Modeling Part-winding motor start 1

[dpc]

I'm trying to model expected voltage drop for starting a 200 hp motor using part-winding starter (not my idea).

I don't have a lot of experience with part-winding starting, but my understanding is that the main purpose is just to reduce the initial inrush current then quickly transition to the full winding to accelerate the motor.

So is it reasonable to assume that when the starter switches to full winding mode, the motor will be drawing close to locked-rotor current?

Thanks,

Dave

 

[DickDV]

The object is to use the start winding to accelerate the motor up to some fraction of the full speed, typically 75%. Of course, the load must be light enough so the low-torque start winding can actually get there.

And, yes, there will be a second inrush when making transition to the run winding. It's magnitude will be related to how far up the acceleration ramp the start winding was able to go.

 

[dpc]

Thanks Dick. I have some old US Motor info indicating that in some cases, the motor might not even turn until the transition is made, but maybe that's an extreme situation.

My understanding is that the time on the part-winding is quite short - maybe l or 2 seconds?

 

[jraef]

Part winding is a bit different from other 2-step starting methods in that there is no reduced voltage, just a reduced number of windings being energized. Keep in mind that in a PW starting scheme the remaining windings are just added in, not reconfigured. So the flux penetration into the rotor should be the same as if started X-Line, meaning that you will not have to contend with increasing the flux when transitioning. Combined with the fact that it is a closed transition (the motor is not disconnected during the changeover) and the transition spike should not be as bad as, say, an open transition Y-Delta or Autotransformer transition spike. Starting current is kept to 65% of X-Line which means 390% of LRA, and my experience is that the transition spike has never been more than 450% when set at 2 seconds transition.

The pitfall in modeling it is that the transition time MUST be short enough to make sure that you do not allow the motor to fall into the nadir of its Pull-Up torque, the dip in the torque-speed curve that occurs at 50% speed, at a time when the load torque requirements are greater than that value, otherwise you enter into a stall. So the transition time may end up being shorter and therefore the spike higher than you want it to be, but you will have no choice in the matter because there is nothing to adjust on the PW starter other than transition time.

 

[dpc]

Thanks, Jeff. I used 4X FLA, so I guess that was in the ballpark.