Calculating Motor Torque on VFD

[Barry1961]

A customer has a 10 hp, 480 V, 3 ph, Baldor Motor # CWDM3714T driving a mixer through a 100:1 helical gearbox. The motor is being driven by a WEG VFD in the volts/hertz mode, (not vector) at 25 Hz. The following reading were taken by a old clamp on amp/volt meter, Simpson I think, which does not have very fine scale for reading voltage. The PWM carrier frequency is 5 khz.

The customer is going to change the ratio to run the motor at 60 hz after testing and wants to calculate how much torque he is really using. He is getting someone to fix the unbalanced incoming line.

The load is pulsing at about 1 hz which is reflected in the amp draw.

Line coming to VFD
Voltage from phase to phase, 480 V, 480 V, 460 V
Current, 6 A, 6 A, 3 A

Motor Leads, Motor is rated at 13.5 amp
5 to7 amp pulse, balanced

VFD current display
13 to 15 amp pulse

Which reading can I believe to calculate torque at 25 hertz?
Do I need to get another amp meter to read the 5khz carrier?
Will the unbalanced line affect this besides heating up the buss diodes?

Baldor Motor Data
7.8 A at 25% FLT
9.2 A at 50% FLT
11 A at 75% FLT
13.5 A at 100% FLT
16.2 A at 125% FLT

Thanks!!!

Barry1961

 

[Skogsgurra]

I would trust the VFD display more than any other measurement device.

A sanity check is to find out if the motor runs normal hot when drawing 13 - 15 A.

If it does, then the clamp's 5 - 7 A is probably very wrong. It corresponds to less than 25 % motor load. And if that were a true reading, the motor would run very cool.

It doesn't matter if the clamp is old or new. It is often difficult to get a reliable reading using a clamp-on ammeter on VFD drives.

Check the speed with an accurate tach. The speed should be a little lower than 1769, which is speed at 100 % load.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Barry1961]

They are running the motor at 25 hertz now which is the speed they need to run for the application. This is a 4 pole 1769 rpm motor so the motor rpm is going to be close to 725 rpm at 25 hertz.

The motor is running cold, at ambient.

Barry1961

 

[Skogsgurra]

Does the display really show 13 - 15 A?

Or does it show some percentage that has been transformed to amps? The latter is failure-prone if the 100 % level isn't known.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[jraef]

But to address your other issue, the unbalanced line should not be affecting the output of the drive. As you suspect, it may not be good for the converter section but once converted to DC and filtered, it should have no direct relation to the VFD's output.

Hmm... unless perhaps the VFD's DC bus filtering is woefully inadequate? Not sure about that though, I've never seen that, even if the input is 1 phase.

 

[Barry1961]

I was told from another source that if I put the VFD in vector mode and read the percent torque parameter it should be fairly accurate even at low hertz. The VFD self tuning would compensate for hertz and voltage variations.

He also said you would need to know the VFD output voltage, efficiency and power factor along with current to calculate torque.

Does reading the percent torque in vector mode sound fairly accurate?

Barry1961

 

[Skogsgurra]

Before I try to answer that last question, I need to know if your display reads Amps or percentage of rated Amps. Which one is it?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[DickDV]

As skogsgurra has already said, do not try to measure anything on the motor leads. Use the VFD display instead.

Also, as skogsgurra has said, you need to know exactly what the drive display is telling you, especially if the display values are in percent (percent of what?).

You can pretty well figure that the motor is producing its rated nameplate torque of about 30ft-lbs when it is drawing 13.5amps. That will be true for any frequency from 10 to 60hz. That's assuming the drive is set for constant V/Hz.

The tricky part is determining the motor torque when the load and the current are less than full load. It is not a simple proportion but rather a vector sum which involves some calculation. Better to let the drive do the math and report the torque for you.

To get the drive to report torque, you will need to reprogram the drive to run in sensorless vector mode. When that is done properly, I would expect the reported torque to be within 2-3% of actual.

An added bonus will be a stiffer, more stable speed under changing load conditions if you tune up the speed loop controller properly.

 

[DickDV]

I looked at the OP again and see that, at the bottom, the motor current at four levels of torque is specified.

Using that information, you could avoid changing the drive to vector mode and instead, simply compare the motor current draw to that list. It is important, however, that the drive V/Hz ratio be constant when doing this.

 

[Barry1961]

The customer was able to get some support from WEG at last.

The WEG was displaying amps and not percentage. But the WEG guy says go by the percent torque in sensorless vector mode and it should be within 2-3%.

The readings from sensorless vector percent torque seem to be about what I expected from the limited mechanical torque readings I did which is about 60% of FLT.

I have seen high currents displayed on other drives while operating at low hertz and very low load but the motor was hot. I assumed the low voltage boost was the reason. This was why I became wary of the displayed amps on any drive at low frequency.

Thanks for all the help!!!

Barry1961

 

[sreid]

The customer may want to know the torque required (for whatever reason) but if the gear ratio is doubled and the motor speed is doubled (so the output shaft is turning the same speed), the motor torque will be cut in half.

 

[Barry1961]

With the tests done so far the final ratio should be around 240:1 which would more than double the reduction. The load has a fair amount of inertia to overcome during start up so the peak motor torque should drop even more since reflected inertia is divided by the square of the ratio.

The cost difference between a 3 hp and 5 hp gearbox at 240:1 will a couple thousand dollars so they are trying to make the next gearbox the right one. To do this they need to get an accurate torque reading.

Thanks!!

Barry1961

 

[DickDV]

At those hp levels, I would choose the gearbox ratio so maximum speed on the load machine equals 90Hz on the motor.

Starting torque, running torque, and motor cooling are all optimized that way.

 

[VinceS]

Do you have the specs on the drive? I have had trouble calls on large motor sized drives using small motors. The rectifier section utilized phases A and C largely with about half the proportional amp load draw on phase B. As to the 20 volt delta on phase C, the local POCO specs may provide a surprise as to the slop they are aloud to provide.