parameter detection test fails

Mika Bay

Hi,
Can someone help me with this.
I have designed a high voltage board (230VAC) around f2805m instaspin board.
The voltage feedback resistances are 600K + 5.1K||100nf giving
#define USER_ADC_FULL_SCALE_VOLTAGE_V (391.53) &
#define USER_VOLTAGE_FILTER_POLE_Hz (314.72)
The currents are amplified by -ve polarity opamp with Rf = 24K and Ri = 1.6K giving
#define USER_ADC_FULL_SCALE_CURRENT_A (22) & pwm frequency is
#define USER_PWM_FREQ_kHz (15.0)
Currently I am testing two motors with low DC voltage (48V).
One motor which works perfectly well (is wound for high voltage) has follwoing parameters
Rs = 2 ohm, L = 6.8 mH and back emf 1000rmp/30Vrms (V/F = 0.2068). The motor parameter are detected quite close & motor runs very smooth and perfectly.
The other motor which is wound for low (48V) operation, has following parameters
Rs = 0.21ohms, L = 0.5 mH and backemf = 2.8V/1000rpm(V/F = 0.032)
. The motor while parmeter detection, goes through following stages and finally stops in CTRL_State Error &EST_STate_Error. Following states are seen
RoverL, Rs, Ramp_Up, motor smoothly rotates upto 300rpm,-> finally EST_STate_Error. Values detected are also wrong, Rs= 0.27, V/F = 0.042, inductance yet to be determined.
In view of this can anybody suggest what might be going wrong? Is resistance divider too high for low voltage operation?
Also, how to find reason for error -> Are there any parameter(s) for latching error for user to detect reason for this behavior,,,

over 9 years ago

0 Yanming Luo over 9 years ago

TI__Guru** 117140 points

Hi Mika,

Please try to increase or decrease the maximum currebt used for Ls estimation. If failed again, you can change below 3 parameters to enable the motor run smoothly.

#define USER_MOTOR_RES_EST_CURRENT      (4.0)
#define USER_MOTOR_IND_EST_CURRENT      (-4.0)
#define USER_MOTOR_FLUX_EST_FREQ_Hz     (40.0)

Best Regards,

Yanming Luo

0 ChrisClearman over 9 years ago

TI__Guru** 119680 points

Mika,

Drag and Drop the user.h you are using for the low voltage motor.

it may just be a setting

but it very well could be the smaller Bemf voltage on such a large voltage sensing circuit. running at higher estimation frequency (to a point) should help a bit with this, but I'm concerned that the Rs value is not ID'd correctly at the beginning.

after RoverL what are the values of ctrl.Rhf and ctrl.Lhf ?

0 Mika Bay over 9 years ago in reply to ChrisClearman

Intellectual 600 points

Chris,

I checked the low voltage motor(which failed the test) ctrl.Rhf and ctrl.lhf both were zero. I also checked ctrl.Rhf and ctrl.LHf of high voltage motor, they are 2.09Ohm and 0.0071H -> very close to actual value. Also find here attached user.h file, it is same for both motors3463.user.h

0 ChrisClearman over 9 years ago in reply to Mika Bay

TI__Guru** 119680 points

Check again. It can't be zero and still return values for Rs.

For the low voltage motor try changing
#define USER_IQ_FULL_SCALE_VOLTAGE_V (48.0)

if this doesn't solve it I would chalk this up to the ADC_FULL_SCALE_VOLTAGE_V being too large for this small Bemf motor.

0 ChrisClearman over 9 years ago in reply to ChrisClearman

TI__Guru** 119680 points

also run it at higher speed to get more voltage to the system...but if it's not even registering an Rs it won't matter
#define USER_MOTOR_FLUX_EST_FREQ_Hz (60.0)

0 Mika Bay over 9 years ago in reply to ChrisClearman

Intellectual 600 points

Ok, I will check again. Meanwhile if EST_STATE goes to error is there any variable which indicates why test failed?

0 ChrisClearman over 9 years ago in reply to Mika Bay

TI__Guru** 119680 points

no, with that error there are no further details given

C2000™︎ microcontrollers

C2000 microcontrollers forum

parameter detection test fails