• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 445 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.

Original question:

DAC8760: CRC-8-ATM Calculation algorithm

DAC8760: Gain / offset Calibration

Shunsuke Nagata
Genius 13230 points
Part Number: DAC8760


Hi,

Could you please give me your advice for DAC8760 calibration?

1. TUE (Total unadjusted error) : 
What kind of error parameter are include in TUE? e.g. Are DNL, INL ,Full-scale error, Gain error, Bipolar zero error, Zero scale error and offset error included in TUE?
If the all of error percentage are +/- 0.06%, my customer doesn't need calibration.

2. If user run gain and offset calibration in his product line, which error parameter are not covered? e.g. Are only INL and DNL not cover even if user run gain / offset calibration? or is there another parameter?

3. Can user refer the gain error and each offset errors in datasheet for Vout and Iout? or is there specific notice for Iout?

4. Gain calibration register and Zero calibration register :
Datasheet shows these calibration register are in 0x58 and 0x59 as write operation. Are there common register for  Vout calibration and Iout calibration? If yes, I guess user run Vout calibration only. Is it correct?

Regards,
Nagata.

  • 0
    TI__Guru 54773 points

    Hi Nagata-san,

    1. The TUE spec would include DNL, INL, Gain, and Offset error.  That gain and offset error would take the full-scale and bipolar zero-scale error.  It would not include the zero-scale error, as we specify that we measure the output from code 0x0100 to full-scale.  We do this because the zero-scale error cannot be calibrated out of the output.  

    2. INL/DNL cannot be compensated, nor can zero-scale error.

    3. The gain and offset calibration is simple digital adjustments, so it can be applied to both VOUT and IOUT values.

    4. There is only one set of calibration register for all ranges.  It is recommended that the MCU track the different calibration values for each range, then apply those values when the range changes.

    Thanks,

    Paul

  • 0 in reply to Paul_Frost
    TI__Genius 13230 points

    HI, Paul -san,

    Could you please give me your advice for below question?

    For 1.
    >It would not include the zero-scale error, as we specify that we measure the output from code 0x0100 to full-scale. 

    What kinds of parameter is under 0x0100 in TUE. Is it only zero-scale error? or are there other parameter?

    For 2.
    > INL/DNL cannot be compensated, nor can zero-scale error.

    Is there a method to cancel zero-scale error in customer calibration process? Could you please let me know your idea?

    Regards,
    Nagata.

  • 0 in reply to Shunsuke Nagata
    TI__Guru 54773 points

    Hi Nagata-san,

    1. Only the ZSE specification would capture this.  

    2. There is not much you can do to fully remove the ZSE.  Having greater "footroom" helps the error, but it is unavoidable without applying an external DC offset.

    Thanks,

    Paul

  • 0 in reply to Paul_Frost
    TI__Genius 13230 points

    Paul -san,

    Thank you for your support. 

    For 2.  Could you please follow up for ZSE? Customer would like to use Positive ( Plus ) range such as 0V ~ +10V or 10mA ~ 40mA. This mean is customer doesn't need to care Negative (Minus) range. I guess the ZSE can cancel with offset calibration because he can adjust input value to register to set 0V output. Is it correct?

    3. Datasheet value for TUE has -40C to 125C parameter and -40C to 85C parameter. If user refer these value, he doesn't need to care temperature dependency TUE which include DNL, INL, Gain and offset error, too?

    Regards,
    Nagata.

  • 0 in reply to Shunsuke Nagata
    TI__Guru 54773 points

    Hi Nagata-san,

    2. If they operate in the bipolar range, then they will be able to adjust the output to remove ZSE, but they will need to provide a negative supply for the device, and will loose 1 bit of precision, as have the range will not be used.

    3. That is correct, this value combines on the error sources. 

    Thanks,

    Paul