Coded data generation or conversion – Analog to or from digital conversion – Digital to analog conversion
Reexamination Certificate
2000-05-31
2002-05-07
Jeanpierre, Peguy (Department: 2819)
Coded data generation or conversion
Analog to or from digital conversion
Digital to analog conversion
C341S153000
Reexamination Certificate
active
06384763
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
The present invention relates in general to digital-to-analog converters, and more particularly to the segmented type of converter having multiple resistor sections for carying out the conversion process.
BACKGROUND OF THE INVENTION
In mixed signal circuits which involve both analog and digital signals, circuits are generally required for converting the analog signals into corresponding digital signals, and vice versa. Digital-to-analog converters provide an analog output as a function of the digital input. Many different varieties of conversion circuits are commercially available to satisfy the various constraints required, such a speed, resolution, differential non-linearity, monotinicity, etc. The digital-to-analog conversion process can be carried out according to different techniques, including the use of weighted current sources, R-2 ladder networks, as well as many other well-known conversion schemes. Because of the wide utilization of microprocessors employed to process digital information, it is a common practice to integrate digital-to-analog converters on the microprocessor chip. Because the use of chip area is always a concern, the minimization of components is therefore of paramount importance.
In a rudimentary digital-to-analog converter having a 12-bit resolution, as many as 4,096 series resistors can be utilized to produce a different magnitude of analog voltage in response to each of the 4,096 digital words. The amount of semiconductor space required for all these resistors would be prohibitively large. In addition to each resistor, there are required corresponding switches across each resistor for selecting voltage levels in response to different input digital combinations.
A segmented digital-to-analog converter (DAC) provides an adequate solution to the problem of a large number of resistors to carry out the conversion algorithm.
FIG. 1
of the drawings illustrates two resistor segments or strings of a 12-bit DAC
10
. The digital-to-analog converter
10
includes a main DAC
12
and a subsidiary (“sub”) DAC
18
. The main DAC
12
includes a number of series resistors to provide 2
x
different analog levels in response to X most significant digital bits.
A sub-DAC resistor string
18
includes a number of series-connected resistors to provide 2
y
different analog levels for the least significant bits of the DAC
10
. The DAC
10
includes X+Y digital input bits, and produces 2(
x+y)
analog levels. A fewer number of resistors are required in a segmented DAC driven by a corresponding number of digital input bits.
A first switch multiplexer
20
is connected between the main DAC resistor string
12
and the sub-DAC resistor string
18
. The switch multiplexer
20
is of conventional design for allowing the sub-DAC resistor string
18
to be connected in parallel to any one or more of the resistors in the main DAC resistor string
12
. The switch multiplexer
20
is required to provide connections to 2
x
different resistor combinations in the main DAC
12
.
The sub-DAC
18
also includes a switch multiplexer
22
for selecting 2
y
different resistance values. The output
24
of the second switch multiplexer
22
is connected to an operational amplifier
26
. An output
28
of the amplifier
26
provides 2
(x+y)
different analog outputs corresponding to the different combinations of the X+Y digital bits applied to the DAC
10
.
While the switch connections between the main DAC resistor string
12
and the sub-DAC resistor string
18
provides a multiplying function and reduce the number of resistors required to complete the X+Y bit conversion, various shortcomings of this arrangement exist. For example, the coupling of the sub-DAC resistor string
18
to the main DAC resistor string
12
can present an unbalanced load thereon, as a function of the overall resistance of the sub-DAC resistor string
18
. This can occur when the individual resistors of the string
18
are switched in or out of the circuit. When an unbalanced load is connected across the main DAC resistor string, a nonlinear conversion results. While
FIG. 1
illustrates that only the main DAC resistor string
12
is supplied with a reference supply voltage, the same loading factor exists when the sub-DAC resistor string
18
is independently powered by a reference supply voltage. Various attempts have been made in the prior art to overcome this loading problem between the main and sub-DAC resistor strings. As noted in the background portion of U.S. Pat. No. 4,338,591 by Tuthill, there is proposed the remedy of placing a buffer amplifier between the main DAC resistor string and the sub-DAC resistor string. The buffer amplifiers do effectively isolate the main and sub-DAC resistor strings. But, substantial semiconductor area is required to isolate the resistor strings with a pair of buffer amplifiers. Also, the dynamic range of the main DAC is severely limited by the input range and the finite common mode rejection of the buffer amplifiers.
Instead of isolating the main DAC and the sub-DAC resistor strings with buffer amplifiers, the use of current sources is suggested in U.S. Pat. No. 5,703,588 by Rivoir et al. By utilizing a constant current to drive the main DAC and another constant current source to drive the sub-DAC, a balanced operation therebetween can be accomplished so that neither resistor string draws current from the other. The loading problem is thus alleviated, irrespective of the switch connections. When utilizing current sources, it is imperative that the output impedance thereof is some orders of magnitude higher than the voltage of the resistor string being driven. Otherwise, inaccuracies in the conversion process become significant, especially when large voltage excursions in the sub-DAC are experienced.
Current mirrors are well known for use as current sources and current sinks in DAC resistor strings. While accurate current control can be achieved, the output impedance of such a structure is not always as high as desired. Utilizing two transistors in series as either a current source or a current sink in a DAC resistor string could increase the output impedance of the current source by a factor of the gain of the second transistor. This solution can cause other problems.
It can be seen from the DAC
10
shown in
FIG. 1
that, depending on the switch setting of the switch multiplexers
20
and
22
, analog voltages very near the reference voltage, or very near the circuit common voltage (ground) can be coupled to the output amplifier
26
. Unless expensive, precision instrumentation amplifiers are utilized, a wide dynamic input range (rail-to-rail) of amplifiers is not always available. However, it is most desirable to design DAC resistor strings that operate “rail-to-rail”, otherwise wasted voltage ranges due to headroom resistors must be used. In other words, to reduce the dynamic range over which the amplifier must operate, resistors can simply be placed in series at the top and/or bottom of the DAC resistor strings. Such resistors waste power and require additional semiconductor area.
The accuracy in the conversion of the digital input to an analog output is a function of the values of the resistors with which the resistor strings are formed in the semiconductor material. While exact value resistors are difficult to form in integrated circuits, the repeatability of making a nominal resistance value is high.
The value of the DAC resistors themselves is only one consideration in forming or scaling accurate resistance values for the resistor strings. Another consideration is the contact resistance for each semiconductor resistor. Since each resistor has two contacts, there may be a contact resistance in series with each semiconductor resistor. U.S. Pat. No. 5,343,199 illustrates DAC resistor strings where the various segments of a resistor string are connected together by metal contacts. When utilizing resistors of different values, the contact resistance presents a problem when attempting to scale the resistor va
Holberg Douglas R.
Leung Ka Y.
Cygnal Integrated Products, Inc.
Howison Thoma & Arnott, L.L.P.
Jeanpierre Peguy
Lauture Joseph
LandOfFree
Segemented D/A converter with enhanced dynamic range does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Segemented D/A converter with enhanced dynamic range, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Segemented D/A converter with enhanced dynamic range will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2837565