I've seen on several websites that a variable frequency drive (VFD) can be used as a starter for induction motors, but didn't find any details calculation on what should be the starting frequency.
In a typical delta-star starter, the start configuration is used first to provide a lower voltage. Should I reduce the frequency such that the RMS gets equal to the star configuration? I don't think this is the right way. From which frequency should I start running?
If the motor is not spinning, then from zero hertz or cycles per second.
If the motor is spinning, this is termed "catching" it and some VFD's can do this and some cannot. Check the feature set of the chosen VFD to see if it includes this capability.
A VFD, at zero hertz, locks at least one of the phases at any angle (full current) so that the motor can't spin. As the frequency is increased, the phase currents are alternated, resulting in rotation.
Some VFD's will let you control the magnitude of this (stand-still) current and some will not. Many will let you define the motor type (such as induction, synchronous, servo, etc.) and this directly impacts the drive methodology and capabilities. Again check the VFD features.
If a large motor, also consider DC injection braking via the VFD. That may be more convenient than using say, a resistor bank or mechanical brake to dissipate that power. The motor may get warmer from DC injection braking however.
I am currently working on a "simple" DIY VFD design, and have had some success in being able to start and run a small three phase induction motor at frequencies from 10 Hz to 90 Hz. Something to consider is that an induction motor operates with a certain amount of "slip", which is why a 2 pole 60 Hz motor may be specified as having a speed of, say 3450 RPM, instead of the 3600 RPM synchronous speed. The 150 RPM difference is due to the "slip" that allows the motor to rotate by inducing current into the shorted turns of the rotor. 150 RPM corresponds to a frequency of 60 * 150 / 3600 = 2.5 Hz, or about 4% slip.
So you could apply 2.5 Hz to the motor and it may not rotate, especially if it is connected to a load. Using a simple V/f method, the corresponding voltage for a 220 VAC motor would be about 9 VAC. This would be a good starting point and then you could increase the drive while monitoring the current. Starting under load may require as much as 3-4 times nominal rated current, which can be tolerated for several seconds.
I am still in a fairly early design and testing stage, but I plan to verify some of the assumptions I have provided here.
