Does anyone have experience or knowledge of a technique for detecting shaft currents, (VFD induced), with a vibration analyzer?

I know shaft voltage can be measured by picking up a shaft contact and current can be measured with a Rogowski coil. And I've seen the SKF TKED 1 detector (EDD pen). However, I heard a report that a standard vibration analyzer (2130?) had been used to detect electrical discharge in VFD motor bearings.

Has anyone else heard of this?

Thanks.

If you want to detect the discharge itself, good luck - I haven't heard of that.

If you want to detect the resulting damage - different story. Not too much different than typical outer race defect from what I have seen on the forum.

What you've stated is what I have always understood to be the case. When I heard a report that a vibe guy claimed he was detecting electrical discharge at a bearing by watching the real time display of the spectrum, I thought "there's something I need to learn about".

Thanks for your input.

I don't know if the 2130 can be set up to measure current or not. You can measure the voltage (potential) between ground and the shaft with a brush, just like you would use a volt meter. Just set up the data collection point like you would a gap voltage point for prox probes. It would require a cable with a ground clip on one wire and brush on the other. Attach the clip to ground and apply the brush to the shaft. Just make sure you are away from keyways and keys and most of all, safety inspectors. If the 2130 can be set up to measure current, the cable and attachment would be the same. I would prefer to use a VOM and just measure the current and enter into the database manually.

Regards,
John J

John J,
If the problem due to VFD, then the Rogowski Coil is the only instrument that will measure the high frequency currents. Just reading from the shaft will not work.
Regards,
Ron

Went through a very long and painful exercise at a large steel mill clients facility last year doing just that......a couple of very large DC motors had been replaced by induction motors and IGBT drives. They were driving via gearbox very large rolls which were smothered by cooling water. The rolls began to corrode rapidly, and due to surface finish requirements requiring frequent enough reconditioning and refinishing that production was being impacted. All sorts of possible causes were explored, to no avail. Due to awareness of bearing fluting and other VFD induced issues, the motors and other components were equipped with grounding brushes, insulated bearings, etc. The remote possibility of currents going down the shaft, etc., was considered and much head scratching ensued. Finally found one of the old CSI 384SP shaft current probes, hooked it into a 2120, set up some points via documentation, and went hunting. The unit measures when properly set up, magnitude, time domain and frequency content of any currents thus measured. To my utter amazement, we clearly observed currents directly measuring on the shaft, and the currents frequency content was absolutely at the drives switching frequency, and magnitudes varied proportional to load.

Went to another mill with similar drives (different mfr) and found identical results although at much lower magnitudes.

One quirk that emerged was the most Rogowski coils do not the frequency range to address the hgher switching frequency drives.

Much more to the story, but yes it can be done, and at least in this case the results were unequivocal.

This message has been edited. Last edited by: Don Rainey,

I have measured the current to ground by attaching a brass bursh to a VOM wire and attaching the other wire to ground on VFD operated motors where some were have shaft current problems and others were not. On the VFD operated motors that were having problems, I measured up to 200 mA current to ground. On the VFD operated motors that were not having problems, I measured very small amounts of current to ground, but not nearly as high as the problem motors.

Regards,
John J

Did a recent case study into medium voltage motors having continuous electrical fluting of non drive end motor bearing x 40 or so motors!

EASA and NEMA both state 300 millivolts to ground or less is acceptable. Reasearch also found that less than 20 amps end to end is also a good rule of thumb.

One OEM also states 20 amps or less end to end current

Used a 3 ott welding cable 10' long, used a digital multimeter and Fluke 199C digital oscope to capture both voltage and current to ground and end to end. Your standard multimeter measure RMS values, due to the high frequency events like vfd's, rotor bars and stator slots, the oscope had a 200Mhz frequency range.

The fix was to insulate the non drive end housing with either and insulated bearing housing or an insulated bearing plus a shaft grounding brush on drive end of motor.

Hope this helps
Dave

DAve,

Is it really 20 amps end-to-end or should it be 20 m-amps? 20 amps seems like a lot of current.

Walt

Walt, it is 20 amps. I have a video showing us using the 3 ott welding cable from end to end showing arcing on the strands of wires, in this case there was close to 40 amps!
Dave