VFD Resonance

Hello all,


we're using a Siemens G120 VFD to control a 3hp inverter rated motor for a small project. To make sure everything worked, we bought about 10ft of Southwire unshielded 10AWG cord to make initial connections. Everything worked fine. Motor is 3-phase running at 208V/7.9A. No choke or filters.



Ordered some 14AWG VFD cable and hooked it up last night. Ran the motor and noticed that the switching noise at around switching frequency (high pitch varies with speed) is now so loud that I almost need to wear earplugs standing next to the motor. Pushing on the fan shroud quiets it down to about half, so it seems to be resonating from there. All of the noise is coming from the motor itself and not the drive. The cable length is only 20ft.



Looking at the drive data sheet, it says maximum shielded cable length of 165ft. Initially I didn't have the shield connected at the drive end, only the motor end (just wanted to do a quick test), but grounding the shield doesn't seem to make a difference. Another thing specified in the Drive datasheet is a conductor cross section of 10-6AWG. However, the drive is rated for 5hp, the motor draws far less than this so didn't see this being an issue.


I'm a bit afraid to run the motor like this as my gut says damage will eventually ensue. I assume some type of electrical resonance here. Anyone have any ideas what's going on and how to mitigate it?


Thanks,

Try changing pulse frequency. Factory setting 4K? Try adjusting by 1K increments.
P1800 I think.

Was the motor mounted when you tested it the first time?

No, the motor was never mounted, just sitting on the bench in both cases.



I tried moving the motor location back and forth a few inches on the bench just to see if made a difference, it didn't.

Aw motor sing I have heard many motor on a vfd sing
It would be common for me to walk into a plant and here more than a few motors singing
I take it from your description that you have not run the motor up to rated speed or anywhere near rated load.
Usually that comes from loose laminates in the motor I have found that is more noticeable at mid speed and lower load as the motor gets up to speed and load they quite down.
The newer designed vfd’s tend to do it less, it think it’s a different switching technology
You stated that if you put your hand on the fan shroud the sound goes down that would tell me that It is defiantly vibration in the motor. You could just have a bad motor. A better quality motor would be better or VPI the motor would really help that. If it is too bad and allowed to continue it will rub through the motor windings and short out the motor.
Before I would make any changes I would see how it is under normal running conditions.

I have had some motors that when I place the order I had the motor shop VPI them, replace the bearing with precisions bearings and balanced for high speed operations . depending on the application Imay have the motor shop laser alien motor on site the extra cost is worth it for reliability I have never regretted it.

Did you mess with any other drive parameters? I would make sure you are in v/Hz mode to start. Running in sensorless vector with the wrong motor data can cause noise, vibration and lack of torque.

I take it from your description that you have not run the motor up to rated speed or anywhere near rated load.

Click to expand...

I ran it through the full range of speeds, but correct, no load.
Just found it strange that it wasn't nearly as loud with the larger unshielded cable.

Did you mess with any other drive parameters?

Click to expand...

I didn't mess with any drive parameters, but I did do a factory reset and re-run the motor identification routine to see if that would help. It's currently in sensorless vector, I can try V/Hz and see if that improves things.

Did you run an Auto-Tune?
For 22 years, at my previous company, I had never ran an Auto-Tune. Now at, my current company, it is standard practice.
I have noticed a difference, in this quieter environment, of the sound of a motor after an Auto-Tune.

@GaryS - VIP (Vacuum Pressure Impregnation), never heard of it until today. Thank you.

Jieve said:

I didn't mess with any drive parameters, but I did do a factory reset and re-run the motor identification routine to see if that would help. It's currently in sensorless vector, I can try V/Hz and see if that improves things.

Click to expand...

You can get better performance with sensorless vector but V/Hz is simpler and doesn't rely as much on having the motor parameters right.

Is the motor new or old?
If new, is it from a reputable brand? (we have had issues with off brand motor quality being sub-par)
Is the motor a standard Design B motor or something special?

Keep in mind, guys, that it isn't so much that it sings. It is that with 10 AWG unshielded SO cord it doesn't make a sound but with 14 AWG shielded VFD cable its the Mormon Tabernacle Choir. Same motor, same drive, different conductors. I'm at a loss on this one myself.

If it were me I would run an auto-tune with the new cable (if you haven't already) just to see if it makes a difference. Other than that I'm not sure.

Keith

Jieve, what is your PWM frequency? Have you tried going up or down? It can affect the ringing...
Maybe you could just follow the manual and use 10 ga...

Agree with cidchace. Lowering the switching frequency may help in this.

Actually it would be raise the switching frequency. I know Yaskawa drives you can change switching freq. If I remember correctly they come from factory set at 15KHz. Used to be drives when first came out had switching freq at like 1Khz when put in office on air handlers you could hear the noise. The higher freq. is not as easily heard at 15KHz by human ear. We had about 1000 VFD's in a process plant most of them were left at factory setting of 15KHz. This setting also affected the way the overload and over current was monitored.

The noise is from the laments in motor. If you get a VFD rated motor this will be less of a problem.

Thanks everyone for the helpful responses. Just revisited this; double-checked all connections, made sure everything is correct.


The motor is from North American Electric, Inc. It's a premium efficiency inverter duty rated motor. It's standard design, and brand new.



I also re-ran the auto-tune routine (this is what I meant by motor identification) above after switching cables, as mentioned no real difference.


I checked all parameters, one thing interesting is that I have the drive set to 240V input voltage (single phase), and the output set to 208V 3-phase, but when running at full speed (3600rpm) with no load, the output voltage on the display reads 240V.


Next, changed the PWM frequency. This made a HUGE difference. Default is 4kHz, changing up reduced the noise, with 8kHz probably being the most quiet and 12kHz almost tied for 2nd place. Changing down to 2kHz the noise was horrible and even louder.



I haven't tried setting to V/Hz, still on sensorless vector.



Other than audible noise, what are the effects of changing the PWM frequency on the longevity and performance of the motor, and how do you choose the ideal frequency? Just because I can't hear the noise, for example, doesn't mean it isn't doing some damage of some kind.

A few things on this
How did you measure the vfd output voltage my guess is a standard AC millimeter
If you did, don’t even concern yourself with the output voltage reading, only a few very high end volt meters can accurately read the output voltage of a vfd. Most meters are designed to read a sinusoidal
wave but the wave form on the output of a vfd is not sinusoidal, so go by the displayed voltage on vfd it will be the most accurate. You stated it read 240V that would be the output a 60 Hz and above.
Is this a single phase input vfd? 240V 1Ph input or 240V 3Ph input just running on single phase.
It does make a big difference while some 3 ph vfd’s will run ok on singe phase ( Just reduced load output) others will not it is important to have all the information to select the proper combination.
How did you measure the motor speed with a tachometer or are you just looking at the vfd output frequency and going by the motor name plate data you have to be careful with this.
You said the max speed is 3600 rpm’s that would be a 2 pole 60Hz motor for use in North America.
It may not seem important but most vfd’s need to know the number of motor poles to run properly all vector drives require it. I had one some time back I couldn’t get to run right until I realized it was 2 pole motor and the vfd default was set for 4 pole I corrected it and the problem went away. They use the counter EMF from the motor to calculate the motor speed in open loop vector mode. They also use that information to control the output transistors with an encoder. A standard squirrel cage motor will never run quite at sync speed. They will always slip back a few rpm’s even so the motor will run at the motor rated speed. Typically a 2 pole motor will have a name plate rating at about 3450 rpm’s or so and an 4 pole motor will have a rating of 11750 rpm’s or so.
It may sound trivial but it is important to fully understand what’s going on.

As for increasing the switching frequency if it solve the noise problem, great but I really don’t think you solved the problem you moved the frequency up to an audible range we can’t hear but the vibration is still there. I have only had to change the switching frequency once. It was in sewage plane and every time the motor ran the interments around it would go crazy moved it up a little and everything worked fine. I have to wonder if maybe you a bad output transistor in the vfd on one leg is not turning on correctly the only way you may be able to see it would be with a scope or if it’s practical try the motor on another vfd and see what happens Either way you need to make sure that data in the vfs matches the motor nameplate data.
I really don’t think that motor sing will give you problems down the road but it is hard to say unless you can identify exactly where it’s coming from. Most motors on a vfd sing some. I know the frequency increase with the motor speed / vfd output frequency.