Predictive sigma-delta ADC filter for power consumption
A sigma-delta analog-to-digital converter comprises a sigma-delta modulator; and an ADC filter that receives a segment of L binary samples from the sigma-delta modulator, L being a positive integer. The ADC filter includes a predictor circuit that determines whether a power consumption of a first summing method would be higher than a power consumption of a second summing method, based on a content of the segment, and an accumulator circuit that calculates an output using the first summing method in a case where the power consumption of the first summing method would be lower than the power consumption of the second summing method, and using the second summing method in a case where the power consumption of the first summing method would be higher than the power consumption of the second summing method.
1. A sigma-delta analog-to-digital converter, comprising:
a sigma-delta modulator; and
an ADC filter configured to receive a segment of L binary samples from the sigma-delta modulator, L being a positive integer, including:
a predictor circuit configured to determine whether a power consumption of a first summing method would be higher than a power consumption of a second summing method, based on a content of the segment, and
an accumulator circuit configured to calculate an output using the first summing method in a case where the power consumption of the first summing method would be lower than the power consumption of the second summing method, and using the second summing method in a case where the power consumption of the first summing method would be higher than the power consumption of the second summing method.
2. The sigma-delta analog-to-digital converter according to claim 1 , wherein the first summing method is a summing up method, and the second summing method is a summing down method.
3. The sigma-delta analog-to-digital converter according to claim 2 , wherein the predictor circuit is configured to analyze a set of K bits of the segment of L binary samples, K being a positive integer much smaller than L.
4. The sigma-delta analog-to-digital converter according to claim 3 , wherein the predictor circuit is configured to determine that the power consumption of the first summing method would be higher than the power consumption of the second summing method if the set of K bits contains more 1s than 0s, and to determine that the power consumption of the first summing method would be lower than the power consumption of the second summing method if the set of K bits contains more 0s than 1s.
5. The sigma-delta analog-to-digital converter according to claim 3 , wherein K=5.
6. The sigma-delta analog-to-digital converter according to claim 2 , wherein the accumulator circuit is configured, in a case of using the first summing method, to receive a sequence of bits from the segment of L binary samples into a temporary bit buffer, match the sequence to one of a plurality of predetermined reference sequences, and add a corresponding one of a plurality of pre-computed partial sums to a running total in the accumulator circuit.
7. The sigma-delta analog-to-digital converter according to claim 2 , wherein the accumulator circuit is configured, in a case of using the second summing method, to receive a sequence of bits from the segment of L binary samples into a temporary bit buffer, match the sequence to one of a plurality of predetermined reference sequences, and subtract a corresponding one of a plurality of pre-computed partial sums to a running total in the accumulator circuit.
8. The sigma-delta analog-to-digital converter according to claim 1 , wherein the accumulator circuit is configured to output a running total when all L binary samples have been analyzed.
9. A method of calculating an output in a sigma-delta analog-to-digital converter, comprising:
receiving, at an ADC filter of the sigma-delta analog-to-digital converter, a segment of L binary samples from a sigma-delta modulator of the sigma-delta analog-to-digital converter;
predicting, at a predictor circuit of the ADC filter, whether a power consumption of a first summing method would be higher than a power consumption of a second summing method, based on a content of the segment;
calculating the output using the first summing method in an accumulator circuit of the ADC filter, in a case where the power consumption of the first summing method would be lower than a power consumption of the second summing method; and
calculating the output using the second summing method in an accumulator circuit of the ADC filter, in a case where the power consumption of the first summing method would be higher than a power consumption of the second summing method.
10. The method according to claim 9 , wherein the first summing method is a summing up method, and the second summing method is a summing down method.
11. The method according to claim 10 , wherein the predicting includes analyzing a set of K bits of the segment of L binary samples, K being a positive integer much smaller than L.
12. The method according to claim 11 , wherein the predicting includes determining that the power consumption of the first summing method would be higher than the power consumption of the second summing method if the set of K bits contains more 1s than 0s, and determining that the power consumption of the first summing method would be lower than the power consumption of the second summing method if the set of K bits contains more 0s than 1s.
13. The method according to claim 11 , wherein K=5.
14. The method according to claim 10 , wherein calculating using the first summing method includes receiving a sequence of bits from the segment of L binary samples into a temporary bit buffer, matching the sequence to one of a plurality of predetermined reference sequences, and adding a corresponding one of a plurality of pre-computed partial sums to a running total in the accumulator circuit.
15. The method according to claim 10 , wherein calculating using the second summing method includes receiving a sequence of bits from the segment of L binary samples into a temporary bit buffer, matching the sequence to one of a plurality of predetermined reference sequences, and subtracting a corresponding one of a plurality of pre-computed partial sums to a running total in the accumulator circuit.
16. The method according to claim 9 , further comprising outputting a running total when all L binary samples have been analyzed.
17. An image sensor, comprising:
a pixel array; and
a sigma-delta analog-to-digital converter, including a sigma-delta modulator and an ADC filter configured to receive a segment of L binary samples from the sigma-delta modulator, L being a positive integer, the ADC filter having:
a predictor circuit configured to determine whether a power consumption of a first summing method would be higher than a power consumption of a second summing method, based on a content of the segment, and
an accumulator circuit configured to calculate an output using the first summing method in a case where the power consumption of the first summing method would be lower than the power consumption of the second summing method, and using the second summing method in a case where the power consumption of the first summing method would be higher than the power consumption of the second summing method.
18. The image sensor according to claim 17 , wherein the first summing method is a summing up method, and the second summing method is a summing down method.
19. The image sensor according to claim 18 , wherein the predictor circuit is configured to analyze a set of K bits of the segment of L binary samples, K being a positive integer much smaller than L.
20. The image sensor according to claim 18 , wherein the accumulator circuit is configured, to receive a sequence of bits from the segment of L binary samples into a temporary bit buffer, match the sequence to one of a plurality of predetermined reference sequences, and add or subtract a corresponding one of a plurality of pre-computed partial sums to a running total in the accumulator circuit.