Motor Identification with KMS GUI

[derekcook]

Hello, 

I am having quite a bit of trouble trying to complete the automatic motor identification portion of the KMS GUI with my own motor on my own hardware. 

During the resistance measurement portion of the Self Commissioning Mode used during the automatic motor identification I am pulling ~600A of current. I have the RS and LS identification current both set to 10%. My motor rated current is 37.5A, and my rated system current is 400A. These are all set in the KMS GUI. Therefore, I'm confused as to why I am pulling 600A of current when attempting automatic motor identification. 

It would be good to have a call about this if possible to talk about the system, what I've done, and what to do to move forward. 

Thanks for your help, 

Derek

[derekcook]

Hey Adam, 

Thanks for the help on the call. We resolved our problem and now are able to run all of the self identification, resistance, induction, flux, and inertia from the GUI. We are also able to control the motor in speed and position mode from the GUI using our own hardware. The issue was that we reversed the polarity on the current feedback. We are going to fix this in the hardware, but for the time being we inverted the ADC values for current feedback by taking 4095-RAW (So That 4095 is full negative and 0 is full positive). 

Once we corrected the current feedback, we could then also run the DC Injection validation to ensure we have 5A applied to phase A, and -2.5A applied to both B and C. 

Thanks for the help, 

Derek

[linestream-adam]

Derek,

Great news!  I'm glad everything is working well.

[derekcook]

Hey Adam, 

It would be good to have a call about this if possible. Would it be possible to get on the phone / Skype with yourself or some others at NXP (today if possible) and some of our engineers here at General Electric (hardware and software) so that we can walk through what we are doing, and you all can maybe point out things we are doing incorrectly. 

We are unable to connect our motor to the reference hardware. Therefore, we are trying to identify it using our hardware. We have put our system parameters into the GUI, as well as verified our ADC scaling. Our PWM frequency is 4kHz. 

[linestream-adam]

Drop me an email at: adam@linestream.com

We can setup a call for today.

[linestream-adam]

Derek,

I'm a little confused as to why your custom board is setup to sense 400A of current. What is the maximum current you expect to use with your motor?  The reason I'm concerned about this is the finite ADC resolution.  If you only ever expect to run with 37.5A, but you can sense 400A, you are only using 8 of the 11 bits (one is lost to sign) of resolution on the ADC. This will impact how well you can control the motor.

But the motor parameter identification can be treated as a separate issue. When the motor parameter identification runs, it needs to control the current in the motor, but without knowing the motor parameters it needs to be de-tuned to the point where it will still work. So it assumes a very slow tuning, and attempts to control the current in the motor. Under some situations this may not work.

Also if you are switching 400A, what is the PWM frequency you are using.  Most IPMs that can handle that much current can't switch very quickly, and if you motor is too small the inductors act like wire instead of inductors, making control impossible.

Where you able to estimate the motor parameters and spin the motor using the reference hardware? If so, you shouldn't need to perform the motor parameter identification on your custom hardware.

[derekcook]

Hey Adam, 

It would be good to have a call about this if possible. Would it be possible to get on the phone / Skype with yourself or some others at NXP (today if possible) and some of our engineers here at General Electric (hardware and software) so that we can walk through what we are doing, and you all can maybe point out things we are doing incorrectly. 

We are unable to connect our motor to the reference hardware. Therefore, we are trying to identify it using our hardware. We have put our system parameters into the GUI, as well as verified our ADC scaling. Our PWM frequency is 4kHz.