Related Applications
This is a Continuation-In-Part of U.S. patent application Ser. No. 60/006,117, filed Oct. 23, 1995, and entitled "High Speed Variable Gain Amplifier."
Background of the Invention
1. Field of the Invention
The present invention relates to analog amplifiers, and in particular, to analog amplifiers in which complementary metal oxide semiconductor field effect transistors (CMOSFETs) and bipolar junction transistors (BJTs) are used together so as to take advantage of the high input impedances available with MOSFETs and current multiplication capabilities of BJTs.
2. Description of the Related Art
Fast settling, variable (e.g. switchable) gain amplifiers present many design challenges. For example, such an amplifier must often have a high input impedance, a tunable gain control, the ability to switch gain and settle fast, and the ability to recover quickly when overdriven by a large input signal.
One such type of amplifier is often referred to as a Gilbert multiplier, or Gilbert cell. Gilbert multipliers work well for applications that use large power supply voltages and provide a tunable gain control. However, such amplifiers rely upon BJTs and thus have bipolar inputs for which the input impedances are relatively low, and their gain switching capabilities are quite limited.
Summary of the Invention
A high speed variable gain amplifier in accordance with the present invention uses an open loop architecture for settling time verses gain requirements, and includes a bias generator circuit, a gain control circuit and a core amplifier circuit. Such an amplifier can be implemented in any BiCMOS technology, e.g. an N-Well CMOS and NPN based process, and includes a transconductance ("GM") boost circuit for improving the linearity of its voltage-to-current ("V-I") converter and providing an output current, and a current amplifier stage with a switchable folded Gilbert cell and a translated input current to output voltage suitable for applications requiring low power supply voltages. Gain control is accomplished using an external voltage and a current switching MOSFET. Input signal overdrive is controlled by current starving, thereby providing for fast recovery. High input impedance is provided for facilitating applications requiring AC coupling.
In accordance with one embodiment of the present invention, a metal oxide semiconductor field effect transistor (MOSFET) amplifier with a boosted transconductance includes two transistors and a current source. The first transistor has gate, drain and source terminals and is for receiving an input voltage via its gate terminal, conducting a first drain-to-source current via its drain and source terminals, and having a first gate-to-source voltage across its gate and source terminals. The second transistor has gate, drain and source terminals and is for conducting a second drain-to-source current via its drain and source terminals. The second transistor gate terminal is connected to the first transistor drain terminal and the second transistor drain terminal is connected to the first transistor source terminal. The current source is connected between the second transistor gate and source terminals and is for providing a fixed current as the first transistor drain-to-source current. The first transistor gate-to-source voltage remains fixed in accordance with the fixed first transistor drain-to-source current and the second transistor drain-to-source current is generated in accordance with the input voltage.
In accordance with another embodiment of the present invention, a differential current mirror circuit includes first and second circuit nodes, first and second input circuit branches, first and second auxiliary circuit branches, a bridging resistance, and first and second output circuit branches. The first input circuit branch is connected to the first circuit node and is for receiving and conducting a first input branch current from the first circuit node and receiving a first input voltage and in accordance therewith providing a first output voltage. The first auxiliary circuit branch is connected to the first circuit node and is for receiving the first output voltage and in accordance therewith receiving and conducting a first auxiliary branch current from the first circuit node. The second input circuit branch is connected to the second circuit node and is for receiving and conducting a second input branch current from the second circuit node and receiving a second input voltage and in accordance therewith providing a second output voltage. The second auxiliary circuit branch is connected to the second circuit node and is for receiving the second output voltage and in accordance therewith receiving and conducting a second auxiliary branch current from the second circuit node. The bridging resistance is connected between the first and second circuit nodes and is for conducting a bridging current which equals one half of the difference between the first and second auxiliary branch currents, is substantially zero when the first and second input voltages are equal, and is nonzero when the first and second input voltages are unequal. The first output circuit branch is connected to the first input circuit branch and is for receiving the first output voltage and in accordance therewith conducting a first output current which is proportional to the first auxiliary branch current. The second output circuit branch is connected to the second input circuit branch and is for receiving the second output voltage and in accordance therewith conducting a second output current which is proportional to the second auxiliary branch current.
These and other features and advantages of the present invention will be understood upon consideration of the following detailed description of the invention and the accompanying drawings.
Brief Description of the Drawings
FIG. 1 is a high level functional block diagram of a high speed switching gain amplifier in accordance with one embodiment of the present invention.
FIG. 2 is a schematic diagram of the bias generator circuit of FIG. 1.
FIG. 3 is a schematic diagram of the gain control circuit of FIG. 1.
FIG. 4 is a schematic diagram of the core amplifier circuit of FIG. 1.
FIGS. 4A, 4B, 4C, 4D, 4E and 4F together form a more detailed schematic diagram of the core amplifier circuit of FIG. 4.
FIG. 5 is a functional block diagram representing the controllable current gain feature of the core amplifier circuit of FIG. 1.
Detailed Description of the Invention
Referring to FIG. 1, a high speed variable gain amplifier in accordance with the present invention includes a bias generator circuit VBVGA, a gain control circuit GSVGA and a core amplifier circuit VGACORE. The bias generator circuit VBVGA uses replica bias techniques to establish and provide bias voltages VBP, VCP, VCN, VBN to the amplifier circuit VGACORE. The gain control circuit GSVGA provides two control currents Iu, Id as a multiplying signal to a modified Gilbert cell within the amplifier circuit VGACORE (discussed in more detail below) to control the gain of the amplifier circuit VGACORE. The amplifier circuit VGACORE amplifies the input signal IN by the gain control signal Iu, Id and provides clamping and gain switching functions.
Referring to FIG. 2, the bias generator circuit VBVGA uses replica bias techniques to create a minimum bias cascode current source structure, thereby maximizing the dynamic signal range. Whereas a number of techniques exist for producing a cascode current source, the preferred method is as shown in FIG. 2.
Referring to FIG. 3, the gain control circuit includes two voltage-to-current (V-I) converters and a current summing circuit. The input voltages VSET, ACOM to the V-I converters can be derived from a resistor divider (not shown) biased with a reference voltage. It can be shown that the output currents Id and Iu are established in accordance with the input voltages as follows:
Referring to FIG. 4, the core amplifier also includes a V-I converter, composed primarily of MOSFETs M1-M14, followed by an enhanced and folded Gilbert multiplier, composed primarily of BJTs Q6-Q9. The output current difference of the Gilbert multiplier is further folded in a double-to-single-ended cascode current stage, composed primarily of MOSFETs M17-M24. The output current of the cascode current stage flows into a resistor R2 and gets converted into an output voltage, i.e. the voltage across this resistor R2 is buffered by an emitter follower buffer (Q12, M25, M26) to provide the output OUT.
The V-I converter includes transconductance-boosted P-MOSFET input transistors M1 and M2, P-MOSFET current source transistors M3-M6, resistor R1, N-MOSFET transconductance-booster transistors M7 and M8, and N-MOSFET current mirror transistors M9-M14. The transconductances (GM) of the input transistors P-MOSFET M1 and P-MOSFET M2 are boosted by the feedback loops made up of transistors M1, M9, M11 and M7 on the input side and transistors M2, M10, M12 and M8 on the reference ACOM side of the differential input pair. The boost in GM for the P-MOSFET input devices M1, M2 enables a more accurate conversion of the input voltage IN-ACOM to an output current. The differential input voltage, i.e. the voltage difference between the input voltages Vin and ACOM, causes a current to be conducted through resistor R1. This difference current is mirrored by the boost transistors M7, M8 and current mirror transistors M13, M14 to the common emitter nodes of the Gilbert multiplier transistors Q6-Q9.
Transistors Q1-Q5 provide a bias voltage to the base of the folded Gilbert multiplier transistors Q6-Q9. This bias voltage is approximately equal to the input reference voltage ACOM, and ensures that all of the NPN transistors Q6-Q11 that form a current amplifier stage have approximately the same bias conditions. The output of the current amplifier stage is formed by the sum of two currents, one current directly from the output of the V-I converter and the other from a multiple of the V-I converter output current from the Gilbert multiplier cell. By way of a gain enable signal GB, transistors M27 and M28 enable a high speed gain boost by adding such multiple current to the output. From the input signal IN side, the collector currents of BJTs Q6 and Q10 sum to form one Gilbert multiplier output current (I+aI), while from the reference ACOM side, the collector currents of BJTs Q8 and Q11 sum to form the other Gilbert multiplier output current -(I+aI). Transistors Q6 and Q7 form a Gilbert multiplier and produce a multiple of the drain current of transistor M13 on the input signal IN side, while transistors Q8 and Q9 form another Gilbert multiplier and produce a multiple of the drain current of transistor M14 on the input ACOM side. The multiplier action is achieved by converting the control current difference Iu-Id to a difference in the base-emitter voltage (VBE) of biasing transistors Q4 and Q5. This VBE voltage difference is further converted into a current difference by the Gilbert multiplier transistor pairs Q6/Q7 and Q8/Q9. The multiple current outputs (I+aI) and -(I+aI) from the current amplifier stage go to the inputs of the differential-to-single-ended current converter.
It can be shown that the "a" term can be expressed as follows:
The differential-to-single-ended current converter includes transistors M17-M24, with P-MOSFET transistors M17-M20 forming a cascode input current pair and N-MOSFET transistors M21-M24 forming a cascoded double-to-single current mirror. The output current of the differential-to-single-ended current converter flows through resistor R2 to produce an unbuffered output voltage. An emitter follower buffer formed by transistor Q12 and current source M25-M26 performs unity gain buffering so that the output node has a low impedance for driving various loads.
A large input, or overdrive, signal in either direction will not deplete the current in the P-MOSFET M3-M6 and N-MOSFET M9-M12 current sources in the V-I converter or in the cascode current sources M17-M20 in the differential-to-single-ended current converter. Since none of the current sources mm off, no transistor goes out of its normal operating bias condition; therefore, recovery from an overdriven input signal is fast. The current in R2 is limited by the difference in input voltage across R1.
Referring to FIG. 5, the controllable current gain feature of the core amplifier circuit of FIG. 1 can be represented as shown, i.e. as a serial arrangement of a voltage-to-current (V-I) converter and current-to-voltage (I-V) converter with a current gain control stage interposed between them. The differential input voltage IN-ACOM is converted by transistors M1-M16 to a set of differential currents which are selectively, in accordance with gain control signals Iu and Id, multiplied by transistors Q6-Q11. Of the resulting, selectively multiplied, differential current (I+aI)/-(I+aI), current phase (I+aI) is formed by summing its current components I and aI at the node connecting the drain and source terminals of transistors M19 and M17, respectively, while current phase -(I+aI) is formed by summing its current components -I and -aI at the node connecting the drain and source terminals of transistors M20 and M18, respectively. This differential current is then converted by transistors M17-M24 and resistor R2 to a single-ended voltage which is buffered by transistors Q12, M25 and M26 to provide the final output voltage OUT.
Various other modifications and alterations in the structure and method of operation of this invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. It is intended that the following claims define the scope of the present invention and that structures and methods within the scope of these claims and their equivalents be covered thereby.