This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1120: TUE and DNL

Part Number: ADS1120

Hi,

Could you let me know DNL and TUE of ADS1120 for 5V supply?

TUE=sqrt(sq(offset error)+sq(gain error)+sq(DNL)+sq(INL))

offset error=4uV=0.08LSB, gain error=0.015%=10LSB, INL=8ppm=0.5LSB

TUE=sqrt(0.08^2+10^2+0.5^2+DNL^2)

Is my calculation correct?

If correct, could you let me know DNL?

Thanks.

  • Hi David,

    The total unadjusted error (TUE) is an RSS value of the errors you have listed. There is typical TUE as you have used, or you can use worst case from the min/max specification. Offset and gain errors can be calibrated out of the measurement. Also, gain error is usually the most dominant error, so if required you could use an external reference with better specs. As far as the DNL error, this is usually not a consideration with Delta-Sigma ADCs as they are an oversampling device that is monotonic by nature (increasing DC input produces an increasing code) and unless indicated otherwise has no missing codes so all codes are represented.

    You are correct in that all errors need to be defined in the same units (you have chosen LSB, but could be in volts or ppm as well). I got a slightly different answer for the offset, but as you can see the dominant error will be gain error. The typical TUE will be a little more than 10 LSB, and the worst case TUE will also be dominated by the gain error at around 66 LSB.

    Gain error always has the greatest deviation from the ideal slope at full-scale. At 0V input the error is now dominated by the offset error. In the end TUE gives you a pessimistic view of the measurement, where as the actual sensor range may reduce the overall effect if the sensor range operates much less than full-scale.

    Best regards,
    Bb B