VFD Puzzler

[jraef]

OK you VFD gurus, here is one that has me stumped.

7.5HP 230V centrifuge application. Motor is designed for 150Hz operation at 230V, 2 pole, 26FLA. Set up the VFD with a custom V/Hz ratio to match the motor design (150Hz at 100%V), coast to stop (no braking). Tested and ran beautifully at the shop, started and stopped maybe 20 times, ran under load etc. etc., no problems. Shipped the unit to jobsite in Salt Lake City UT. Field install team "claims" they never touched the motor connections, but 2 screws had come loose during transit inside of the VFD, so they removed it and turned it upside down to retrieve the screws. They also claim they had not energized it yet at that point so no possibility of internal arcing.

When told to run, the VFD went through normal "speed search" routine (a.k.a. flying restart), found the initial speed and began to ramp. It immediately tripped on VFD OL (not motor OL, not OC, not GF). Found input voltage to be only 213V, so adjusted the VFD line voltage parameter to match, which should have reset the V/hz pattern accordingly since it was a percentage of line voltage anyway. No difference. After much trouble shooting we got down to connecting a separate 230V 60Hz 1HP motor, it runs fine. Reconnect the centrifuge motor, same trip. Disabled the Speed Search function, no change. Uncoupled the motor from the centrifuge bowl, same trip. Reconnected the VFD to test motor, ran fine again. Meggered the VFD motor, no problems. Did a "differential phase winding current test" (I am unfamiliar with that one) and was told the motor windings are "perfect". Cleared the VFD memory and re-entered programming exactly as it had worked in the shop prior to shipment, reconnected the centrfuge motor, same trip!

Brain hurts... must get sleep... any ideas to try in the morning (US MST)?

"Venditori de oleum-vipera non vigere excordis populi"

 

[MikeHalloran]

Warning: I claim zero expertise with VFD drives. Conjecturing just from the behaviors you reported...

As soon as the VFD gets ready to apply serious power to the motor, and tries to go into 'closed loop', it thinks it's looking at a short. As if the impedance that it measures across to sense motor current is now abnormally large, or open. But that behavior is linked only to the centrifuge motor, is not dependent on the actual load on the centrifuge motor, and is not linked to the dummy motor.

I'd be looking for loose screws, wire nuts, etc., within the centrifuge motor, and in wiring that was not verified with the dummy motor. Maybe the windings are perfect, but it sounds to me as if the connection to one of them is somewhat less than perfectly perfect.




Mike Halloran
NOT speaking for
DeAngelo Marine Exhaust Inc.
Ft. Lauderdale, FL, USA

 

[BigLe]

Did you put the 2 screws away? It is not terribly difficult to remove the front panel of the drive and you can at lease see if there are more screws stuck in between the boards..

 

[jraef]

BigLe,
Yes, the screws were actually for the motor leads (not tight enough prior to shipment) so they were of course reinstalled prior to energizing and testing, and yes, they appeared to be the only 2 that had come loose and missing.

MikeH,
I did eliminate loose connections as one of my first troubleshooting paths, I agree that was a likely suspect at first glance. Wheen we connected the test motor we tapped in at the centrifuge motor connection box so as to eliminate the possibility of a bad wire from the VFD to the motor as well.

"Venditori de oleum-vipera non vigere excordis populi"

 

[Marke]

Hi Jeff

I have seen this sort of thing before with a very high inertia load where the motor just did not develop enough torque near zero speed to get th load spinning quickly enough. I cured this by adding voltage boost to increase the torque available at near zero speed.
If the voltage is too low, the drive can wind up to greater than the full load slip frequency without the motor beginning to spin. Beyond that, the impedance of the motor as seen by the drive, falls rapidly as does the torque and you get an overload.
- Just a thought.
- some drives have a timed voltage boost just for getting high inertia loads to breakaway.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[aolalde]

Jraef:
You do not clarify it but I am assuming that your test did not include the centrifuge. If the motor started nicely but it was not engaged to the load inertia, that was not a valid test.

We can not go against the basic mechanic laws.

alpha = T/j

alpha= angular acceleration.
T = accelerating torque.
j = Total system inertia.

To accelerate successfully the centrifuge, the first step needs enough torque at very low frequency to pass the high slip area of the motor speed- torque curve. After that is achieved the frequency and voltage must increase pulling the inertia into speed and staying in between the low slip area of the speed- torque curve. The time in seconds required to reach the full speed could extend to several seconds. A thermal analysis must be done to define how fast the increase of speed can be done.

 

[DickDV]

This is likely a vector drive and should require a motor ID run to build a motor model. If this model is lost somehow, a new motor ID run should be done. It sounds like this is the case here.

For testing purposes, run the accel rate out to several minutes and set the drive speed reference at .5hz. See where the current goes when the drive is started under these conditions. If it doesn't fault, slowly increase the speed reference to see if the motor will follow and start to move. If this works and the motor model is good, it seems that a longer accel ramp or at least an S curve start to the ramp would help.

Advise on these results and I'll try to help further.

 

[jraef]

Well, I think I found it!

Customer had "assumed" the 6 wires in the motor terminal box were because it was a 6 lead dual-voltage motor. I looked at the nameplate and saw that it was a single voltage design, making me wonder what the other 3 leads were for (no information from the motor mfgr had been provided). Removed the terminal box to see what was going on and observed that one was just a spare wire going nowhere (capped), the other 2 were going to a thermistor! So the OEM had hooked up the thermistor to 2 of the motor leads, applying the 230V PWM from the VFD directly to it! It apparently worked in the shop many times because the resistance of the thermistor was high enough to fool the drive, but I imagine that it eventually melted, which probably now has made a phase to phase winding short inside the motor. The "test" that the field tech had done on the motor was a High-Pot test to ground, not a phase-to-phase megger test. We are getting a megger in later, but I am almost sure that we will find a short.

"Venditori de oleum-vipera non vigere excordis populi"

 

[aolalde]

For 3 terminal leads, 3-phase motors, internal interconnection of the three phases is required. Then a megger test will not be able to detect winding phase to phase short-circuits.

 

[Marke]

Well that was pretty left field!!
I don't think anyone would have thought of that without looking at it!!

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[MikeHalloran]

I guess that has to be the second stage of troubleshooting:

- Remove every cover and look inside.

- Witness a rerun of every alleged test, and watch with a child's eyes.





Mike Halloran
NOT speaking for
DeAngelo Marine Exhaust Inc.
Ft. Lauderdale, FL, USA

 

[jraef]

Follow up...
Well they megged the motor (and yes, aolalde you were correct of course, but I found the other ends of the windings and opened them up for testing). Unfortunately they did it while I was out to lunch, so once again I did not witness it, but they said "everything was fine" and put the motor back together again by the time I got back! Grrrrrr....

And of course, when we tried the VFD again (still uncoupled from the load), it tripped again. I am still not convinced that the motor is OK. I had to leave the jobsite, but I am having the motor and drive yanked out and sent to the shop so I can test them myself with my own tools next week. Just to satisfy their curiosity I will hook up a new VFD to that motor and see if it does the same thing. If it doesn't (i.e. the new VFD works), my credibility will be out the window!

"Venditori de oleum-vipera non vigere excordis populi"

 

[peebee]

jraef -- Please clarify that your load test at the shop was performed using a true centrifuge load?

 

[jraef]

Yes, the original load test at the shop was with the centrifuge coupled and we loaded the bowl with water to simulate the operation (the usual load is milk). Drive and motor have not arrived yet for further testing.

"Venditori de oleum-vipera non vigere excordis populi"