# What should be the phase difference in the two windings of a split-phase induction motor?

I'm having a little basic problem here.

I'm trying to reconnect the coils of a 4 pole motor (concentric winding) to make a 2 pole motor.

I've managed to make it work in 2 configurations so far (in two different voltages), but the one thing I never get is the fact the sum of the currents (main + aux) is always resulting in less than the greatest one.

In other words, the phase difference is always well above 90º.

If it was 90º, -90º, -270º or 270º the total current would always be bigger than the greatest current (and if both were identical it would be sqrt(2) that current).

But according to the phasor sum I made it's always like 130º, 140º, 150º out of phase etc.

I tested caps from 4 uf to 100 uf and tested several configurations but I never got a 90º difference.

Am I wrong to expect a 90º difference to begin with?

• With a phase angle of 130 to 150, the sum will be less than either individually so, why not try a different value capacitor or, in case I've got hold of the wrong end of the stick, what's the problem here? Aug 28, 2020 at 14:35
• @Andyaka My problem is that I thought a 90º difference should be ideal but I'm never even getting close to it. At least not with similar current and field in both windings. With 4 uf for instance the current in the aux winding is always too small, whereas with 100 uf it will be closer to the main winding (or bigger in one of those configurations, where the aux had less turns) but the phase difference is never getting close to 90º. I'm not sure if adding greater capacitors is getting it closer to 90º but I have the impression that it isn't. Aug 28, 2020 at 14:41
• @Andyaka And not using the a running cap won't do, it's torque is way too uneven and I have the impression it also has some off-axis forces. It's vibrating terribly and the coils are "crying" too much. But with the running cap it stays quiet. Talking about quiet one of the quietest I got was using the main winding (after starting) plus connecting the aux winding with itself through a cap. That was the smoothest run so far but I have no way to check how actually functional it is, I need an oscilloscope to analyze its power factor and the efficiency in producing mechanical power Aug 28, 2020 at 14:45
• It doesn't really matter as long as there is enough to establish a definite direction while starting. If 4uF starts the motor reliably, use it. If you need a high starting torque, use 100uF or 300uF - remembering that starting current in the main winding will be immense - and such motors tend to use a centrifugal switch to disconnect the cap at speed (which at that value will be an expensive fragile bipolar electrolytic).
– user16324
Aug 28, 2020 at 17:28
• @BrianDrummond But the problem is that using only the main winding while running is making it a terrible motor. In none of the configurations I made, for instance, it runs up to 3600 rpm by itself by giving it an initial impulse (like a normal 2 pole motor does). Instead it goes to something like 800 rpm and stops there. If accelerated it brakes down to 800 rpm again. It only goes to 3600 rpm if I give quite a speed or if I use the aux winding with a huge capacitance. And, even after it reaches 3600 rpm it just sounds terrible running only with the main winding. Aug 28, 2020 at 18:28

Single-phase motor designs are a difficult compromise. They are designed in various ways depending on which aspects of performance are most important for a given application. See my answer to your question: Why 2 pole induction motors don't use auxiliary windings while running (while others do)?

All single-phase motors are subject to vibration. The stator magnetic field pulsates rather than rotates as it does in a three-phase motor. Mathematically, the pulsating field can be fewer as two magnetic fields that rotate in opposite directions. Once the motor starts to rotate, the resulting torque increases in the direction of rotation. Torque pulsation and vibration can be reduced by distributing the windings as well as possible to make the shape of the field more sinusoidal as mentioned in an answer to your question: What are the properties/advantages of concentric winding?

The phase difference between the main winding current and the auxiliary winding current can approach 90 degrees with some capacitor value selection, but only for a particular load torque and point of the torque capability vs. speed curve. The capacitor value, the winding design and the starting scheme must be selected according to the best performance for the application as mentioned in my answer to the first question linked above.

Single-phase motors are avoided by industrial users who have access to three-phase power. One of the pioneers of AC motor design said something to the effect that the best single-phase motor design is not as good as a below-average three-phase motor design.

The best single-phase motors are designed for one spee, one voltage and known load characteristics. A motor such as yours with multiple windings that can be reconnected in various ways has forfeited much opportunity to design for good torque vs. speed performance, high efficiency and low vibration.

• If instead I used a source that produces two phases in 90º (like the old 2 phase system), would the performance be similar to a 3 phase motor? It should keep 90º even under different loads in that case, right? Aug 30, 2020 at 22:58
• If you have two sets of equivalent windings that are distributed to produce a sinusoidal magnetic field, a symmetrical 2-phase source will produce performance that is comparable to that of a 2-phase motor. I believe that falls short of the performance of a 2-phase motor because the power conversion is not as constant. I will see if I can find a reference about that.
– user80875
Aug 30, 2020 at 23:56
• I was mistaken about the rate of power conversion not being constant. If the windings are identical, and the winding configuration is optimized, the performance should be equivalent to 3-phase motor performance.
– user80875
Aug 31, 2020 at 11:58