Advice Wanted Re Vfd For 2 Speed Motor.

[strantor]

Actually there this all depends on the belt and gear ratios. If you change the pulley size you will get just as much torque, if you push the motor up to 120Hz, you will loose significant torque and Hp with these old motors and I doubt it will last long. Why go through all the trouble to see if you can burn up the old motor. As, already mentioned, for some odd reason 2 speed motors do not seem to work well on the high speed with a VFD. On the low speed you will have 4 Hp at best. The 2 speed motors I have seen in these machines was a Delta on low speed, a YY on the high speed.

Rereading my post I see that I left a lot of room for misinterpretation. Sorry for being unclear.

I was not suggesting wiring for high speed or low speed. I was not suggesting going up to 120hz. I was not suggesting any wiring changes at all. Nor was I suggesting any pulley changes.

I was suggesting leaving all the wiring intact and unchanged, and simply connecting the VFD between the mains supply and the power input of the machine. That way original 2-speed functionality remains (until the poor old motor is inevitably put to rest at the hands of PWM spikes). I realize that I was a little hasty in saying so; that is a bit of an over simplification, since the drive would need to be tied in with the start/stop/fwd/rev circuit. I didn't think about it hard enough to mention that part, but I still believe in leaving it 2-speed.

I then went on to suggest that, if I was forced to pick one way or the other to permanently wire it high speed or low speed (because/if for some reason leaving it 2 speed didn't work out <--- that wasn't stated, it was implied, and vague, sorry) then I would pick low speed and I would exceed 60hz. But I agree with your suggestion to wire for low speed and change pulley size, it's a better solution.

Thank you for pointing these things out.

 

[British Steel]

Hi, re VFD PWM spikes, you can save the motor by fitting a load reactor - just a 3-phase choke (or three separate chokes), usually sized by " percent " - 3% to 5% is usual, the inductance is calculated by taking the voltage across the choke e.g. 3% of the mains voltage at the rated current and frequency.

An example: a 4% choke for a 415v motor rated at 6A, 50Hz
4% of 415 is near as dammit 16v, the current is 6A so the reactance Z is 16/6, 2.67 Ohms (reactive)
Z = 2 * Pi * f * L
lets rearrange that: divide both sides by 2 * Pi * f:
Z/(2 * Pi * f) = L so 2.67/( 6.3 * 50) = L
L = 2.67 / 315 = 8 milliHenries (8 mH) - this is the value of inductor you need, rated for at least the max current your motor will pull. One for each phase!

How does it work? The chopped PWM signal is approximating a sine wave, so the square edges of the current pulses are very high-frequency components (multiples of the switching frequency), the formula z = 2*Pi*f*L show that the impedance (resistance to current change) of the inductor scales with frequency, so for (e.g.) a 5 KHz switching frequency, the switching component sees 267 Ohms, 100 times the impedance that the wanted 50Hz "average" sees - the harmonics (all odd-order with square waves, so 3f, 5f, 7f etc.) see even higher impedances, 3f sees 800 Ohms, 5f sees 1330, 7f sees nearly 1900 - this limits the current the harmonics deliver to the motor winding (and insulation!) so reduces the rate that the voltage on the winding can increase, making the voltage seen by the insulation much more like a sinewave and much safer for old insulation
I've wound suitable inductors for a 6-amp 415v motor on salvaged cores from 24v control transformers, hefty high-temperature enamelled copper windings and a "gapped" core - I removed the interleaved E and I laminations and stacked all the E's, all the I's together and put a thin sheet of PTFE in between to form the gap, which increases the allowable current before the core saturate (when it can't take any more magnetic field and stops acting like an inductor, basically).

You probably didn't need to know that, I do go on sometimes

Dave H. (the other one)

 

[richo132]

@dave H
Thanks for taking the trouble to share your wisdom on this subject.
I now know enough about this subject to be dangerous but nowhere near enough to be useful
Slavishly applying your formulae, I have calculated the choke inductance values for my 2 speed motor:
2800rpm FLA 9.02A -> 5.8mH
1450rpm FLA 6.2A -> 8.5mH
As I am unable to "roll my own", I looked for commercial products and ended up here:
http://www.zez-silko.com/en/products/reactors-for-frequency-converters
- where I discovered there are both Input Reactors and Output Reactors.
However, I notice that in both cases they are single value chokes.
So if I may draw further on your expertise, do you recommend an Input or Output Reactor and is it necessary to have separate reactors for each speed (current) ?

 

[British Steel]

No problem Richo,

Output reactors are what you want - input reactors are to (somewhat) correct supply power factor, reduce switch-on current and help contain RF interference on the supply to the VFD, in your circumstances they shouldn't be necessary!

You *could* use different inductors for the two speeds but myself I'd probably pick the 5.8mH if they were available - that link shows a max. of 2.5mH, which would probably be enough to protect your motor (there'll still be the attenuation of the PWM harmonics, just not quite as much, instead of 100:1 you'd be looking at 45-ish:1 or 25-ish:1 impedance ratios for the two speeds and 5kHz switching (vs 50Hz sine), again the ratios would multiply for the harmonics.

My own application is on a 3-speed motor with 4.9, 5.2 and 6A rated currents in the three speeds, I've used a single value of about 7mH as a bit of a compromise, so far the 60 year old motor hasn't died on me, nor has the VFD!

Hope that helps,
Dave H. (the other one)

P.S. - oops. forgot something important! Put the inductors between VFD and speed switch, try NOT to switch speeds with the motor running! That requires a bit more thought, putting together a snubber circuit (to deal with voltage spikes from the motor/switching and prevent them getting to the VFD) and programming the VFD with an additional switch wafer/set of contacts/microswitch to drop its output during switching.