Coded data generation or conversion – Analog to or from digital conversion – Analog to digital conversion
Reexamination Certificate
1999-07-30
2002-08-13
Tokar, Michael (Department: 2819)
Coded data generation or conversion
Analog to or from digital conversion
Analog to digital conversion
C341S131000
Reexamination Certificate
active
06433723
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to analog-to-digital conversion devices, and in particular, those devices which exhibit error or distortion due to differential non-linearity.
BACKGROUND OF THE INVENTION
Devices that convert analog signals into digital signals for use in digital processing circuitry are commonly referred to as analog-to-digital (“A/D”) converters. A/D converters are typically integrated circuits that attempt to approximate the instantaneous value of an analog input signal and assign a digital value to that approximation.
The dynamic range of the A/D converter defines the range of analog input voltage values that may practicably be converted. For example, an A/D converter may be configured to convert signals within a dynamic range of −5V to +5V. In such a case, the A/D converter provides useful digital values only for those input voltages that are within that dynamic range. For any voltage values of below −5V, the A/D converter may simply provide the a zero digital value, and for any voltage above +5V, the A/D converter may provide its maximum digital value.
The sampling rate of the A/D converter typically defines the rate at which the A/D converter produces output digital values or samples. A sampling rate of 10,000 samples/second generally means that the A/D converter measures the input analog signal 10,000 times every second, and produces 10,000 corresponding digital values.
The resolution of the A/D converter refers to the number of different possible digital values that the A/D converter can produce for the dynamic range. An N-bit resolution provides a potential range of 2
N
values. Thus, an A/D converter having an eight bit resolution produces a range of 256 output values. Each output value is referred to as a quantization step.
Ideally, the digital values produced by the A/D converter are linearly distributed over the dynamic range of the converter. The difference in the input voltage levels represented by adjacent digital output values is referred to as the quantization step value of the converter. Thus, for example, an 8-bit converter having a dynamic range of −5V to +5V would have 256 values to represent the scale of voltages from −5V to +5V. In such an example, the quantization step value would be 10/256 volts or 0.039 volts. Thus, in such a converter, the digital value 0 would represent an analog input voltage of −5.0 to −4.962, the digital value 1 would represent an analog input voltage of −4.961 to −4.923, and so forth.
A/D converters are susceptible to error from many causes, including differential non-linearity. For example, successive approximation A/D converters exhibit significant amounts of differential non-linearity, particularly in low cost implementations of the circuit. Differential non-linearity is a phenomenon in which the range of analog input voltages that generate each digital output value is not uniform for all digital output values.
In particular, as discussed above A/D converters ideally approximate analog inputs by allocating an equivalent input voltage range to each digital output value. In the above example, each digital output value ideally represents a voltage range of approximately 0.039 volts. As a result, the 8-bit converter “covers” the dynamic range of ten volts with 256 digital values.
Because of differential non-linearity, however, individual digital values will be generated for ranges of input voltage values that is more than or less than the ideal voltage range. Thus, referring again to the above example, the digital value 0 may represent a non-ideal range of input voltages from −5V to −4.958V and the digital value 1 may represent a non-ideal range of input voltages from −4.959V to −4.923V. In many circuits, however, a significant amount of differential non-linearity error is tolerable.
Nevertheless, in some circuits, differential non-linearity can result in inaccuracies that are detrimental to the operation of a system. For example, a measurement device that is intended to measure various aspects an input analog signal can provide inaccurate measurements due in part to the differential non-linearity of an A/D converter. One example of such a measurement device is an electrical utility meter.
Utility service providers employ electrical utility meters to measure various values related to energy consumption. The service providers use the measured values for billing and resource allocation purposes, among other things. Because the measured values are used for billing, high standards of accuracy have evolved for the electrical utility meter industry.
Many electrical utility meters utilize A/D converters to convert input analog measurement signals to digital measurement values. The input analog measurement signals are typically signals representative of the actual current and voltage waveforms on the power lines being metered. The A/D converter digitizes the signals and then provides the digital measurement values to digital processing circuitry. The digital processing circuitry performs mathematical operations on the digital measurement values to generate the various metering quantities used by the service provider for billing and other purposes.
In such meters, the differential non-linearity error of the A/D converter can cause error in the measurements that exceeds the allowable standards. One solution to the problems raised by differential non-linearity error in A/D converters is to formulate the A/D converter with high quality, low tolerance components. However, such A/D converters can be significantly more expensive than ordinary successive approximation converters.
A need exists, therefore, for an A/D converter circuit that exhibits a reduced amount of differential non-linear error or distortion without employing high cost, low tolerance components in the converter design.
SUMMARY OF THE INVENTION
The present invention addresses the above need, as well as others, by providing an A/D converter circuit that adds a relatively high frequency noise signal to the input analog signal to spread the differential non-linear error over larger portions of the dynamic range of the A/D conversion device. In addition, the A/D conversion device oversamples the combined analog signal and error signal, thereby allowing the effective resolution and accuracy of the converter circuit to exceed that of the A/D conversion device.
In one embodiment, the present invention includes an analog-to-digital converter for generating samples at a first sample rate. The analog-to-digital converter includes a noise generator, a summing circuit, a conversion circuit, and a low pass filter. The noise generator is configured to generate a noise signal having a signal bandwidth, the signal bandwidth including a high frequency that exceeds the first sample rate. The summing circuit is configured to sum the noise signal with an input signal in order to generate a composite signal. The conversion circuit is configured to convert the composite signal to a first digital signal, the conversion circuit using a sampling rate that exceeds the first sample rate. The low pass filter is operable to filter the first digital signal and generate a second digital signal having the first sample rate.
An exemplary method of converting analog input signals to digital output signals according to the present invention includes a first step of generating a noise signal having a signal bandwidth, the signal bandwidth including a high frequency that exceeds a first sample rate. The exemplary method further includes the steps of summing the noise signal with an input signal in order to generate a composite signal, and converting the composite signal to a first digital signal using a sampling rate that exceeds the first sample rate. Finally, the exemplary method includes the step of filtering the first digital signal and generating a second digital signal having the first sample rate.
The conversion of the input signal after the addition of the noise signa
Cho James H.
Siemens Power Transmission & Distribution Inc.
Tokar Michael
LandOfFree
Analog-to-digital conversion with reduced error does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Analog-to-digital conversion with reduced error, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Analog-to-digital conversion with reduced error will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2952594