Constantin
Well-known member
I have a general question re: adc references, some of which are externally accessible.
As best as I understand it, the reference is directly responsible for the accuracy of the adc. In the case of the mcp3910, the reference is specified with a +/- 2% spread re VREF and a 9ppm/C drift. Yet, the unit is specified to allow 1% or better accuracy re energy measurements.
How does that work? Is it that the reference may be off but it's error is consistent and hence can be calibrated? In that case, would an external reference merely buy better unit to unit measurement consistency and potentially reduce / eliminate the calibration burden? For example, the adr3412 series offers 0.1% accuracy and 8ppm/C drift.
Anyhow, if anyone has any good articles, sites, or books to point me to, I'd appreciate the insights.
As best as I understand it, the reference is directly responsible for the accuracy of the adc. In the case of the mcp3910, the reference is specified with a +/- 2% spread re VREF and a 9ppm/C drift. Yet, the unit is specified to allow 1% or better accuracy re energy measurements.
How does that work? Is it that the reference may be off but it's error is consistent and hence can be calibrated? In that case, would an external reference merely buy better unit to unit measurement consistency and potentially reduce / eliminate the calibration burden? For example, the adr3412 series offers 0.1% accuracy and 8ppm/C drift.
Anyhow, if anyone has any good articles, sites, or books to point me to, I'd appreciate the insights.