• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 149 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

ADS8588S: Offset error + Gain error

Felix Debus
Intellectual 980 points
Part Number: ADS8588S

Tool/software:

Hi team,

I have a question regarding the datasheet of the ADS8588S.

The offset error temperature can be calculated for 60°C with 3ppm/°C * 60°C * 10000mV = 1.8mV, is that correct?

3ppm/°C max Offset error temperature drift

10000mV = +-5V full scale range

Now to my questions:

1.   1.8mV are the max offset error for +-5V and +-10V. Does it mean that the max offset can be 1.8V over the full temperature and voltage range? So the 3ppm/°C are just valid as long as an offset error is below                        1.8mV? Can you please explain shortly why this is this also the max value for the +-10V full scale input range? From my feeling it needs to be doubled.

2.   Can I convert the 3ppm/°C in a percentage value?

3.   The max ADC offset error for 25°C is then: 3ppm/°C * 25°C * 10000mV = 750uV? Is this correct?

4.   How can I calculate the max Gain error? Do we have a max value for the Gain error at 25°C with internal reference? Is it always +-10ppm/°C? 

5.   Do we have curves for the offset drift and the Gain drift over the full temeprature range?

Best regards

Felix

  • 0
    TI__Guru**** 181001 points

    Hi Felix,

    1. Yes, the offset is +/-1.8mV over the full temperature range and all voltage input ranges.  Short answer reason is because the only change from +/-5 to +/-10 is the internal sample/hold cap.
    2. Sure - use the Analog Engineer's Calulator for that if you like:  
    3. Not really - usually things are normalized at room temp (25C) so you start off with 0 offset (calibrate by shorting the ADC inputs then add/subtract any offset error).  At 60C then, you would look at the delta in temperaure 60-25 = 35 and the offset drift would then be 35 * 3ppm.
    4. Gain error is +/-4LSB typical - that would be what you see at room temp, this is also calibrated out initially.  The PPM/C number is only applied when there is a change in temp as described in point 3.
    5. Review the datasheet - Figures 21 through 31
  • 0 in reply to Tom Hendrick
    TI__Intellectual 980 points

    Hi Tom,

    thank you for the answer!

    May I asked again one point from above?
    Are the offset error and the Offset error temperature drift two different errors? Do I need to get the total offset error added both together?

    I am asking because I still do not fully understand the calculation. If we have 85°C  and more (60°C above calibrated temperature) the Offset error temperature drift would be higher than the max offset error. Can you please help me to understand this?

    Best regards

    Felix

  • 0 in reply to Felix Debus
    TI__Guru**** 181001 points

    Hi Felix,

    Yes - offset would be the initial error, and offset drift is the devication of the offset error over temperature.  For worst case error, you could add them together.  For better understanding, please take a look at the Error Sources (3) videos 'Statistics behind error analysis...' and 'Understanding and calibrating the offset and gain...' at this location:

    https://www.ti.com/video/series/precision-labs/ti-precision-labs-analog-to-digital-converters-adcs.html

  • 0 in reply to Tom Hendrick
    TI__Intellectual 980 points

    Hi Tom,

    got it! Thank you! The video series is a good source of information.

    They mentioned the calibration methods in there. You also mentioned this before for the gain error. In the datashhet I couldn't find a reference to that.

    How is the calibration done for the internal gain error? Is it because the internal reference source is temperature compensated or does the device perform an automated calibration? Or do we need to perform an external calibration?

    Best regards

    Felix

  • 0 in reply to Felix Debus
    TI__Guru**** 181001 points

    Felix,

    Gain error is a deviation in the slope of the transfer function.  Generally speaking it would be verified at FS and then you could calibrate through SW to accomodate any non-ideal slope.  The ADS8588S does not have any internal calibration capabilities.