Buzzing noise after VFD conversion

Line and load filters are not dV/dT filters and the primary purpose of a dV/dT filter is to tame voltage spikes associated with longer motor cable runs and reflections that occur in the wire. Line reactors help decrease the current draw into the VFD and decrease electrical noise, load help tame the heating effects and insulation damage that can occur. I prefer to use a DC choke as opposed to an input line reactor, these would have no effect on the high frequency whine. In addition the filters are dV/dT often designed not to be used with carrier frequencies above 8 kHz, and there are also limitations as to the variable frequency range that they can be used. The TCI output filters information is attached.

The KEB link clearly states "It should also be noted that audible noise varies between VFD manufacturers. So even though the switching frequencies are the same, one still might be audibly louder. For example, I have heard some competitor drives that are set to “16kHz” that still produce quite a bit of sound. I suspect there is more to the audible noise than just the rate of IGBT switching. I suspect the quality of the PWM signal and the driver circuitry also plays a big part. These details cannot be found in datasheets and are only found by testing side-by-side".
 

Attachments

  • TCI KLC VFD Output Filter Brochure.pdf
    2 MB · Views: 1
I can hear the whine on your 16kHz video, which makes me suspect it is not a 16kHz noise. I've been tested for hearing loss, and know I don't have good hear up at those frequencies. (And yes I do have tinnitus too). Be interesting to put an oscilloscope on there, especially one with some frequency analysis capability. But that type of equipment is not sitting around in your typical hobby shop. While carrier frequency noise is one component, a VFD uses some complicated control algorithms that could also have some periodic effects, possibly even creating beat frequencies that interact with the carrier.

Myself, I'd go with trying a different brand VFD. That noise would drive me nuts. Oh, wait, maybe it's too late already ...
 
@7milesup if running your current VFD at 80 Hz limits the noise, and your mill works fine, I say go for it. You can always do other remedial actions later.

@mksj I only used a line reactor at GE support insistence - and it did the right things for my motor noise at 10 KHz modulation freq. I agree that other solutions would seem more logical, but this is my experience. I have a 2HP pancake motor with a slightly larger inductance than my other 3PH 2HP motors, and this ?might? be a factor.

(Actually they called it a line filter, but the part they had me order called it a 'line reactor' - go figure....)
 
rabler said:
I believe technically a harmonic is a higher frequency component. Not sure what you'd call a lower frequency, a sub-harmonic? (lol). But that's what I was pointing toward with a "lower frequency resonance".
Click to expand...
Good point , maybe they would be a combination of multiple harmonics causing beat interference which may be at a lower frequency than the original ?

or it may be as you say a "sub harmonic" or undertone series. https://en.wikipedia.org/wiki/Undertone_series


Stu
 
Dabbler said:
(Actually they called it a line filter, but the part they had me order called it a 'line reactor' - go figure....)
Click to expand...
My take is those terms are 80% marketing (we need a new name to make our product sound special), and 20% technological. Reactance is the EE term for the capacitive/inductive component of impedance. dv/dt is from calculus, related to the mathematics used to analyze those components. So the engineers used capacitors and inductors, to make a specific filter, and the product marketing guys gave it a product line name. And the names get thrown around sort of generically and meanings don't rigidly hold true.
 
have you tried default settings other than current?

I’m wondering if the auto tune has enabled a wonky settin?

I have two VFDs, both used from eBay, both more or less silent. One is a brand name the other a cheap made in China, paid about $50 for each one. Both work about the same. I would try another VFD.
 
Maybe this will help with determining the frequency without a scope.
Test Tones: https://www.youtube.com/channel/UCgTxhjNGWS8EX-i9XMNKKQw/videos
How Old Are Your Ears:

Line reactor, AutomationDirect (I've used a lot of their stuff over the last 14 years or so without issue): https://tinyurl.com/25e52cox
Won't be in stock until 06/30/2022.

I would be surprised if a line reactor removed the mechanical noise. The other posters are correct in that long cables require line reactors on the motor side due to the large voltage spikes that can sometime occur and to prevent harmonics from being dumped onto the supply line when used on the drive input. Years ago, using F class or better field insulation on regular motors prevented voltage spikes from "poking" holes in the insulation. Of course they would still sing. DIdn't have PWM freqencies as high as they are today. Back in the 90's, AllenBradley Vector drives (new hotness better than plain old V/Hz) had a max of 8 Khz if I remember right. Newer VFD rated motors are both electrically and mechanically designed to be used with VFD's. On the other hand I you're going to use the mill 1 hour per week, who knows, the motor might outlast you if you can deal with the noise somehow.

Magnetostriction contributions due to motor not designed to be used with a VFD?
Maybe try a different motor. Here are some possibly helpful articles.

vfds.com

How to Stop VFDs From Ruining Your Motors

Common Causes and Solutions for VFD-related Electric Motor FailureElectric motor breakdowns are the source for plenty of business for service companies, but even they would never wish it on a customer. Motor experts would rather save you the expense and downtime with well-maintained motors that...
vfds.com vfds.com

Variable Frequency Drive Motor Mechanical Vibration

www.variablefrequencydrive.org
www.esmagazine.com

Motors For Use With VFDs

The questions can pile up quickly when it’s time to specify a motor to pair with a variable drive.
www.esmagazine.com www.esmagazine.com

Reactors for VFDs

www.electram.com www.electram.com

I'm 59 and I can't hear anything above 12 kHz. I can clearly hear the sound in all three of your videos.

Hope This Helps,
Jeff
 
Last edited:
Jeff.64 said:
Maybe this will help with determining the frequency without a scope.
Test Tones: https://www.youtube.com/channel/UCgTxhjNGWS8EX-i9XMNKKQw/videos
How Old Are Your Ears:

Line reactor, AutomationDirect (I've used a lot of their stuff over the last 14 years or so without issue): https://tinyurl.com/25e52cox
Won't be in stock until 06/30/2022.

I would be surprised if a line reactor removed the mechanical noise. The other posters are correct in that long cables require line reactors on the motor side due to the large voltage spikes that can sometime occur and to prevent harmonics from being dumped onto the supply line when used on the drive input. Years ago, using F class or better field insulation on regular motors prevented voltage spikes from "poking" holes in the insulation. Of course they would still sing. DIdn't have PWM freqencies as high as they are today. Back in the 90's, AllenBradley Vector drives (new hotness better than plain old V/Hz) had a max of 8 Khz if I remember right. Newer VFD rated motors are both electrically and mechanically designed to be used with VFD's. On the other hand I you're going to use the mill 1 hour per week, who knows, the motor might outlast you if you can deal with the noise somehow.

Magnetostriction contributions due to motor not designed to be used with a VFD?
Maybe try a different motor. Here are some possibly helpful articles.

vfds.com

How to Stop VFDs From Ruining Your Motors

Common Causes and Solutions for VFD-related Electric Motor FailureElectric motor breakdowns are the source for plenty of business for service companies, but even they would never wish it on a customer. Motor experts would rather save you the expense and downtime with well-maintained motors that...
vfds.com vfds.com

Variable Frequency Drive Motor Mechanical Vibration

www.variablefrequencydrive.org
www.esmagazine.com

Motors For Use With VFDs

The questions can pile up quickly when it’s time to specify a motor to pair with a variable drive.
www.esmagazine.com www.esmagazine.com

Reactors for VFDs

www.electram.com www.electram.com

I'm 59 and I can't hear anything above 12 kHz. I can clearly hear the sound in all three of your videos.

Hope This Helps,
Jeff
Click to expand...


If you need quite good audio analyses tool for free (also sound editing) I highly recommend the open source "Audacity".
Cross platform too (windows / mac / linux) , https://www.audacityteam.org/

Here are two screen shoots , one of a spectrogram (waterfall) plot of the sound file and the other a frequency plot of a small time segment.

You can see the energy at different frequencies quite easily on the spectogram type plot very intuitively, the frequency plot shows a big spike at about 12.5khz but other strong frequencies are a little harder to spot I would say with this type of plot.

Interestingly the audio file contains nothing above 16khz , a lot of microphones produce hideous sound above this frequency also most people can't hear any of it and it saves the compression algorithm from wasting bandwidth on it. So best to just filter it out with a low pass filter (most likely reason it's missing).

Stu

waterfal-vfd.jpgfrequency plot mill.jpg
 
Jeff.64 said:
Maybe this will help with determining the frequency without a scope.
Test Tones: https://www.youtube.com/channel/UCgTxhjNGWS8EX-i9XMNKKQw/videos
How Old Are Your Ears:

Line reactor, AutomationDirect (I've used a lot of their stuff over the last 14 years or so without issue): https://tinyurl.com/25e52cox
Won't be in stock until 06/30/2022.

I would be surprised if a line reactor removed the mechanical noise. The other posters are correct in that long cables require line reactors on the motor side due to the large voltage spikes that can sometime occur and to prevent harmonics from being dumped onto the supply line when used on the drive input. Years ago, using F class or better field insulation on regular motors prevented voltage spikes from "poking" holes in the insulation. Of course they would still sing. DIdn't have PWM freqencies as high as they are today. Back in the 90's, AllenBradley Vector drives (new hotness better than plain old V/Hz) had a max of 8 Khz if I remember right. Newer VFD rated motors are both electrically and mechanically designed to be used with VFD's. On the other hand I you're going to use the mill 1 hour per week, who knows, the motor might outlast you if you can deal with the noise somehow.

Magnetostriction contributions due to motor not designed to be used with a VFD?
Maybe try a different motor. Here are some possibly helpful articles.

vfds.com

How to Stop VFDs From Ruining Your Motors

Common Causes and Solutions for VFD-related Electric Motor FailureElectric motor breakdowns are the source for plenty of business for service companies, but even they would never wish it on a customer. Motor experts would rather save you the expense and downtime with well-maintained motors that...
vfds.com vfds.com

Variable Frequency Drive Motor Mechanical Vibration

www.variablefrequencydrive.org
www.esmagazine.com

Motors For Use With VFDs

The questions can pile up quickly when it’s time to specify a motor to pair with a variable drive.
www.esmagazine.com www.esmagazine.com

Reactors for VFDs

www.electram.com www.electram.com

I'm 59 and I can't hear anything above 12 kHz. I can clearly hear the sound in all three of your videos.

Hope This Helps,
Jeff
Click to expand...

interestingly , I notice on the sound tone video their is also nothing above 16khz so I think it's youtube doing a low pass filter.

Hearing Test Audio.jpg

Stu

edit , you can generate tones with audacity too.
 
You must log in or register to reply here.
Back
Top