How do you troubleshoot common VFD faults related to carrier frequency and modulation index?

Did I actually make the problem go away? Not really b/c the antenna is still there operating and causing the same problem to this day, but it’s just outside the human audible range radiating at a higher frequency then we can hear. So in other words don’t bring your dog near or inside that office building for fear of it going ballistic and you will have no idea why unless you read my rambling here.

At any rate we started dialing up that carrier frequency parameter. I kept telling my technicians out there only go as high as you have to (again b/c it’s harder on that motor and drive). I think we ended up setting it clear up to 12 KHZ and voilà the sound was amazingly gone. I finally could breathe a sigh of relief and go home and change my pants. I not only made the sale but I saved the day making it work in the field. Or wait did I?

I don’t think it was me but someone much smarter than me may have recommended to try adjusting the carrier frequency. I would leave this parameter at 6KHZ as that was as low as this particular drive could go and lower is better as it protects or minimizes the motor and VFD heat generation as the statements above elude to. I believe some drives can go as low as 2 KHZ. The downfall of this is lower control resolution but in the HVAC world we don’t care b/c our motors are controlling fans and pumps that don’t require precise control. In the manufacturing or process industries that may not be the case.

I will eventually get to my point here but imagine a large radiating antenna. That’s what the 3 phase output wires turn into on the output of that VFD. So the longer distance your VFD is away from the motor the bigger and noisier your antenna is getting as your wire leads increase in length. I may be wrong here but I believe this is what they call EMI generation or electro magnetic interference. The EMI we were creating with the VFDs was radiating and buzzing electrical noise that we could physically hear through the light ballasts down in the office space.

If you were to look at the 3 phase output of the leads going out to the motor from the VFD on an oscilloscope you could conclude that electrically speaking the output is very dirty. It’s not very clean power feeding that motor. Those 6 little IGBTs (inner gate bipolar transistors) on a 6 pulse drive (2 on each phase, 1 positive, 1 negative) are firing that motor with 650 volts of DC (for a 460v AC drive) so fast (fractions of a second) that it creates what I would call a violent electrical event that you can physically see on that oscilloscope screen.

In such an instance, I would contact the manufacturer of the VFD and get their input. They (should) know their product inside out and be able to provide the absolute best troubleshooting advice.

As much as it is unfortunate to wait and see, a track record is a record. If there are sporadic component failure or minute undulation with a device that utilizes dc voltage and the regulatory of application. For example testing consistent starting torque application of an electric motor. Benchmark is manufacture benchmark, remainder is bench and field tests through varying means. Another I find is rectifiers, if a unit fails then check the controlling and supplying controllers. If there is a problem with the source it will effect everything downstream, the trick is translating the results.

Usually there is Modulation index limit a parameter setting inside the VFD, this limit should be site to motor rated voltage or slightly higher than the rated voltage if compensation or increased motor voltage is required under overload conditions. In case of three level NPC configuration some margin is reserved for neutral balance regulator to prevent unbalanced between positive and negative buses. Running at the Modulation limit. Limit this regulator and over voltage is hardly avoidable.