VFD Harmonics - Load Side

[cmahaffey]

All,

Due to previous bearing incidents on some rotating machinery we have installed, our customer has requested we monitor harmonics on the LOAD side of the VFD that feeds our equipment (in addition to temperature and vibration analysis which we already do). They have been completely convinced by a local electrician (albeit unjustified in my opinion) that the harmonics are the root cause of the previous bearing failure. I understand it is well documented that VFDs can certainly cause premature bearing failures, however all of our in house tests have indicated no issues with shaft currents. The customer, naturally, wants verfication that the harmonics are within 'specs'. My question is, does anyone know an acceptable level or any 'specification' for LOAD side THD? Anyone know a good way to measure THD on VFD? Are these requests even reasonable? I am no PQ power guy, but all the Dranetz meters, etc. I have seen are for monitoring 60Hz systems, naturally. Our VFD runs at approx 2.8kHZ under normal coniditions, which is going to certainly limit the number of ordered harmonics we can read seeing as most PQM meters have a finite sampling ability (about 1 microsecond for the Dranetz I have seen). If anyone has any comments or advice on this situation it would be greatly appreciated. Let me know if there are any missing links I need to fill in with the summary of the situation.

Colin

 

[Skogsgurra]

One should not use the term "harmonics" for the load side. The voltage is a PWM voltage that consists of switched DC. So, if you try to measure harmonics, you will get something like "hunred of percents".

The PWM often has a carrier (switching) frequency between a few kHz and up to 15 or 16 kHz. The square character of the switched voltage means that there are frequency components in the MHz range. It is these high-frequency components that cause your bearing currents. Shaft grounding or du/dt filters or common-mode filters (for larger macheines) usually helps. Sine filters are effective, but expensive.

This is work for a seasoned guy that knows about bearing currents. Hard to find.

Gunnar Englund

http://www.gke.org

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

 

[CJCPE]

A recent article on this difficult subject:

Dahl, D.; Sosnowski, D.; Schlegel, D.; Kerkman, R.J.; Pennings, M., "Gear up your bearings," Industry Applications Magazine, IEEE , vol.14, no.4, pp.45-53, July-Aug. 2008

First paragraph:

Recent experience suggests that steep wave fronts lead to motor bearing material erosion and early mechanical failure. Until recently, bearing failures from shaft currents were predominantly electromagnetically induced. However, we report three sources for bearing currents, and, under pulsewidth modulation (PWM) operation, these can be both electrostatically and electromagnetically induced. We identified each source using a field methodology, and the field measurements reviewed indicate that both electrostatic and electromagnetic fields are induced within the motor. We present data from two field sites, with measurements supporting the mitigation strategies that were used to resolve the application issues.

Bearing current mitigation strategies described: insulated bearing, shaft-grounding brush, CM choke, revised grounding configuration, insulated coupling between motor and encoder, ASD output load reactor, electrostatically shielded motor, conductive lubricants, ceramic bearing, and reduced ASD carrier setting. Several may be required.

 

[DickDV]

I would suggest a shaft grounding ring such as made by ElectroSTaticTechnologies of Mechanics Falls, Maine or Parker Corporation (the hydraulics and pneumatics people). Both are simple and cheap.

If that doesn't work, you can go on to other more sophisticated analysis.

 

[LionelHutz]

There will be a lot of harmonics, especially voltage harmonics. The current is filtered by the motor inductance.

Measuring them will be useless in determining if they caused bearing damage.

You need to actually measure the shaft voltage and look for signs of bearing discharge.