• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 2 answers
  • Subscribers 33 subscribers
  • Views 324 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.

AM2434: Motor Control Processing Time in the app note AM243x/AM64x Single Chip Servo Motor Control Implementation and Benchmark

Hideaki Matsumoto
Genius 9090 points
Part Number: AM2434

Tool/software:

Hi,

My customer has some questions in the app note of AM243x/AM64x Single Chip Servo Motor Control Implementation and Benchmark (sprad03 – March 2022).

Could you answer their questions below ?

 

1.  Processing time

    Is the CPU processing time for control period the total of Open Loop Id/Iq, Closed Loop Id/Iq, Closed Loop Speed, Closed Loop Position and Closed Loop CiA402 ?

 

 

 2.  Difference between Table 10-2 and Table 14-5

    There are Table 10-2 and 14-5. Do the numbers in Table 10-2 come from the Max normalized in Table 14-5 ?

    If the customer wants to know the maximum value of CPU processing for control period, they should summate all of Max values in Table 14-5 ?

    It would be 4568 [ns] ?

 

3.  Difference between Min and Max

    In the Table 14-5, there are Min an Max value. Could you tell them what the factor is for these differences ?

 

4. Update

    This app note looks old, issued in 2022. Do you have any update or improvement in this benchmark ?

Thanks and regards,
Hideaki

  • 0
    TI__Guru* 83865 points

    Hello Matsumoto-san

    I have requested Ming to provide some guidance on this content. He is no longer the owner of this doc, but may have the context. 

    Will discuss internally if this appnote needs to be refreshed. Currently no such plans, but if there are portions of the appnote that are of interest to your customer and are needed to to make the right design decisions, feel free to let us know. 

    Regards

    Mukul 

  • 0 in reply to Mukul Bhatnagar
    TI__Guru 55055 points

    Hi Matsumoto-san,

    Your e2e is about the appnote:  AM243x/AM64x Single Chip Motor Control Benchmark (ti.com). It is the appnote about a older single chip single motor demo based on the AM64x EVM and 4 other adaptor cards. Unfortunately, we do not have the hardware setup anymore and the code never published on ti.com.

    There is new TI design  EtherCAT® Connected, Single-Chip, Dual-Servo Motor Drive Reference Design (ti.com) which uses the AM243x LP plus a BLDC Dual Motor launch pad which is available in ti.com and the source code is included in the Motor Control SDK since 09.00.00. I have a collected some average performance numbers for that setup (Closed speed loop only):

    focCountTotal is the total time for Closed Speed FOC loop.

    Best regards,

    Ming

  • 0 in reply to Ming Wei
    TI__Genius 9090 points

    Hi Ming, Mukul,

    Thank you for your response. The reason why we reffered to the older single chip single motor demo is that we need the benchmark data. The newer design app note doesn't it.

    Thank you for sharing the collected performance numbers.

        -  Could you tell us what Raw and Pure are ?

        -  Do you have Closed Position Closed Loop and Closed Loop CiA402 ?

        -  Actually, they want to know the processing time of EtherCAT CiA402 profile, Could you share it if you have ?

    We understood the app note which we referred to was older design, but could you please answer to their questions 1 to 3 described above ?

    Thanks and regards,
    Hideaki

  • 0 in reply to Hideaki Matsumoto
    TI__Guru 55055 points

    Hi Matsumoto-san,

    1.  Processing time

        Is the CPU processing time for control period the total of Open Loop Id/Iq, Closed Loop Id/Iq, Closed Loop Speed, Closed Loop Position and Closed Loop CiA402 ?

    [MW] Yes. It is the total time from getting current/position feedback to the PWM output for each FOC type. 

    2.  Difference between Table 10-2 and Table 14-5

        There are Table 10-2 and 14-5. Do the numbers in Table 10-2 come from the Max normalized in Table 14-5 ?

        If the customer wants to know the maximum value of CPU processing for control period, they should summate all of Max values in Table 14-5 ?

        It would be 4568 [ns] ?

    [MW] No. See my answer above. Each time, you only use one FOC type.

    3.  Difference between Min and Max

        In the Table 14-5, there are Min and Max value. Could you tell them what the factor is for these differences?

    [MW] the Max and Min difference is caused by the program and data caching. Unless all the program and data are in TCM, there will be variation between each FOC loop. 

    We have not implemented the Closed Position Loop in the new TI design, so we do not have performance numbers, but based on the performance numbers provided in the appnote. The numbers should be similar to the Closed Speed Loop, so is the Closed Loop CiA402. In fact, our new TI design, supports the CIA402 Speed Control.

    Best regards,

    Ming