A/D converter correction scheme
AssigneePhotobit Corporation
InventorHarry Langenbacher
Life of this patent
Estimating the term…
Abstract
Differential nonlinearity is corrected by a look up table. The look up table obtains one extra bit of information and uses it to address corrected information in the look up table. The corrected look up table information represents information that is corrected for a differential nonlinearity error.
Claims20
A corrected A to D converter, comprising:
- an A to D converter unit, receiving an analog signal and producing an uncorrected digital output signal that is indicative of the analog signal;
- and
- a memory, storing a relationship between said uncorrected digital output and a corrected digital output as corrected to compensate for differential non linearity errors, said memory connected to receive information indicating said uncorrected digital output signal, and to output said corrected digital output which corresponds to said uncorrected digital output wherein said uncorrected digital output has more bits than said corrected digital output.
A method of correcting for an inaccuracy in a converter circuit, comprising:
- determining information about inaccuracy in an uncorrected A to D circuit;
- using said uncorrected A to D converter circuit to produce more resolution than that necessary for desired output, to produced an uncorrected, extra resolution output;
- using said information determined by said determining to determine a correlation between said uncorrected extra resolution output, and a correlated final output;
- programming a look up table using said correlation;
- and
- using said uncorrected extra resolution output to obtain corrected information from the look up table.
An A to D conversion correction circuit, comprising:
- an A to D converter, producing an uncorrected digital output, wherein said A to D converters produce more resolution than is necessary for a desired output, and said look up table output has less resolution than a resolution of said output of said A to D converters;
- a look up table, storing a correlation between said uncorrected digital output, and a corrected digital output, said corrected digital output including at least correction from a differential nonlinearity in the A to D converter, said look up table arranged such that said correction values are stored at addresses of said uncorrected values such that said uncorrected values form the address for reading the corrected values out of the look up table.
An A to D converter system, comprising:
- a first A to D converter having first characteristics;
- a second A to D converter, also having said first characteristics;
- and
- a look up table, having stored therein information to compensate for said first characteristics, said information compensating at least for a differential nonlinearity error in said first and second A to D converters, and said look up table being addressed using uncorrected digital information from each of said first and second A to D converters to produce corrected digital information from each of said first and second A to D converters wherein said first and second A to D converters and said look up table are all formed on a common substrate wherein each of said first and second A to D converters have more resolution than is produced by an output of said look up table.
A method of analog to digital converting an output, comprising:
- using an analog to digital converter device, which has a do controlled differential nonlinearity, to convert an analog value to a digital uncorrected value;
- testing characteristics of the digital uncorrected value,
- determining codes in said digital output value which will never occur;
- writing only those codes which will occur into a look up table, and not writing the codes which will not occur into said look up table;
- and
- using said look up table to correct for the differential nonlinearity.
A method of correcting for an inaccuracy in a circuit that converts between analog and digital, comprising:
- determining information about differential non-linearities in an uncorrected circuit of a type that converts between a first signal and a second signal, where one of said first and second signals is analog, and the other of said first and second signals is digital;
- using said uncorrected converter circuit to produce more resolution than that necessary for a desired output, to produce an uncorrected, extra resolution output;
- using said information determined by said determining to determine a correlation between said uncorrected extra resolution output, and a corrected final output;
- programming a look up table using said correlation;
- and
- using said uncorrected extra resolution output to obtain corrected information from the look up table.
A corrected A to D converter, comprising:
- an A to D converter unit, receiving an analog signal and do producing an uncorrected digital output signal that is indicative of the analog signal, said A to D converter including capacitors to effect the A to D conversion, sizes of said capacitors being constrained to produce an intentional controlled differential nonlinearity, where at least a plurality of said capacitors are less than double the size of the next-smallest capacitor;
- and
- a memory, storing a relationship between said uncorrected digital output and a corrected digital output as corrected to compensate for differential non linearity errors, and wherein wherein said uncorrected digital output has more bits than said corrected digital output, said memory connected to receive information indicating said uncorrected digital output signal as a physical address which addresses the corrected digital output, and to output said corrected digital output which corresponds to said uncorrected digital output, said memory not storing all values of uncorrected digital output, specific values of said uncorrected digital output which will never occur not being stored, to conserve space in said look up table.
Dependencies are inferred from the claim text.