• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 82 subscribers
  • Views 902 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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 motor identification reports low inductance

DC
Prodigy 175 points
Other Parts Discussed in Thread: MOTORWARE

Using Motorware 15 with the HVKit.

 

Attempting to run motor identification with a couple different class D AC induction motors and getting Ls_d / Ls_q values that don’t seem right.  Wanting to discuss within the context of comparing to a manual test for collecting motor parameters...

 

For example using manually executed no-load/locked rotor/DC resistance tests by hand we came up with:

Rs=2.626

Rr`=4.854

Ls=0.008108

Lm=0.07840

 

Using HVKit, we receive the following values:

Rs=2.6069

Rr=5.470872

Lsd_H=0.002194

Lsq_H=0.002194

Flux_VpHz=2.950347

MagnCurr_A=5.184385

 

1. How to evaluate this comparison?

a. Rs compare looks good.

b. Rr compare looks close??

c. Ls feels like something is wrong? (This is a direct compare, right? )

d.Regarding Lm…To compare back to manual test Lm value, can we calculate it to be stator flux in webers divided by rated mag current? (0.8165*230VAC/(2*pi*60)) / 5.184385 = 0.0961H ? vs. 0.0784 from manual test. What is the right way to do this?

e. Note However, I believe the manual locked rotor test was run at 60Hz and not 15Hz (25% of rated freq), so I’m not sure how this might have impacted the manual test results.

  

 

I then tried a different AC induction motor which I have not run the manual motor parameter test with yet.  However, the motor identification process with HVKit produces odd values for Ls….

 

Rs=7.522885

Rr=15.33149

Lsd_H= 2.865012e-07

Lsq_H= 2.865012e-07

Flux_VpHz=3.133936

MagnCurr_A=2.644088

 

2. How to evaluate the results?

a. Rs is reasonable when quick check against ohm meter (haven’t run dc resistance test yet)

b. Lsd and Lsq values do not appear to have a “reasonable” value, right?

c. If motor id won't produce a good value, what is the recommended method for determining these values?

 

 

 

3. What is the expected performance impact of:

a. Having an Lsd_H / Lsq_H inductance value that seems completely wrong?

b. Having an Lsd_H / Lsq_H inductance value that is off by only a couple-hundred percent?

 

 

Note: For reference here is what my HVDevkit motor id reports for motor params for the Marathon 230VAC HVdevkit AC motor VH48T17D2000K

#define USER_MOTOR_TYPE MOTOR_Type_Induction

#define USER_MOTOR_NUM_POLE_PAIRS (2)

#define USER_MOTOR_Rr (6.527813)

#define USER_MOTOR_Rs (9.722257)

#define USER_MOTOR_Ls_d (0.00892178)

#define USER_MOTOR_Ls_q USER_MOTOR_Ls_d

#define USER_MOTOR_RATED_FLUX (2.996896)

#define USER_MOTOR_MAGNETIZING_CURRENT (1.538325)

#define USER_MOTOR_RES_EST_CURRENT (1.0)

#define USER_MOTOR_IND_EST_CURRENT (NULL)

#define USER_MOTOR_MAX_CURRENT (6.0)

#define USER_MOTOR_FLUX_EST_FREQ_Hz (5.0)

 

Chris Clearman had similar results posted in this thread…

https://e2e.ti.com/support/microcontrollers/c2000/f/902/t/509955

 

If this motor is supposed to be equivalent to the Marathon_5k33GN2A motor, then something again seems off regarding the motor id generated Ls_d value.  The 5K33GN2A motor ID params in motorware report Ls_d ~= 0.04282

 

4. By chance has TI determined what the “expected” values of Rs,Rr,Ls_d,Ls_q,Flux,MagCurr are for this motor? (in other words…Perform a manual motor parameter test vs. motor ID test results)  (If not, can you please consider running a manual test and posting results?)

 

 

5. Note: It would make an excellent app note / blog topic to discuss manual motor parameter collection test details and mapping to motorware motor model parameter details.

 

 Thx

  • 0
    TI__Guru** 118760 points

    1.

    a. yes

    b. close, but a small Rr error will result in a small error in the speed estimate only

    c. for Ls did you use an LCR meter and do line-line? Did you divide by 2?  It would still be off by a factor of 2 though...

    d. I'm not sure on this one

    e. Using the LCR @ 60 Hz should be fine

    2.

    a. Rs is usually correct, it's a very simple test

    b. correct, these are unreasonable. It would also make Rr unlikely to be correct

    c. you either need more nameplate data and follow SPRUHJ1 Chapter 4.8 or work on improving motor ID.

    3.

    a. inaccuracy of the FAST observer. Instability of the current control loops.

    b. inaccuracy of the FAST observer especially at higher frequencies. Instability of the current control loops.

    4. We have not. Will try to do so on Friday.

    While the manufacturer sells the 48T as an equivalent replacement, that seems to be strictly from the 1/4 Hp and current ratings. It clearly has some differences in motor parameters.

    5. agreed