between them to the first low-pass filter circuit 64 and the second low-pass filter circuit 70, respectively. Output signals of the first low-pass filter circuit 64 and the second low-pass filter circuit 70 are amplified by the variable gain amplifiers 65 and 71 in each system. Output signals of the variable gain amplifiers 65 and 71 in each system are respectively converted to digital signals by the first AD converter 66 and the second AD converter 72, and are outputted to the baseband portion 80. The signals are hereinafter subjected to digital signal processing by the DSP (Digital Signal Processor) in the baseband portion 80.
The various circuits using the variable impedance circuit 100 illustrated in
As mentioned above, the present invention has been described based on some embodiments. These embodiments are intended solely for the purpose of illustration, and it should be understood by those skilled in the art that various modifications are possible in combining those various components and various processing and those modifications also fall in the scope of the present invention.
Although the first transistor M1 and the second transistor M2 are constituted by N channel type MOSFETs in
Number | Date | Country | Kind |
---|---|---|---|
2007-137233 | May 2007 | JP | national |