• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Subscribers 82 subscribers
  • Views 1719 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.

ACIM Identification with Motorware 12

Matthias B.
Intellectual 330 points
Other Parts Discussed in Thread: MOTORWARE, TMDSHVMTRINSPIN

Hi all,

I’m using Motorware 12 now. I try different Voltages for the DCbus (100V, 300V, 600V).
At State_Ramp_up the Motor just moves back and forward. This problem also occurs during State_IdRated no matter at which DCbus level and no matter how big RES_EST_CURRENT.


#elif (USER_MOTOR == Z_4KW5)
#define USER_MOTOR_TYPE                 MOTOR_Type_Induction
#define USER_MOTOR_NUM_POLE_PAIRS       (1)                
#defineUSER_MOTOR_Rr                    (NULL)             
#define USER_MOTOR_Rs                   (NULL)                
#defineUSER_MOTOR_Ls_d                  (NULL)                
#defineUSER_MOTOR_Ls_q                  (NULL)             
#define USER_MOTOR_RATED_FLUX           (11.26)                 
#define USER_MOTOR_MAGNETIZING_CURRENT  (NULL)                
#define USER_MOTOR_RES_EST_CURRENT      (1.0)                 
#define USER_MOTOR_IND_EST_CURRENT      (NULL)             
#define USER_MOTOR_MAX_CURRENT          (6.5)           
#defineUSER_MOTOR_FLUX_EST_FREQ_Hz      (5.0)  


I get good results for a PM with my hardware but I recognized that I get better results for PM with PWM frequency of just 4KHz. Is a PWM frequency of 4KHz a problem for ACIM ID? I also tried today 10kHz and 15kHz but get always the same problem.


Here is the nameplate of the motor:
50 Hz
5,5kW
cos 0,89
690V (Y)
6,1A
2925/min
USER_MOTOR_RATED_FLUX = 690V * sqrt(2)/sqrt(3) * 1/50Hz = 11.26
I connect the motor in Y.
Attached you can find oscilloscope plots of the three phase currents.As mentioned the motor just moves forward and back.
It’s important for me to understand why it’s not working. I really had high hopes that ID for ACIM is working with Motorware 12.

Thanks for your support

Matthias

2821.3 Phase Currents ACIM ID Motorware12.pdf

  • 0
    TI__Guru** 119670 points

    4 KHz isn't a problem for running, although I would probably start out with 10 KHz so the Estimator is running faster during Motor ID.

    Did you use the spreadsheet to set-up rest of user.h?

    Did you check gMotorVars.UserErrorCode?

    How high did you go on USER_MOTOR_RES_EST_CURRENT?

  • 0 in reply to ChrisClearman
    Intellectual 330 points

    What do you mean with a spreadsheet?

    I didn’t check the ErrorCode because often ID ends with state idle.

    Biggest USER_MOTOR_RES_EST_CURRENT was 3.0.

    Are there new documents for Motorware 12? I just could find the lab documents.

    Thanks for your fast answer

  • 0 in reply to Matthias B.
    TI__Guru** 119670 points

    Setting User.h

     gMotorVars.UserErrorCode can be viewed once the gMotorVars.Flag_enableSys is set.

    This checks variable limits post-compile and pre-runtime.

    MW _12 has updates to the instaspinlabs.pdf and any new comments that get pulled through Doxygen in MotorWare.exe  

    The UG and TRM docs are being updated as well, but they aren't on the same schedule as MotorWare (take longer for updates to publish).

    you can post your user.h and I'll take a look....I don't think this is an issue though with this motor.

     

    I've sent this to the person who made al the ACI updates....he'll be surprised that this isn't working for you.

  • 0 in reply to ChrisClearman
    TI__Guru** 119670 points

    I"m pretty sure this is a timing issue.

    Please post your user.h or at least try with

    10 KHz PWM

    set all decimations to (1) except for speed and trajectory at (10)

     

  • 0 in reply to ChrisClearman
    TI__Guru** 119670 points

    Matthias,

    Any success yet?  If not, please Options --> Attach your user.h

  • 0 in reply to ChrisClearman
    Intellectual 330 points

    0844.user.h

    Hi Chris, here is my user.h file. I also tried the PM ID today and I get the same problems as with ACIM ID, Motor just moves forward and back. I have to check my hardware once again.

    But have a look at my user.h file.

    Is the Voltage_Filter_POLE okay? If I understand the spreadsheet right my ideal Hardware Pole is 22,1Hz. Is 447Hz also okay?

    4657.motorware_selecting_user_variables_gamma.xlsx

  • 0 in reply to Matthias B.
    TI__Guru** 119670 points

    Matthias,

    I have made changes and comments to your user.h attached.

    Search for "CCC" for each of my comments.

     

    May I suggest that you first try to verifty operation on a TI Hardware Platform before attempting on your own hardware?  You are bringing in many more chances for error with unproven hardware.  Even if you do have a 600V motor, you should be able to ID and run unnloaded with a 300V bus on TMDSHVMTRINSPIN kit just fine.

    Especially when you say you also have problems with your PM motor now.  I feel very strongly that there is an error in your HW or in how you are updating the variables in user.h to match your hardware.

     

    "Is the Voltage_Filter_POLE okay?"

    USER_VOLTAGE_FILTER_POLE_Hz is the software equivalent of the phase voltage filters in hardware. It needs to match the pole per the calcualtions you can use from SPRUHJ1 5.2.4

    When you design your HW you should follow this rule:

    If Max_Hz < 800.0, design voltage fitler poles to 200.0 Hz
    Else, design filter poles =  Max_Hz + buffer  /  4.0

     

    user.h
  • 0 in reply to ChrisClearman
    Intellectual 330 points

    Chris,
    as you recommended I tried the HVKit today to identify the ACIM.
    It works perfectly!
    I also get the new version of Motorware to work with my hardware. Obviously I missed something in the hal.c file.
    The parameter identified from HVKit looks very good.
     

    I also used your changes in the user.h file. I changed the Voltage Hardware Filter Pole to 196Hz on my hardware but this did not have an effect to the accuracy of the ID. Here is a comparison for an identified PM.


    I think the ID of ACIM needs more accuracy of the measurement as the ID of PM, right? So I have to improve the measurement.  Do you have any other suggestions?


    Thank you for your support

  • 0 in reply to Matthias B.
    Intellectual 330 points

    Unfortunately the graphics are note included in my last post. You can find them attached.

    6406.ID ACIM and PM.pdf

  • 0 in reply to Matthias B.
    TI__Guru** 119670 points

    Matthias,

    In the pdf you just attached, are ID Inverter and ID My Hardware the same inverter?

    For the PMSM the Rs is the only thing very different. Are you using different values for RES_CURRENT?  I would guess for My Hardware you are using too much current and heating the wires.  The values vs. the datasheet look good for the TI HVKIT. Note that our values are Line-Neutral, while most datasheets are Line-Line, so they will be ~2x ours.  Also note that datasheets are never accurate, our ID process is better.  For the flux, they both look ok, and a difference like this can be found in the error in the phase voltage filters. One or both may be slightly off. I have posted a way to rudimentary tune this SW filter_pole to best match the specific HW filter pole....but they can still drift over temp/speed/time.

    for the ACI motor...there is a gross error.  Either in your HW circuitry, set-up of the driver (perhaps the current bias direction), or a mis-match in the software settings. 

    But if this is the same inverter as My Hardware this shouldn't be the case...it appears My Hardware is working (although if your PMSM Rs value is still showing high I would verify the amount of current actually being injected during Rs state.)

    Do you have your ADC_CURRENT values set correctly? 

    As an example, If you are measuring +/- 10.75A then ADC_CURRENT (21.5) and IQ_CURRENT (11.0)

     

  • 0 in reply to ChrisClearman
    Intellectual 330 points

    Chris,

    Yes My Hardware and Inverter are the same in the pdf. Thank you for the ADC_CURRENT example. I made some changes on the current measurement some time ago and didn’t verify it till now. But I think this is the point!

    I measure +/-40A therefore I have to set

    ADC_CURRENT (80.0)

    IQ_CURRENT(41.0)

    Tank you very much

  • 0 in reply to Matthias B.
    TI__Guru** 119670 points

    ok, so you had both your ADC and IQ _CURRENT at half their value? You were certainly putting in 2x and measuring 2x the curren then...

    what about your IQ and ADC _VOLTAGE?  They are very high for 350Vdc bus motors.  Why do you need nearly 900Vdc measurements?  This hurts the resolution of all of your signals if you don't really need to ever measure voltages this high.