• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 333 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:

UCC21750-Q1: Measuring accuracy for sensing 1000Vdc

UCC21750-Q1: analog sensing accuracy

Dongbao Zhou
Expert 4170 points
Part Number: UCC21750-Q1

Hi team,

Just want to clarify what's the expected total error caused by UCC21750-Q1? Do we need to take both APWM duty cycle error and internal current source error into account.

1. APWM duty cycle error: +-2.5% across temp, +-1% with single point calibration. http://www.ti.com/lit/an/slua988/slua988.pdf?ts=1587982882008

2. Internal current source error: +-4.5% *200uA across temp, can we improve this by single point calibration?

Thanks

Dongbao

  • 0
    TI__Genius 10652 points

    Dongbao,

    Thats a good question.

    The min/max data for Internal current source error represent absolute worst case temp and process effects together, the range specified in the table doesn't mean the same thing as the error characterization plot such as for duty cycle error. This is likely why there is no characterization plot for just the current source.

    The APWM error in SLUA988, the report you reference, is measured by directly driving AIN, which negates the function of internal current source, so it doesn't contribute in that case.

    In the use case where the internal current source is used to generate VAIN, such as for the NTC thermistor and thermal diode, the datasheet does have some data for the duty cycle error, which would include any error contributed by AIN and might be what youre looking for.

    These are in Figures 58 & 59 of the datasheet, on  Page 43 and 44 respectively.

    I hope this helps. Please let me us know if you have any additional questions.

    Best

    Dimitri

  • 0 in reply to dimitri james
    TI__Expert 4170 points

    Hi Dimitri,

    That's to say, in the case of  temp sensing sensing, it's very promising to achieve +-1% total accuracy with AIN after calibration. Is my understanding correct?

    How about the DC link voltage sensing accuracy?

    Thanks.

    Dongbao

  • 0 in reply to Dongbao Zhou
    TI__Genius 10652 points

    Dongbao,

    +/-1% total accuracy for DC link sensing accuracy is definitely achievable with single-point calibration.

    The reason why NTC and thermal diode show some difference before single point calibration is simply due to the difference in DC voltage at VAIN produced by each. After the fixed point calibration, the error is very low for both. This concept is same for DC-link sensing, after a single point calibration you can expect similar accuracy of APWM duty cycle.

    Also, the "absolute error" in the app is really concerning the error impacted by the VAIN voltage

    In the datasheet the data for NTC/diode error comes from the difference between a "perfect" PWM signal and the corresponding with the response of NTC/thermal diode. Initially, the "average" error is high because of different VAIN levels for NTC and thermistor. This difference causes a consistent error in the duty cycle, which can easily be accounted for with single-point calibration. This is effective, and result is shown in teh next figure.

    Please let me know if you have any other questions.

    Best

    Dimitri