PWM Freq setting in VFD drives

TheWaterboy

Lifetime Supporting Member + Moderator
Join Date
May 2006
Location
-27.9679796,153.419016
Posts
1,906
I just replaced a PowerFlex4 that was showing a F012 error on a good motor. The replacement drive worked fine but I wanted to adjust the PWM setting to make the drive less noisy as long as there is no detriment to the motor.

The default freq is 4hkz and is quite audible and in a range that is objectionable in the closed room it sits in. Putting it at 15khz puts it out of my old hearing range but its probably driving some dogs in the neighborhood nuts.

Whats the criteria for changing the PWM freq? Certainly there is a better reason to change it than just auditory concerns.
 
I have installed many VFD's on all kinds of motors, even on wound rotor motors.
I have had some do what I call sing that usually related to the VFD output frequency and load and I never had anybody complain about before. Usually I'll hear it when they don't even notice it.
The only time I have had to adjust the PWM frequency was a case where there was a ultrasonic flow sensor in close proximity to the motor when the motor ran it would play havoc with the meter readings. I changer the PWM freq to I think 10Khz problem solved.
The F012 Fault looks like a hardware overload fault (300%) that would be instantaneous current.
A question did you get the same fault with the motor disconnected?
If you did then the VFD is faulted.
If not it still could be in the motor.
If you are hearing that much noise I would be thinking of a lose laminate in the motor someplace vibrating. If that is the case then the motor will fail soon, better start to think about getting a replacement on hand. I have seen bad motor run a long time on a VFD, most motors fail on the startup The across the line starters much quicker due to the high inrush current on startup.
Is this a problem that just came up after running a while, or was it there from the original system startup and thy just now complained about it.
That would also lead to where to look. If it just came up now then something changed. Changing the PWM freq. may only cover a problem will come back to bite you.
 
Increasing the PWM (carrier) frequency dramatically increases the switching losses inside of the drive. You must de-rate the drive and if you turn it up to 15kHz, the de-rate is to 58% of it's stated rating. So that means if you have a 6A rated drive, at 15kHz it is only good for 3.5A , or in other words a 3HP 480V drive becomes good for only a 1-1/2 HP motor. That might be why your drive was faulting on F012.

Increasing it also INCREASES the potential negative side effects of cable capacitance and motor rotor-frame capacitance. Cable capacitance issues leads to standing wave / reflected wave propagation that can lead to motor insulation damage. the rotor-frame issue leads to EDM damage of bearings and races as the capacitive charges build up in the rotor and discharge through the bearings to the frame.

There are mitigation techniques that can address both problems, but higher carrier frequencies exacerbate them. So for example if your PF4 was OK for a motor distance of 40ft before a load reactor is recommended, that may be cut down to less than 10ft if the carrier frequency is that high.
 
We had some new air scrubbers installed using PF753 drives. These were noticeably louder than our older equipment. After a call to AB tech support, they suggested moving the carrier frequency up 2Khz. This solved the noise issue, could barely hear them running afterwards. They did tell me that any more than 2Khz, would need to start de-rating the drive as jraef stated.
 
Thanks for the feedback guys. The drive just faulted with Code F064. However I did inadvertently use a smaller drive (why DO WE HAVE THOSE ?!) that was on the shelf. It's within spec but barely. So I increase the carrier and have to derate the drive... that fits what just happened. I have the larger drive also so I will deal with that tomorrow.
The motor is 3 feet away and 1.7 amps so maybe there's a small reactor that can mitigate the noise.
 
As a general rule, I find no value to increasing the carrier or PWM frequency above 8khz. Pretty much all the motor quieting and shaft rotation smoothing at very low frequency is in at 8khz.

As another general rule, I always recommend using the lowest carrier frequency your application can tolerate. It's easier on the motor and may, in some brands, avoid the derate on the output current.
 
On this motor I was nearly able to play music on it! Every freq had a tone to it, for me the higher the better, but alas, the drive did not want to play that game. I backed it down to 6.3 khz and its happy and doesn't sound too bad.
I do find it interesting that the current that the drive reports it is providing doesn't change very much with frequency, I would have thought it would.
 
The current probably doesn't change because the motor leads are so short. If they had been a couple hundred feet long, the current likely would have increased a bit with the higher frequencies. The extra current didn't make it to the motor. It was capacitively shunted to ground along the leads.
 
...The extra current didn't make it to the motor. It was capacitively shunted to ground along the leads.
While realizing this is academic at this point I'm still interested... regardless where the current is lost, shouldn't the drive have shown an increase since it is the source ? How else are we to see that the increased carrier is doing more harm than good?
 
The display would show the increased current if it was actually happening. In your case, the motor leads are too short for any increase to be measureable

Within reasonable limits, say 1k to 8k, higher carrier frequency isn't necessarily bad, especially if you have an inverter-duty MG1-P31 motor. In my view, frequencies over 8khz cause more problems and give you nothing in return
 
While realizing this is academic at this point I'm still interested... regardless where the current is lost, shouldn't the drive have shown an increase since it is the source ? How else are we to see that the increased carrier is doing more harm than good?
The display is only showing the output current. As Dick said, with short motor leads the capacitive charging current losses to ground would be almost undetectable. But if the leads were long enough, you might see it in the display. However if you are increasing the switching losses inside of the drive by using a high CF, those losses would only show up in looking at the input current, which the drive is not monitoring. You would need to look at that with another meter. It would need to be an expensive one too, because the harmonics would be significant.
 

Similar Topics

I have a project to control the speed of motor DC using PWM Output on PLC and when im on working i have a several trouble and of of them is the...
Replies
6
Views
168
Hi All, I came up with this for ramping the PWM pin on the micro 820. I'm new to ladder logic so I am not sure if this will even work? Once...
Replies
6
Views
324
Hi, I am trying to have my PLC start a heating fan in a linear manner. A certain percentage at a certain received setpoint. To do this, I use a...
Replies
8
Views
1,521
Hi guys, i have never had to do a PWM Output or input before i have a Device that to dimm the lamp output from the ballast, the ballast has a 24V...
Replies
6
Views
937
Hello Friends, I am looking for a way to take my PID Output( it is the result of my PI Algorithm in Real format) and generate a PWM Signal in...
Replies
16
Views
2,330
Back
Top Bottom