II.

BASIC SWITCHED-CAPACITOR OPERATION The essence of the switched-capacitor is the use of capacitors and analog switches to perform the same function as a resistor. This replacement resistor, along with op-amp based integrators, then forms an active filter. Before delving too far into actual filter designs, however, it makes sense to ask why one would want to replace the resistor with such an apparently complex assembly of parts as switches and capacitors. It would seem from the multiplication of parts that the switched-capacitor would be area intensive. As a matter of fact, for the resistor values that one seeks in certain filter designs, this is not the case. Furthermore, the use of the switched-capacitor will be seen to give frequency tunability to active filters. Figure 1[2, 3] shows the basic setup for a switched-capacitor, including two N-channel Metal-Oxide Semiconductor FieldEffect Transistors (NMOS) and a capacitor. There are two clock phases, , which are non-overlapping. The MOSFET's, either M1 or M2, will be turned ON when the gate voltage is high, and the equivalent resistance of the channel in that case will be low, . Conversely, when the gate voltage goes LOW, the channel resistance will look like . With such a high ratio of OFF to ON resistances, each MOSFET can be taken for a switch. Furthermore, when the two MOSFET's are driven by non-overlapping clock signals, then M1 and M2 will conduct during alternate half-cycles.

Figure 1. Two NMOSFET's, driven by alternating, non-overlapping clock signals, comprise the basic switched capacitor network. This makes the two-MOSFET arrangement equivalent to a single-pole, double-throw switch (SPDT). One can now use a symbolic switch picture, as in Figure 2a below, to represent the circuit. The operation of this circuit is as follows. When the switch is thrown to the left, the capacitor will charge up to . When the switch is thrown to

the right, the capacitor will discharge down to/charge up to consecutive switching events, there will be a net of cycles/sec,

. As a result of these charge transfer of

. Now, if one flips the switch back and forth at a rate then the charge transferred in one second is , which of course has the units of current. One can

claim that an average current, . If is much higher than the frequency of the voltage waveforms, then the switching process can be taken to be essentially continuous, and the switched-capacitor can then be modeled as an equivalent resistance, as shown below in Figure 2b. The value of the equivalent resistance is given by:

(Eq. 1) Therefore, this equivalent resistance, in conjunction with other capacitors, and Opamp integrators, can be used to synthesize active filters. It is now clear from Equation (1) how the use of the switched-capacitor leads to tunability in the active filters, by varying the clock frequency.

(a)

(b)

Figure 2. Equivalent resistor model for switched capacitor circuit in Fig. 1. This equivalent resistance has features which make it advantageous when realized in integrated-circuit form: (a) High-value resistors can be implemented in very little silicon area. For example, a 1-M resistor can be realized with a 10-pF capacitor switched at a clock rate of 100 kHz.

(b) Very accurate time constants can be realized, because the time constant is proportional to the ratio of capacitances, and inversely proportional to the clock frequency:

. Capacitor ratios, especially in monolithic form, are very robust against changes in temperature, and clock frequencies can also be strictly controlled, so that accurate time constants are now available in the switched-capacitor technology. The principal constraint in using the switched-capacitor is that inherent in all sampleddata systems: the clock frequency must be much higher than the critical frequency set by the RC products in the circuit. Furthermore, on either side of the analog switches, i.e., the MOSFET's, there must be essentially zero-impedance nodes (voltage sources). There are a number of other constraints which the unsuspecting designer/user might overlook [3, p. 725]: (a) The equivalent resistance formed by the action of the switched-capacitor cannot be used to close the negative-feedback path in an op-amp all by itself. One must recall that to ensure stability, the op-amp's feedback path must be closed continuously, while the switched-capacitor is a sampled-data construction of a resistor, and thus not continuous. (b) Circuit nodes cannot be left floating. That is, there must always be a resistive path to ground so that charge does not build up on the capacitor plates. (c) The bottom plates of the MOS capacitors must be connected to ground or to a voltage source. There is an intrinsic, parasitic capacitance associated with the MOS capacitor's bottom plate [4]. This parasitic capacitance can be between 5% and 20% of the desired value; furthermore, it behaves nonlinearly with voltage [4]. Therefore, it must be connected to AC ground or a voltage source so that this nonlinear portion of the capacitance will not affect the overall response of the switched-capacitor filter. In practical terms, this means that capacitive voltage dividers with three or more capacitors, and circuits that switch both ends of a capacitor in sequence to the inputs of an op-amp, are used. (d) The noninverting pin of the op-amp should be kept at a constant voltage. If this pin is connected to the signal in some way, then the virtual short circuit between op-amp inputs means that the inverting input is no longer a virtual

ground, and so an undesirable alteration of filter response due to the MOS capacitor's parasitic capacitance will occur (see item (c) above).

OTA AMPLIFIER
The application of the operational transconductance amplifier (OTA) in the design of simple amplifiers with voltage-controllable gain and to the design of firs-order and second-order active filters with controllable gains and controllable critical frequencies is demonstrated. A typical biasing scheme is also shown so that readers can more easily set up the circuits themselves. Introduction This paper demonstrates the usefulness of the operational transconductance amplifier (OTA) as a replacement for the conventional op-amp in both first and second-order active filters. It is at least partially intended to acquaint the technology student w ith the rudiments of operation of the OTA, as well as the practicalities of using the presently available commercial OTAs. The structure of this paper is as follows. First, the basic operation of the OTA, including DC and AC operation, is explained. Simple example circuits will be presented in this section to demonstrate the similarities and differences between circuits w hich use the conventional op-amp and the OTA. Second, OTA active filter circuits will be presented and analyzed. The richness of the filter possibilities inherent in the second-order structures will be demonstrated. Finally, the last section will presen t practical considerations which must be considered when using the present generation of OTAs. II. Basic OTA Operation 1. DC Operation The OTA is a transconductance type device, which means that the input voltage controls an output current by means of the device transconductance, labeled gm. This makes the OTA a voltage-controlled current source (VCCS), which is in contrast to the conventional op-amp, which is a voltage-controlled voltage source (VCVS). What

is important and useful about the OTAs transconductance parameter is that it is controlled by an external current, the amplifier bias current, I ABC , so that one obtains (Eq. 1) From this externally controlled transconductance, the output current as a function of the applied voltage difference between the two input pins, labeled v+ and v-, is given by (Eq. 2) Clearly, an output voltage can be derived from this current by simply driving a resistive load. The equivalent circuit for the OTA is shown in Figure 1.

Fig. 1 At this point, two key differences between the OTA and the conventional op-amp must be kept in mind. First, since the OTA is a current source, the output impedance of the device is high, in contrast to the op-amps very low output impedance. Because a low output impedance is often a desirable trait in general amplifiers used to drive resistive loads, certain of the newer commercial OTAs, such as National Semiconductors LM13600, have on-chip controlled impedance buffers. Second, it is possible to desi gn circuits using the OTA that do not employ negative feedback. In other words, instead of employing feedback to reduce the sensitivity of a circuits performance to device parameters, the transconductance is treated as a design parameter, much as resisto rs and capacitors are treated in op-amp based circuits.

The biasing of the OTAs internal circuitry is such that the total quiescent supply current [Soclof-91] is given by ISUPPLY = 3IABC. This seems to imply that the OTA can be used in micropower applications, even down to IABC = 1 A. However, the losses in speed and bandwidth, which are controlled ultimately by IABC, can be severe at such low current levels. 2. AC Analysis and Frequency Response Much of the dependence of open- and closed-loop bandwidth and frequency responses in the OTA are similar to those in the conventional op-amp. For a circuit employing negative feedback, a very important relationship between the closed-loop bandwidth, the amplifier bias current, and the closed-loop gain exists: (Eq. 3) where CNET is the sum of device junction capacitances at the output of the OTA and whatever load capacitance is attached to the circuit: CNET = CO + CL. Equation (3) has the interesting consequence that cer tain types of active networks, such as active filters, can have their critical frequencies controlled by the external current, I ABC, which of course can be in turn controlled by an external voltage.

OR

The operational transconductance amplifier (OTA) is an amplifier whose differential input voltage produces an output current. Thus, it is a voltage controlled current source (VCCS). There is usually an additional input for a current to control the amplifier's transconductance. The OTA is similar to a standard operational amplifier in that it has a high impedancedifferential input stage and that it may be used with negative feedback.[1]

The first commercially available integrated circuit units were produced by RCA in 1969 (before being acquired by General Electric), in the form of the CA3080, and they have been improved since that time. Although most units are constructed with bipolar transistors, field effect transistor units are also produced. The OTA is not as useful by itself in the vast majority of standard op-amp functions as the ordinary opamp because its output is a current. One of its principal uses is in implementing electronically controlled applications such as variable frequency oscillators and filters and variable gain amplifier stages which are more difficult to implement with standard op-amps.

Principal differences from standard operational amplifiers

Its output of a current contrasts to that of standard operational amplifier whose output is a voltage. It is usually used "open-loop"; without negative feedback in linear applications. This is possible because the magnitude of the resistance attached to its output controls its output voltage. Therefore a resistance can be chosen that keeps the output from going into saturation, even with high differential input voltages.

Basic operation
In the ideal OTA, the output current is a linear function of the differential input voltage, calculated as follows:

where Vin+ is the voltage at the non-inverting input, Vin is the voltage at the inverting input and gm is the transconductance of the amplifier. The amplifier's output voltage is the product of its output current and its load resistance:

The voltage gain is then the output voltage divided by the differential input voltage:

The transconductance of the amplifier is usually controlled by an input current, denoted Iabc ("amplifier bias current"). The amplifier's transconductance is directly proportional to this current. This is the feature that makes it useful for electronic control of amplifier gain, etc. [edit]Non-ideal

characteristics

As with the standard op-amp, practical OTA's have some non-ideal characteristics. These include: Input stage non-linearity at higher differential input voltages due to the characteristics of the input stage transistors. In the early devices, such as the CA 3080, the input stage consisted of two bipolar transistors connected in the differential amplifier configuration. The transfer characteristics of this connection are approximately linear for differential [2] input voltages of 20 mV or less. This is an important limitation when the OTA is being used open loop as there is no negative feedback to linearize the output. One scheme to improve this parameter is mentioned below. Temperature sensitivity of transconductance. Variation of input and output impedance, input bias current and input offset voltage with the transconductance control current Iabc.

[edit]Subsequent

improvements

Earlier versions of the OTA had neither the Ibias terminal shown in the diagram nor the diodes shown adjacent to it. They were all added in later versions. As depicted in the diagram, the anodes of the diodes are attached together and the cathode of one is attached to the non inverting input (Vin+) and the cathode of the other to the inverting input (Vin). The diodes are biased at the anodes by a current (Ibias) that is injected into the Ibias terminal. These additions make two substantial improvements to the OTA. First, when used with input resistors, the diodes distort the differential input voltage to offset a significant amount of input stage non linearity at higher differential input voltages. According to National Semiconductor, the addition of these diodes increases the linearity of the input stage by a factor of 4. That is, using the diodes, the signal distortion level at 80 mV of differential input is the same as that [3] of the simple differential amplifier at a differential input of 20 mV. Second, the action of the biased diodes offsets much of the temperature sensitivity of the OTA's transconductance. A second improvement is the integration of an optional-use output buffer amplifier to the chip on which the OTA resides. This is actually a convenience to a circuit designer rather than an improvement to the OTA itself; dispensing with the need to employ a separate buffer. It also allows the OTA to be used as a traditional op-amp, if desired, by converting its output current to a voltage.

An example of a chip combining both of these features is the National Semiconductor [4] LM13600 and its successor, the LM13700, the data sheet for which can be found here: