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.

I'm seeing variation in Ls measurements

This question has come up a few times outside of this forum so I thought I should address here as well.

First, if you are using identical settings and running the motor ID tests, you should get reasonably consistent results.  If not, you need to to look at your ID settings and read Chapter 6 of SPRUHJ1 as you are likely not getting a proper identification.

This topic addresses those who are changing their ID settings (current and frequency used) and seeing different Ls values identified. THIS IS NORMAL.

Motors tend to saturate when you use positive Id (like in our RoverL and Rs testing) and tend to de-saturate when using negative Id (like in our inductance testing) in no-load. In reality all motors with iron stator cores do show saturation, hence there is no single value for inductance, and there is a non-linear relation between flux and current.

 If you change the current or the frequency used during ID and you get different values, which is “correct”?  In principle they are both correct but you can also state that neither is correct in all situations.

How does an X% delta in Ls effect the outputs of FAST?

Suppose the short circuit current equals

        Isc = Flux/Ls (Wb/H).

Just for clarity suppose Id=0: then stator flux:

        Flux_d = magnet flux, Flux_q = Ls* Iq,

hence if Ls is actually a delta %, then the calculated rotor flux would be not equal to the magnet flux, but equal to

        Flux_estimated = magnet flux + delta Ls* Iq,

you will see the flux amplitude change a little when you command more torque and the resulting error in angle becomes

         angle error = atan(delta Ls/Ls * Iq/Isc),

written simply:

        angle error = atan(Iq/Isc * relative L-delta),

 

If you have 20% L-delta and you run at 50% of the short circuit current (Iq = 50% Isc) you would induce an angle error of about atan(0.5*0.2) = 5.7 degrees. This value should still work fine and provide high efficiency.

In summary:

-          Machines with high Isc are less sensitive to Ls changes

-          Unfortunately, small cheap machines often have low Isc and stronger saturation

  • in this case it is wise to aim for smallest error during the highest current, so compensate with a lower L value than the estimated one.

-          At lower current levels the angle error only gets smaller (needed accuracy of Ls is reduced)

-          Most motors also show unequal saturation in D- and Q-directions (saliency)

  • slight changes in estimated flux-amplitude changes on various torque levels gives an indication of this phenomenon.
  • While you can create fancy curves or look up tables for Ls-d / Ls-q compensation (necessary for best efficiency in highly salient IPM motors) it won’t be necessary to make the motor run smoothly.  Just try to not over-compensate the (saturated) inductance on the highest torque levels that you want to use.