Motor Bearing Currents 1

[DriveSystemExplorer]

Hello to all. This is my first attempt at this, so please be gentle.

My question has already been discussed here back in February (thread has since been closed), but I have need for some more input.

A customer of mine has a 250HP AC motor operating from a PWM Inverter. The motor has a NDE insulated bearing that failed after a few months of operation. The bearing was replaced and the motor shaft was fitted with an Aegis grounding ring. The bearing failed again after a few months. Machine/equipment grounding techniques seem OK. What else can be wrong?

I know Gunnar mentioned on February 16/06 that he had a 2003 presentation as well as measurement techniques that address this issue. I would be grateful to get that information.

Thank-you in advance.

Craig Davis

http://www.leadingedgeautomation.com

 

[ozmosis]

Craig
Are you sure the bearing failure is due to the effect of motor bearing currents?
If yes, and you have tried insulated bearings and the Aegis grounding brush, as well as ensuring the motor is grounded correctly (btw, are you sure the motor ground is the same ground potential as the inverter?) then it might be worth checking out an output filter (sinewave filter) to decrease the steepness of the voltage rise. I have used Schaffner to good effect.

 

[Skogsgurra]

Hi Craig,

I feel compelled to answer your question.

It is no wonder that insulated bearings do not work. And no surprise either. It is more of a surprise that the Aegis ring doesn't help.

The simple test you can do is to measure shaft voltage with a reasonably fast memory scope. Attach probe ground to motor ground and use a screw-driver to contact the rotating shaft and measure voltage. You have to be very careful so you don't hurt yourself.

Scope settings 1 or 2 us/div and 2 or 5 V/div, Single sweep. Trig level a few volts, negative edge.

Press Single and see what you get. Mostly, you will just see a lot of signals that are difficult to interpret. But once in a while, you will see a typical discharge wave-form. It looks like this:

The green trace is shaft voltage, the red trace is shaft current measured with a rogowski coil.

The characteristic fast edge going from a positive voltage (usually anything between 5 and 15 volts) to zero is what you shall be looking for. If you have lots of them, then you have EDM in your bearing.


Why do not insulated bearings work? Simply because the insulation is a low-frequency insulation that is very good if you want to protect the bearings from current damage like welding currents, return currents in tracrion motors etcetera. But the insulating layer is not very thick (50 - 100 microns) and it consists of Al2O3 (alumina) which is a good dielectricum. So you have a capacitor between outer ring and housing. The capacitance can reach high values. I have measured close to 10 nF in a 6279 bearing. So the HF from the PWM does not see much impedance.

I am surprised that the Aegis did not work. That is an indication that you have severe problems. What motor voltage? What carrier (switching) frequency? Use as low a switching frequency as possible - that saves the bearings. Cable length sometimes play a role. Especially if it is frame voltage that is your problem. Please describe the installation better. Motor, motor bed, cables, inverter, couplings, gear, driven machine etcetera.


Gunnar Englund

http://www.gke.org

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

 

[Skogsgurra]

Sorry, read "a 6214 bearing"

Gunnar Englund

http://www.gke.org

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

 

[DriveSystemExplorer]

Thanks Guys,

sed2developer: yes I'm pretty sure it's EDM based on the race fluting observed. Thanks for the filter suggestion. I'll look into it.

Gunnar: The site that I described is down in Florida and I'm near Toronto, so I haven't seen it yet. But I do have another problem site nearby that I have seen. There they have lost 2 bearings in the past year and we are about to install an Aegis ring. This one is a 125HP, but it uses the same parameters as in Florida.
460VAC, 2.5Khz carrier, cable run only about 20M, I've witnessed the installation (motor ground comes right back to the inverter, which is as SSD 690+) there is a 360 degree ground connection around the VFD cable screen where it enters the drive cabinet, and it is a belted-duty application (static perhaps???), the 125HP is running part of a commercial printing press, the Florida 250HP is running an extruder.

I'll try the measurement procedure you suggest at the location nearby.

Thanks again


Craig Davis

http://www.leadingedgeautomation.com

 

[Skogsgurra]

Craig,

I tried to send you a mail at the address indicated on Leading Automation home page. It bounced. The main contents was a presentation discussing the main three failure mechanisms. I think that it could be of some interest to more members, but I do not know how to make it available. Are there sites like Tinypic where you can load a file and then put the link to it here?

Gunnar Englund

http://www.gke.org

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

 

[itsmoked]

Gunnar:

None are as painless as tinypic..

But this one is not bad.

http://www.verzend.be/

The download-e has to wait <10s and look at.. what? I don't even see anything. But it works fine. I tested it for you.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

Thanks a lot! I have put the pdf at

http://www.verzend.be/v/1927226/Sunnepresentation_EDM_2003_a4_w_comments.pdf.html

I haven't tested it yet. But if Keith sez it works - it works.

Gunnar Englund

http://www.gke.org

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

 

[itsmoked]

High praise .. and I thank you.


Seems it has worked, as 21 people have downloaded it.

Nice clear document.. You really should translate it, Gunnar. Especially since all the hacking/formatting with the figures and pictures is kind of the hard part and you've already done that.

I would be glad to proof read and edit it for you, if you want to just do a "quick" translation.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com