VFD's - Significance of switching frequency 2

[Hewlett]

I have read from other threads that lowering switching frequency will improve efficiency, and reduce bearing currents.
We have some houndred VFD powered motors 0-20kW, and it appears setting tend to be left at default. Probably installers wanted to minimize their work... My perception is that we are spending relatively much time on bearing replacements...
Do any of you have any pointers to more information/ case stories on what the magnitude of difference a change of switching frequency could have?

 

[DickDV]

The most common reason for using high switching frequencies is to reduce motor magnetic noise.

I always recommend using the lowest switching (carrier) frequency you can tolerate.

Sad to say, this will not necessarily solve your problems. Even with a carrier frequency of 1 or 2khz, the individual pulses have very steep rise and fall times which equate to very high frequency components. These will couple across the air gap and continue to cause bearing failures but at a somewhat slower rate due to the reduced number of pulses.

In my view, the best solution to bearing current failures is the little grounding rings made by ElectroStatic Technology of Mechanics Falls Maine (

http://www.est-aegis.com).

They are simple to mount and are a pretty reliable solution.

Bearing currents have complex and varied origins and a complete understanding of them is not mine to give. For that, I would suggest someone on this BBS like skogsgurra who seems to have studied this problem in depth.

 

[jraef]

I agree, the lower the carrier frequency the better, as long as you can tolerate the whining, but that alone is not a solution to bearing problems.

That Aegis bushing is the easiest to make fit into a motor, but an alternative is external shaft grounding brush assemblies, shuch as Sohre or Helwig supplies.

http://www.sohreturbo.com/

http://helwigcarbon.com/studies/study_bearing.aspx

JRaef.com
"Engineers like to solve problems. If there are no problems handily available, they will create their own problems." Scott Adams
For the best use of Eng-Tips, please click here -> faq731-376

 

[ScottyUK]

Hi jraef,

Do the Sohre shaft-riding bushes deal with VFD-induced shaft currents? I've looked at them a couple of times and will probably end up buying a couple to try as shaft grounding bushes on our turbo-generators, but not for that reason. On those machines the problem is charge buildup on the shaft due to friction (windage). It is effectively a 'static' charge. Eventually the charge builds up to a voltage high enough to punch throuch the bearing oil film and damage the white metal. Circulating shaft currents are dealt with differently by isolating the generator bearing pedestals from frame earth except at one location, thus opening the loop.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[Hewlett]

Thanks! I think the aegis rings are something we will start fitting. Then on the efficiency side.....any idea of how much difference in power consumption this could make? I would "guess" its a small number, but possibly significant with many motors...

 

[jraef]

ScottyUK,
To be honest, I have only used the Aegis units. I researched the shaft brushes for someone once and had that information available, but he went with the Aegis as well on my recommendation. I never considered that they may not work for inverter fed motors.

JRaef.com
"Engineers like to solve problems. If there are no problems handily available, they will create their own problems." Scott Adams
For the best use of Eng-Tips, please click here -> faq731-376

 

[CJCPE]

It is not unusual for VFDs to be derated at the higher carrier frequency settings. Some models may be rated for as little as 60% of their advertised output current rating when set to the maximum carrier frequency. You may need to study the manual very carefully to discover that. They should be capable of full output at the default carrier setting. If they are not derated in terms of output current, VFDs may be derated in terms of maximum ambient temperature or maximum altitude. VFDs that are available in NEMA 12 enclosures may be limited to lower carrier frequency settings. Sometimes only the “protected chassis” enclosure is suitable for the highest carrier frequency.

Whatever the manual and specifications say, you can be sure that the VFD will run hotter if the carrier frequency is increased. That temperature increase is due to increased losses and reduced efficiency. The increased temperature is also likely to reduce the reliability at least by a little. VFDs are usually advertised as 95% efficient or a higher, but the advertised efficiency is likely valid only for the lowest carrier frequency setting. If derating information is published for increased carrier frequency settings, you might be able to estimate how much the efficiency is reduced. I wouldn’t be surprised at an efficiency reduction of 1% or 2% for operation at twice the minimum carrier frequency.

 

[ozmosis]

The other impact of increased carrier frequency is the increase in High Frequency interference known also as Radio Frequency Interference (RFI) or simply put "noise". The higher energy required to switch your transistors a the faster rates means the impact of radiated interference will be greater and thus can spread further afield. So, greater precautions are necessary if the carrier frequency is increased.
The efficiency drop can be significant depending on the increase in CF and your existing ambient or calculated ambient. A large number of VFD's these days have internal protection that kicks in to automatically derate drive current if the CF is increased, this can catch you out sometimes. Others will protect based on heatsink temperature but heatsink temperature is often a general guide as it is transistor junction temperature that is critical for the reliability of a VFD.

 

[macmckim]

Came accross a new bearing isolator recently that has built in carbon brushes inside the seal. From what I could find out from the mfg, this shaft grounding seal attenates 98% of the shaft current and well offers IP66 protection against foreign elements. Have anyone tried this yet?

http://www.parker.com/packing/cat/english/5278.pdf

 

[benta]

Concerning efficiency vs. carrier frequency, I'd expect that a higher carrier frequency would increase iron losses in the motor, thus reducing motor efficiency.
Does anyone have information on this point?

Benta.

 

[CJCPE]

The motor impedance at harmonic frequencies is essentially the locked rotor impedance of the motor. Since that is primarily inductive, higher frequency harmonic voltages produce less harmonic current and thus lower losses.

There is a trade-off in inverter losses vs. motor losses such that higher carrier frequency settings produce lower motor losses at the expense of higher inverter losses. Up to a point, the combined inverter and motor efficiency is increased by increasing the carrier frequency. Above that point, the combined efficiency is lower.

 

[ozmosis]

benta
I'm not sure that is correct. The higher the carrier frequency the 'cleaner' the sinewave seen at the motor and therefore a slight improvement in efficiency. (but only slight). The comments made previously about increased motor noise when CF is reduced means the overall efficiency of the motor is reduced because you are losing efficiency in noise energy.
If you actually fitted a sinewave filter between motor and VFD then the motor will run quieter and slightly cooler due to the voltage and I sinewave produced. This doesn't mean the complete system is efficient as the filter drops the overall efficency by a fairly large amount (circa 10~15%)

 

[DickDV]

sed2dev, the comment about high CF making the motor quieter is a bit imprecise. The motor seems quieter but the noise actually has been shifted above the audible range.

Whether this energy is greater or lesser than before is not know to me but we are splitting hairs on this. Great for academic discourse but insignificant in the field, in my opinion.

 

[ozmosis]

Dickdv
Good point. I'm mixing up data and was thinking about high CF in combination with redesigning sinewave filters and the benefits this brings. Slight tangent alert going on with me.

 

[jraef]

To people thinking that a high CF makes the motor quieter, I respond "Not to your dog!"

JRaef.com
"Engineers like to solve problems. If there are no problems handily available, they will create their own problems." Scott Adams
For the best use of Eng-Tips, please click here -> faq731-376