• State
  • Replies 7 replies
  • Subscribers 31 subscribers
  • Views 62 views
  • Users 0 members are here
Support feedback
Options
Options
Related

MCF8329A: Optimization of the parameters to reduce the current

rongwei mao
Prodigy 50 points
Part Number: MCF8329A

Tool/software:

I am testing the motor drive evm board. With the help of Dhinesh, I am able to make it run:  RE: MCF8329A: Not able to complete MPET testing, not be able to start the motor using knob 

The motor runs pretty well, but there are couple issues.

1) the current is very higher than expected. I have another motor driver testing the same motor. The current is at least 20% lower.

 

the Id is high. Sometime it is positive, sometime negative. Maybe the angle is not correctly set(according to chatgpt), but didn't find a way to adjust.

2) the nmos is very hot. I only use up to 4A to drive the motor. On another motor driver, the nmos is very cool at 4A current.

Certainly we can't simply compare this because nmos is different, but the big nmos shouldn't be this hot at 4A. If leave there for 1min, you finger can't touch it.

There must be a big room to improve.

This is what i did so far.

I measured the motor inductance and resistor with LCR meter, so the value shall be close enough.

Then i use the motor studio to run Ke measurement at 300Hz (it failed at 1000Hz, but it works at 300Hz as the max).

Then i switch to I2C speed control, I can run the motor at 11000rpm. The current from power supply is about 3.43A, The bus current shown in the figure is only 1.6A. Don't understand the exact meaning.

Id is very high.

I tune the BEMF  value from 8-10 (measured by MPET is 9), there is no change of the total current, so BEMF value is not the issue.

Please help to instruct how to optimize

 register.2csv.csv

  • 0
    TI__Expert 3455 points

    Hi,

    I see deadtime configured in shared tuning file is very low (50ns), Please change it 350ns.

    We need to use motor phase resistance and inductance values for CLOSED_LOOP2->MOTOR_RES and CLOSED_LOOP2->MOTOR_IND register settings, not the one measured across two phases. Please confirm.

    Thanks,

    Dhinesh

  • 0 in reply to Dhinesh
    Prodigy 50 points

    Quick reply. For the inductance and resistance, i use the measured value /2 as the value needed.

    I will test the deadtime tuning and let you know

    thanks

  • 0 in reply to rongwei mao
    Prodigy 50 points

    I tuned some time of deadtime

    200ns is the best. 150ns setting has slightly larger current.

    250ns triggers starting error, which may be fixed by other setting.

    At 200ns setting, the current is much much lower than before. dropping from 3.65A-> 2.23A, this is significant improvement.

    After turn on deadtime compensation enable , it can run at 250ns-300ns, current drops very slightly.  like from 2.23A to 2.20A (1%).

    My current issue is the Id fluctuates a lots. Suggestion on that? 

    After enable the compensation, it is easy to start. 250ns slightly has lower current (very minor)

    register3-250ns.csv

  • 0 in reply to rongwei mao
    Prodigy 50 points

    Dhinesh

    The deadtime of 250ns with compensation is good.

    The problem now is I can only push up to 850Hz. The max limit I set to 1200Hz.

    The reference torque pi  40% is OK, but when i increase to 50%, it makes worse.

    How to tune these parameters to push up to 1000Hz and reduce the fluctuation of Id current?

    Thanks

    register4-250ns-Ilimit50p.csv

  • 0 in reply to rongwei mao
    TI__Expert 3455 points

    Hi,

    can you try using PWM frequency (CLOSED_LOOP1->PWM_FREQ_OUT) as 20Khz?

    Also, please check whether you are able to reach maximum speed by running the motor in current control (PERI_CONFIG1->CTRL_MODE= current control).

    Thanks,

    Dhinesh

  • 0 in reply to Dhinesh
    Prodigy 50 points

    I tested with 30kHz, it looks worse. it is able to push to 800Hz, not above

    For current control, i am not be able to tune a good status, very difficult to start motor, and not stable

    it looks current loop has issue?

  • 0 in reply to rongwei mao
    TI__Mastermind 23942 points

    Hi 

    Dhinesh is OOO today, until he returns let me take this question.

    I am starting with context from recent post about current control loop not working.

    I see current KP, KI is programmed with value 0.2 and 10, how did we get this number?

    I recommend write 0 to these fields and allow device to calculate and use and then optimize it if needed.

    Also, closed loop acceleration is 500, check by reducing.

    Also, try to use PWM frequency 25kHz/20kHz.

    Thanks and Best regards

    Venkatadri S