from the bearing problems department...
motor bearing problems driven by vfd
Motion control, servos, steppers, etc. topic
Posted by ravi on 29 June, 2005 - 7:46 pm

i want to know how bearings of motors (running slower than normal speed given on data plate of motor, for any reason means to maitain pressure or temp.or any control) driven by vfd, worn out early than motors runs with normal (as per data plate on motor) speed & without vfd with DOL or star delta starter. i experienced same problem with many motors.

is there any remedy for this so that i can save motors from damaging?

thanks for ur help.

Posted by Trevor Ousey \(lists\) on 1 July, 2005 - 1:01 am

I would look at earthing, there are various whitepapers from VFD manufacturers advising cable and earth runs for motors on VFD's. From what I understand, current can get generated into the motor and passing through the rotor shaft and bearings, and this is what can shorten bearing life.


Posted by Chuck Raskin on 31 July, 2005 - 4:50 pm
Interesting Problem.
I just finished reading the Greenheck white paper ( on 'bearing current' and am wondering why, if this current was large enough to hurt the bearings, did it not trip the ground fault mechanisms in the experiments leading up to the conclusions of bearing current?

Is it more plausible that VFD spikes are generating bearing eddy currents? Are these
Eddy currents then causing bearing magnetization, leading to a rubbing or sliding effect of bearing surfaces? This would be especially true in lighter weight systems. I can understand why motor insulation and windings would have a problem, as the paper suggests, since they were not originally designed for pulsed mode operation. But the bearings are another matter.

One question raised is: Does bearing material, size, shape, etc. have an impact?

I agree with TB that the problem is probably mechanical. Perhaps a change in bearing
material, style, position (longer motor), size (larger motor diameter), etc., will alleviate the
problem. I, myself, have used hundreds of VFD's and have never had a motor prematurely fail
due to bearings, and I don't consider it 'luck'!

Posted by Peter Green on 2 August, 2005 - 4:49 pm
Chuck, I think you will find that you have got away with no bearing problems because you have been pumping water or some other conductive liquid.

I worked for many years in a Pulp and Paper mill and we had no problems with something like 50 or so VSD dating from the mid 1980s. Then we had two bearing failures.

One was a new drive that we left running overnight uncoupled. In the morning the bearings were wrecked.

The other was on a 100 Kw fan. Here the bearings failed within a couple of weeks.

In both the above cases the induced eddy current voltage had nowhere to go other than trough the bearings to the grounded motor casing.

All the drives that gave no problems had the induced eddy currents grounded through the motor couplings to the pump shaft and then to the process liquid which being aqueous and conductive was a very good ground.

Subsequently the mill made sure all fans and purely mechanical drives had a small brush contact fitted to the non-drive end of all VSD motors. As far as I am aware there have been no bearing failures since then.


Peter Green

Posted by Chuck Raskin on 30 August, 2005 - 2:38 pm
Actually Peter, I've used countless VFDs starting in the 70's, in applications including paper, food, welding, canning, bottling, packaging, conveyor lines, and general manufacture.

I prefer not to think of it as luck, but as an understanding of the devices I'm about to use, the application they're going into, and their overall integration into the system.

What was not indicated in the original writeup was:
The application data,
the environmental conditions,
the load type,
how long the motors ran before they failed,
the speed ranges they were operating in,
the Hp ratings of the motors and VFDs,
the manufacturers of the motors and VFDs,
how the units were wired (wiring diagrams),
if the main power legs were balanced,
the harmonic content of the system,
if the system power transformer had a 'wild' leg,
the power factor of the 3ph supply,
if the motors that failed were at the end of a long Bus run,
any preventative maintenance routines that were done,

There are tons of reasons as to why things fail in use, but having to apply a grounding 'brush' to wipe charges away from a motor shaft to keep currents from discharging through a motors bearings is not one I've ever come across before.

What I've learned over the years is:
1. At a minimum, always double the Hp rating of the required VFD, i.e. if the motor is 3Hp, use a 6Hp. This will aid the drive in dissipating heat when the motor is running at other than rated speed.

2. Never operate an AC motor at less than 75%, or greater than 125% of its No-Load RPM rating. AC Motor bearings are designed to operate at the nameplate RPM.

Question: Is it possible that AC motor bearings will fail prematurely if internal bearing lubrication does not coat the bearings properly at lower than rated, extended RPM use?

Question: Is it possible that AC motor bearings will fail prematurely if the internal bearing lubrication overheats and fails due at higher than rated extended RPM use?

3. If the motor has to run at less than 75%, or more than 125% of rated speed, because the gearing or motor cannot be changed out, use an Auto-Lube system to insure bearings receive their proper lubrication amount.

4. Use a STAR GROUNDING wiring scheme to insure issues that arise in one system do not adversely affect another.

5. Make sure the power draw from one or more systems does not adversely affect other systems attached to the same run.

There's more, but I think my 60 seconds is up.

Posted by William on 15 September, 2005 - 1:04 am

Thanks for your information and experience.

I would like to add that the VFD motor bearing failure has become serious in the recent years. VFD motor bearing failure was not a problem prior to 5-7 years ago. It is agreed by many experts in the field that the recent VFD motor bearing failure is due to the new high frequency IGBT that drives VFD. It is common to see more than 12k Hz carrier frequecy in a VFD drive rather than 2.5k Hz or so. The high carrier frequency gives a quieter operation but creates high motor shaft voltage and other issues such as insulation, amplified line voltage, grounding issues, etc...

Marathon motor, a Regal-Beloit company, recently completed a thorough testing on the VFD motor. You can contact Jerry Muelbauer at Marathon motor. You will find that a solution must be found to resolve the serious bearing failure of VFD driven motors.


Posted by Chuck Raskin on 21 September, 2005 - 11:45 pm
It's kind of interesting that reference to IGBT switching is constantly being made. I have to agree that the high speed switching of IGBT's is probably exacerbating the problem, and that induced currents are happening. But perhaps that's why I haven't seen the problem.

I've been using IGBT switching for AC & DC Brushless motors for years, and once again, without issue. If VFD IGBT switching waveforms are at issue, remember that AC motors are designed to run on 50Hz or 60Hz sinwaves only. Once chopping takes over at any frequency, all bets are off.

I think about the ability of pulsed waveforms to not only induce currents into the iron, but also magnetize the bearings and either move them or make them freeze in position. In either case, bad things can happen including arcing. As I write this, I think about shaft and bearing material differences in Brushless and standard low frequency AC motor design. Could it be that the VFD to use with a standard vanilla wrapper AC style motor is one that generates a sinusoidal current waveform, rather than a pulse or switched waveform? This is not to say that IGBT's cannot be used, but that perhaps sinusoidal current control should be used.

If the problem is simply a waveform issue, then buy an amplifier that generates the correct waveform!


Posted by Vesna on 30 September, 2005 - 1:14 am

I have problem with bearings on the position in vertical motor for Bingam pump (type 4x6x10L JCV). Bearings type: 35BC02XPP and 50BC03JP. I need more information about this type of bearings and drawings (if it is posible).

My e-mail:


Posted by AdamW on 15 September, 2005 - 1:08 am
Hi Peter, Chuck,

All PWM drives will cause electrical current discharges in the bearings as described in the Greenheck white paper. This is an inherent problem with VFDs because of the induced voltages on the shaft from the IGBTs. We have measured 25 to 70 volts pk routinely on even small VFD motors. VFDs will cause the bearing race to frost and pit and potentially fail - sometimes within only a couple of months.

A good shaft grounding solution is on where a conductive microfiber shaft grounding brush is described. This brush is easy to install, maintenance free and a reliable solution to prevent the potential failures in VFD motor bearings due to electrical discharges.

Hope this info helps,

Posted by Les Smith on 27 January, 2007 - 1:35 am

Watch your lubrication cycle also. To many motors are lost due to improper lubrication method or interval.


Posted by Tom Bullock on 1 July, 2005 - 11:43 pm
Are you sure the problem isn't mechanical?
Do you have a flexible coupling between motor and load?
Is there a side-bearing load on the motor?

Tom Bullock

Posted by Steve Myres on 2 July, 2005 - 12:12 am
I haven't heard of that before. When I've selected bearings for designs though, the rule has been the higher the speed, the higher rated bearing you must select, so I would think lower speed should be easier on the bearings.
Steve Myres, PE
Automation Solutions
(480) 813-1145

Posted by William Hinton on 1 July, 2005 - 11:58 pm
There is a great link to US Motors about variable Speed Drive requirements, follow the link:

Direct On Line starting and Wye-start Delta-run do not cause harmonic distortion. All brands and types of adjustible frequency drives have a carrier or switching frequency in the thousands of cycles per second that cause high voltage spikes and harmonic distortion. Harmonic distortion is like trying to run the motor at its base speed and three times for the third harmonic and at the fifth and seventh harmonic for example. Each current is causing tourque and the rotor is trying to respond to the magnetic pulses. This causes stress and strain on the bearings as well.

There is another device to remove the spikes and phase voltage harmonice from adjustible speed drives and from any other source all at the same time, follow the link:

I hope this helps.

Posted by CCOMM on 6 July, 2005 - 5:08 pm
Do you have any of the failed bearings available for inspection? If so, check for the signature "fluting" wear patterns of EDM (Electro Discharge Machining) in the bearing races. Shaft-to-ground voltages exist on all motors, but PWM inverters up the ante, and are more prone to causing this condition.

This company has a web page giving a good explanation of EDM-related bearing failure, and how VFD drives can cause it. oduct_guide/fa117_03

A google search on 'EDM and "bearing failure"' will kick up a lot of returns, including:
(mostly about alignment, but some cites for EDM). publications/2%5B2%5DPESC'01.pdf

One first step in solving the problem is verify the motor and drive ground bonding scheme equals or exceeds the drive manufacturer's recommended practices.

Long wire runs between drive and motor also figure into the condition, as does (to a lesser extent) the type of wire used - inverter-rated cable is preferred.

Other possibilities worth investigating if EDM has een confirmed include:
Failed motor bearings can be replaced with ones featuring a ceramic non-conductive coatings. Adds expense. Install a shaft "grounding" brush arrangement.Adds exepnse, and requires ongoing maintenance. Use motor bearing grease featuring low bulk resistivity. Much more expensive than regular grease, and efficacy is in dispute.

Posted by Larry Hough - The 84 Group on 13 July, 2005 - 2:07 pm
The cause of shortened bearing life when using VFD's with AC motors is definately due to leakage currents through the bearing. There are several causes, all due to the high frequency harmonics that the IGBT create when they turn on and off.

The frequencies going to the motor can be reduced by applying output reactors or, more expensively, harmonic filters between the motor and the drive. Shielded cable with symmetric grounds and proper grounding techniques will also help.

Typically severe bearing problems are only experienced with large motors (greater than 200 HP).

The best solution is to apply grounding brushes to the bearing, providing a current path to ground that is not through the bearing balls or rollers.

Posted by Greg Scmitz on 12 March, 2008 - 1:18 am
I have to most respectfully disagree with the statement.

We have seen this with as low as 7.5 hp motors that had the old VFD replaced with new ones. In all the applications where we did that (up to 15 hp) we had the same thing happen.

Distinctive (and not a lot on it but SKF documents it) is the grease being burned up as well. It has a very distinctive electrical arching odor, like you got from the trains or slot cars I ran when I was a kid.

We also had one 25 hp motor that periodically took the bearings out and finally confirmed that it too was VFD related, just not as quick to create bad bearing noise as the newer ones (less than 3 months for the noise to occur).

Having shielded bearings may (and I mean may) help as they seem to offer an altnertive patch that does not go through the bearing and grease.

Posted by rich weiner on 20 July, 2005 - 7:45 pm
It has been many years since I first saw a study done by an IEEE Pulp and Paper engineer, but it went into depth on the "fluting" problems on bearings that developed when using VFDs. It had actual voltage/current measurements, photographs of the damage, quantitative data, and the recommended fix (shaft grounding brushes). It was written and presented by one of the major bearing manufacturers (I cannot remember which one).

You may be able to do a search on bearing fluting and find much information but most importantly to note is this is a KNOWN occurance and can be fixed. Good luck.

1 out of 1 members thought this post was helpful...
Posted by Phil Corso on 26 July, 2005 - 6:34 pm
Responding to Ravi's Jun 29, 7:45pm query... before I toss my BGE into the PSH please provide the following:

1) Manufacturer - All the same? Different?

2) Motor - Rating & speed?

3) Construction - Vertical? Horizontal?

4) Bearing type - Ball? Roller? Sleeve?

5) Control mode - Constant Hp? Constant torque?

6) Failures - Drive end? Non-drive end Both?

7) Statistical analysis - Scatter Plot? Frequency plot?

Phil Corso, PE {Boca Raton, FL, USA}
[] (

Posted by Anonymous on 14 September, 2005 - 9:02 pm
Another solution is Yaskawa's G7 inverter with 3 level control. With 3 level control (12 transistors per phase instead of 6 with standard 2 level control) the transistors are switching at 325 VDC vs 650 VDC. This allows for extremely long motor lead lengths without an output reactor and significantly reduces any bearing currents. Refer to for additional information on the G7 drive.

Your use of this site is subject to the terms and conditions set forth under Legal Notices and the Privacy Policy. Please read those terms and conditions carefully. Subject to the rights expressly reserved to others under Legal Notices, the content of this site and the compilation thereof is © 1999-2014 Nerds in Control, LLC. All rights reserved.

Users of this site are benefiting from open source technologies, including PHP, MySQL and Apache. Be happy.

As the poet said, "Only God can make a tree" -- probably because it's
so hard to figure out how to get the bark on.
-- Woody Allen
Advertise here
our advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive