Amplifiers – With periodic switching input-output
Reexamination Certificate
2002-04-16
2003-09-30
Choe, Henry (Department: 2817)
Amplifiers
With periodic switching input-output
C330S069000, C330S109000, C327S124000
Reexamination Certificate
active
06628164
ABSTRACT:
TECHNICAL FIELD
The present invention relates generally to electronic circuits and, more particularly, to circuits having variable amplification of an input signal.
BACKGROUND OF THE INVENTION
In many signal processing applications there exists a need for providing exponential gain variations based on a linearly varying input control. Exponential gain variation implies that each increment in the control signal translates into a multiplication of the present gain value by a fixed quantity. Where this gain is controlled so as to vary based on a programmed input code or a control signal, the amplifier can be considered to a programmable gain amplifier.
One approach to providing exponential gain variation may take into account the exponential dependence of collector current on the base to emitter voltage of a bipolar junction transistor (BJT) device. That is, a BJT device may be used for gain control. One example of such an approach is shown in the publication, “Comlinear CLC520 Amplifier with Voltage Controlled Gain”, National Semiconductor Corporation (NSC) Data Sheet, August 1996.
FIG. 1A
shows exponential gain characteristics, as illustrated in the above referenced NSC data sheet. The graph of
FIG. 1A
includes a graph
120
that shows a gain versus the input voltage, Vg. The gain may be the ratio between an amplifier output voltage and the input voltage Vg.
FIG. 1B
shows these same characteristics, but with the gain measured in decibels (dB), which equals 20log(Vout/Vin). The Gain(dB) versus input voltage (Vg) curve is labeled
124
.
While BJT gain control approaches, and the like, can provide exponential gain control, alternate approaches can provide a piecewise linear approximation to the exponential gain. Such approaches can include cascaded attenuators. An example of a conventional approximation approach is shown in “An Analog-to-Digital Processor for Camcorders and Digital Still Cameras”,
IEEE Transactions on Consumer Electronics,
Vol. 44, No. 3, August 1998, by Mike Koen. In Koen, the relative gain versus a control voltage is as shown in
FIG. 1C
as curve
128
. While Koen can provide an approximation of an exponential gain control, such an approach may be limited by noise requirements. It would be desirable to arrive at a more robust way of providing exponential gain control.
Still other approaches to approximating exponential linear gain control can include amplifier circuit that include switched capacitor networks. Accordingly, by way of further background, some basic principles of switched-capacitor networks will now be described. Typically, switched-capacitor networks can include metal-oxide-semiconductor (MOS) type switches. In MOS technology, it is relatively easy to implement switches, capacitors, and operational amplifiers (op amps). However, it can be difficult to construct resistors with the necessary accuracy. Consequently, switched-capacitor circuits can allow for a basic resistor approximation by using two MOS switches and a capacitor.
Extensive switched-capacitor networks, particularly those that employ the use of op amps and feedback circuitry, are well known in the art. Common applications include performing certain mathematical operations. For example, op amp circuits with switched-capacitor networks can implement signal summation, differentiation, programmable gain, and integration, to name only a few.
Programmable gain amplifiers can be implemented as a circuit using many different topologies, but with different degrees of relative success. For example, a programmable gain amplifier may have an open loop configuration, where there is no feedback network present. However, open loop topologies usually suffer from a compromise of signal range and linearity. As but one example, it is believed that achieving a signal gain which is linear to a 10-bit level for a one volt scale signal for any programmed gain, can be very difficult.
Closed loop, switched-capacitor programmable gain amplifiers, which include a feedback network present, are believed to provide better signal linearity than open loop approaches. Still further, in many cases switched-capacitor networks may be easily controlled by a digital interface. This can result in improved linearity of gain control (on a log scale) compared to other schemes.
Switched-capacitor circuits can handle large input signals that can be programmed over a wide range. However, the accuracy of a switched capacitor can often be dominated by capacitor matching. Thus, in many conventional approaches, in order to achieve exponential gains, exponentially varying capacitor sizes are used. Unfortunately, it can be difficult to design with exponentially varying capacitor sizes because of silicon area and power requirements.
A programmable gain amplifier that may include a switched-capacitor network of unit capacitors is shown in U.S. patent application Ser. No. 09/354,461, filed on Jul. 15, 1999 and titled “A Capacitor-Based Exponential Programmable Gain Amplifier” (referred to herein as Application 354,461).
In general, it can be possible to implement exponential gain variation with the approximation log
e
(1+x)/(1−x)=~2x, where |x|<1 is utilized. Here, x varies linearly and can correspond to the input gain setting code. A switched-capacitor stage of a programmable gain amplifier can implement a gain according relationship (a+x)/(a−x), which can be seen to vary exponentially with x. Thus, for a switched-capacitor gain stage, the gain can be determined as the ratio of the number of unit capacitors used to sample an input to the number used for feedback. In the particular equation described above, a sampling capacitance can be represented by a gain numerator term (a+x), while a feedback capacitance can be represented by a gain denominator term (a−x). Accordingly, the number of unit capacitors used for sampling corresponds to (a+x) and those used for feedback correspond to (a−x).
In an approach such as Application 354,461, a sampling capacitance can be conceptualized as including (a−x) and 2x capacitors, totalling (a+x) capacitors. As is understood the feedback capacitance (a−x) is included in the sampling capacitance term. Such a splitting of terms can enable implementation of the programmable gain amplifier with a reduced number of unit capacitors. One such implementation is shown schematically in FIG.
2
A. It is understood that during the operation of a circuit, switching may result in the use of some unit capacitors and the non-use of others. Unused capacitors in any gain setting can be left connected to an op amp input node, and thereby serve to reduce the variation in the feedback-factor of the closed loop amplifier.
Referring again to
FIG. 2A
, a schematic diagram of a programmable gain amplifier according to the prior art will be described in more detail. The conventional programmable gain amplifier circuit is designated by the general reference character
200
and is shown to include an operational amplifier (op amp)
202
, capacitive switching circuits (
204
and
206
), switches (
208
and
210
), feedback switch
212
, switch
214
, and a sample precharge switch
216
.
The op amp
202
has a noninverting input connected to a charge summing node
218
. The op amp
202
has an inverting input connected to node
230
. The inverting output of the op amp
202
is connected to an analog output terminal
220
, which is also labeled as Vout+, while the noninverting output of the op amp
202
is connected to an analog output terminal
232
. Analog output node
220
is connected to the closed position input terminal of feedback switch
212
.
The programmable gain amplifier receives an input signal Vin+ at analog input terminal
226
. The analog input terminal
226
is connected to the closed position input terminals sample switches
208
and
210
. A ground terminal
228
is connected to the closed position input terminal of switch
214
. The output terminals of feedback switch
212
and sample switch
208
are connected to
Bilhan Haydar
Han Yong
Lee Gary
Ramesh M. C.
Tsay Ching-Yuh
Brady III W. James
Choe Henry
Moore J. Dennis
Telecky , Jr. Frederick J.
Texas Instruments Incorporated
LandOfFree
Method and apparatus for exponential gain variations with a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for exponential gain variations with a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for exponential gain variations with a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3003411