• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 83 subscribers
  • Views 639 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.

Rs-Online Accuracy

MKo
Intellectual 310 points

Hello,

I've got two questions regarding your Rs-Online identification:

(1) Do you have some numbers how accurate the identified Rs-values are under different load and inductance conditions? 5%, 10%, 15%? I am especially interested in high R (1-2Ohm) over L (10-100µH) motors.

(2) I guess you can't provide any further details (apart from what is written in spruhj1d) on how you perform the estimation (IEEE papers, patents, etc.)?

Thanks.

  • 0
    TI__Guru** 119500 points

    1. This is calculated by a direct measurement of voltage and current, so it's as accurate as the sensor signal chain. How accurate is an ohm-meter you are using?

    2. Correct, this is TI intellectual property that is already published as extensively as we will.

  • 0 in reply to ChrisClearman
    Intellectual 310 points

    ChrisClearman said:

    1. This is calculated by a direct measurement of voltage and current, so it's as accurate as the sensor signal chain. How accurate is an ohm-meter you are using?

    If you are using a flux observer you need a machine model which is never perfect and your Rs estimation might depend on things like inverter non-linearity effects, permanent flux linkage, inductances Ld/Lq and I_d current amplitude. Some of these can't be determined exactly or even vary. For this reason other Rs estimators I know of reach accuracy values of about 10%. Also motor temperature estimation may become useless if Rs uncertainty is too large.

  • 0 in reply to MKo
    TI__Expert 5180 points

    I woud say it is about 5% accurate, although we are taking masurements to have an accuracy listed in the documentation.

    Do you have a way to measure the accuracy you are getting?

  • 0 in reply to Jorge Zambada
    Intellectual 310 points

    Jorge Zambada said:

    I woud say it is about 5% accurate, although we are taking masurements to have an accuracy listed in the documentation.

    Thanks. This looks promising. Do you have an idea when these measurements will be published?

    Jorge Zambada said:

    Do you have a way to measure the accuracy you are getting?

    No, unfortunately not.

  • 0 in reply to MKo
    TI__Expert 5180 points

    We will have that information published in about 6 months.