CPS Elec.
Member
- Location
- Bemidji, MN
How do you know if a motor is compatible with a VFD?
VFD's
- Thread starter CPS Elec.
- Start date
- Status
petersonra
Senior Member
- Location
- Northern illinois
- Occupation
- engineer
these days -
most single phase motors are not.
most decent quality 3 phase motors are.
check the manufacturer's data sheet and see if it says invertor rated, or something similar.
most single phase motors are not.
most decent quality 3 phase motors are.
check the manufacturer's data sheet and see if it says invertor rated, or something similar.
CPS Elec.
Member
- Location
- Bemidji, MN
The motors in question are up to 40 years old and I've had a hard time locating any data sheets, is there any other way, possibly something stamped on the nameplate?
If they are 40 years old the probability is that they were not made to work with VFDs. A common problem, and one that I've met a few times, is the very fast switching edges (the high dv/dt) that modern VFDs present to the motor. Winding insulation failure can, in my experience, occur in a matter of a few weeks. There are ways of getting round this. Reactors or filters between the VFD and the motor can mitigate this problem.
Another issue that I've read about, but not come across*, is motor bearing failure. Insulated bearings at one end is a solution but a motor 40 years old is pretty unlikely to have that.
The motor nameplates are unlikely to address either of those issues. But they are matters that I've had to deal with head on both at the bid stage of a project and in the case of insulation failure, after the equipment was put into service. A couple of publications that I found useful were from Gambica. They were not free but relatively inexpensive - around ?10 each as I recall. My QA guy got them for me. They may be available on line.
*How we cable motors here in UK differs from that in the US which is maybe why we don't see the bearing failure issue.
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
Here's the "Reader's Digest" version of a paper I wrote on the subject.
There are 3 basic issues of major concern with using motors behind VFDs, all of them subject to a myriad of application specifics;
Cooling issues are because when you use a VFD, you are by definition going to be slowing the motor down, which means you are also slowing down cooling fans that are attached to the motor (if any). Even though the motor power is going down at the same basic rate that the cooling efficiency is going down, some of the losses in the motor remain the same, so you can get in trouble at low speeds. This issue can be overcome by using separately powered fixed speed cooling fans, a.k.a. using a "blown" motor. If you have an old motor, you may have to fabricate your own cooling plenum to hold a fan. It's not difficult, but it does require some sheet metal talent to avoid looking "funky".
Winding insulation breakdown occurs as a result of what is called "standing wave generation" in the motor leads as a result of the power coming from a PWM inverter. Without going too deep into PWM theory, a VFD does it's magic by sending out very fast pulses of DC that act like AC power to a motor. In that process, the pulses are created by firing the DC at very fast switching rates using high speed transistors. The steep rate of rise of those pulses has an effect in the output conductors that makes them act kind of like capacitors and they create waves of energy that discharge into the motor as pulses. The longer the motor leads, the more these waves can build up and in extreme cases, the pulses can exceed 3 times the line voltage. In the olden days, it was considered prudent design practice for the motor winding insulation to be 2-2.5 times the line voltage. So for a 480V motor, the insulation was often 1200V. But now with a VFD, it's very common to see those pulses get to 2000V! What happens then is that these pulses, albeit very short, find the weakest point in the insulation and punch through it, causing turn-to-turn shorts. Modern "inverter duty" motors then have new insulation materials that can withstand these pulses. On old motors though, you can mitigate the damage by adding filters to the output of the VFD, or use motor terminators at the motor end that are designed to stop the pulses there. But you didn't say what the voltage was. If it is a 230V motor, this issue isn't as bad, because many motor mfrs only have one insulation rating for their magnet wire, so 1200V insulation on 230V systems is way more than 3X the line voltage, so the insulation is more likely to be able to withstand the onslaught. Still, YOUR insulation is now 40 years old, so even if it is 230V, I would take precautions.
Bearing damage is something that is just showing up more recently (10 years or so) and is related to similar issues as the standing waves. But in essence, a static charge builds up in the rotor as a result of the way the PWM power is applied to the stator. This static charge also builds up until it is high enough to overcome the insulating properties of the bearing grease, then it discharges through the bearings. As it does, it has a microscopic effect similar to EDM (Electrical Discharge Machining) and it carves away at the bearing races, causing a washboard like pattern to emerge, called "fluting". This eventually gets so bad that it destroys the bearings. Again, modern inverter duty motors now often come with a grounding bushing at one end so that the charge has a safer path to ground than the bearings and races. There are things you can do to avoid and prevent this in older motors as well, but it is often very expensive. But one way to help it is to replace the regular conductors with shielded conductors, often now called "VFD cables", which is what Besoeker is referring to when he says they don't see this issue much in other parts of the world because that is their standard way of wiring up inverter motors now.
So should you change the motor? If it were me, I would not, at least not yet. Get a good filter for the output of the drive (not just a load reactor), assess the cooling situation based on how slow you intend to run it and use it until it dies. THEN replace it with an inverter duty motor. Having the filter will be good insurance for the new motor as well.
There are 3 basic issues of major concern with using motors behind VFDs, all of them subject to a myriad of application specifics;
- Cooling
- Winding insulation breakdown
- Bearing damage
Cooling issues are because when you use a VFD, you are by definition going to be slowing the motor down, which means you are also slowing down cooling fans that are attached to the motor (if any). Even though the motor power is going down at the same basic rate that the cooling efficiency is going down, some of the losses in the motor remain the same, so you can get in trouble at low speeds. This issue can be overcome by using separately powered fixed speed cooling fans, a.k.a. using a "blown" motor. If you have an old motor, you may have to fabricate your own cooling plenum to hold a fan. It's not difficult, but it does require some sheet metal talent to avoid looking "funky".
Winding insulation breakdown occurs as a result of what is called "standing wave generation" in the motor leads as a result of the power coming from a PWM inverter. Without going too deep into PWM theory, a VFD does it's magic by sending out very fast pulses of DC that act like AC power to a motor. In that process, the pulses are created by firing the DC at very fast switching rates using high speed transistors. The steep rate of rise of those pulses has an effect in the output conductors that makes them act kind of like capacitors and they create waves of energy that discharge into the motor as pulses. The longer the motor leads, the more these waves can build up and in extreme cases, the pulses can exceed 3 times the line voltage. In the olden days, it was considered prudent design practice for the motor winding insulation to be 2-2.5 times the line voltage. So for a 480V motor, the insulation was often 1200V. But now with a VFD, it's very common to see those pulses get to 2000V! What happens then is that these pulses, albeit very short, find the weakest point in the insulation and punch through it, causing turn-to-turn shorts. Modern "inverter duty" motors then have new insulation materials that can withstand these pulses. On old motors though, you can mitigate the damage by adding filters to the output of the VFD, or use motor terminators at the motor end that are designed to stop the pulses there. But you didn't say what the voltage was. If it is a 230V motor, this issue isn't as bad, because many motor mfrs only have one insulation rating for their magnet wire, so 1200V insulation on 230V systems is way more than 3X the line voltage, so the insulation is more likely to be able to withstand the onslaught. Still, YOUR insulation is now 40 years old, so even if it is 230V, I would take precautions.
Bearing damage is something that is just showing up more recently (10 years or so) and is related to similar issues as the standing waves. But in essence, a static charge builds up in the rotor as a result of the way the PWM power is applied to the stator. This static charge also builds up until it is high enough to overcome the insulating properties of the bearing grease, then it discharges through the bearings. As it does, it has a microscopic effect similar to EDM (Electrical Discharge Machining) and it carves away at the bearing races, causing a washboard like pattern to emerge, called "fluting". This eventually gets so bad that it destroys the bearings. Again, modern inverter duty motors now often come with a grounding bushing at one end so that the charge has a safer path to ground than the bearings and races. There are things you can do to avoid and prevent this in older motors as well, but it is often very expensive. But one way to help it is to replace the regular conductors with shielded conductors, often now called "VFD cables", which is what Besoeker is referring to when he says they don't see this issue much in other parts of the world because that is their standard way of wiring up inverter motors now.
So should you change the motor? If it were me, I would not, at least not yet. Get a good filter for the output of the drive (not just a load reactor), assess the cooling situation based on how slow you intend to run it and use it until it dies. THEN replace it with an inverter duty motor. Having the filter will be good insurance for the new motor as well.
weressl
Esteemed Member
Bearing damage is due to magnetic imbalance in the stator and rotor and the resultant voltage buildup is much greater at higher than synchronous - 2-8kHz carrier frequency - and non-sinusoidal supply. The leading switching edge is not helping this either. It can be easily remedied by discharge brushes on one end. The voltage buildup is not a static charge though and it is due to the bearing lubrication forming the dielectric, insulating film on the race surface, pretty much identically to the EDM process.
Interestingly enough older motors were more balanced magnetically than motors produced today to high efficiency standards with random winding and and tightly packed stator slots.
Interestingly enough older motors were more balanced magnetically than motors produced today to high efficiency standards with random winding and and tightly packed stator slots.
These days some motors have an insulated bearing at one end to prevent circulating currents due to asymmetry.
This is an excerpt from a specification of one we installed fairly recently:
Extras Included
45X Heaters (voltage to be confirmed) terminated in the main terminal box
Separate cooling
1024 pulse tacho encoder
Insulated bearing
Routine test report
It's from one of the mainstream motor manufacturers.
weressl
Esteemed Member
It has been available in the past 10 years or more, primarily on MV and larger frame LV motors. I was refurbishing an existing 800HP motor with an ASD and it was suggested by the OEM of the motor. Instead I've installed a grounding 'brush' - a beryllium/copper leaf riding against the shaft and it has been dancing eversince.
The addition of the isolated bearing was discussed in the IEEE541 meetings, I don't know if it ever made it in, I'll check when I am in the office.
I can see that as being the easier option on a refurb. The one we did recently was a new installation and, at 180kW, not very big. Insulated bearings have been around here for quite a while too. I have a Gambica technical guide on bearing currents from some years ago and it mentions insulated bearings. But I don't know how many of VFDs we've supplied have them fitted to the motor - mostly we don't supply the motors
Appreciated.
petersonra
Senior Member
- Location
- Northern illinois
- Occupation
- engineer
I must have been lucky over the years. Never ran into the bearing problem. have heard about it, but never ran into a case where it actually happened.
weressl
Esteemed Member
Nor have I but this an example of what it looks like:
Have heard about it and then run into one on an unexpected size. A 10HP motor failed this way.
The tale always was that it happens on large motors, several hundred HP or larger. Maybe because only on those sizes would you spend the money AND have the available room in the first place, to install bearing isolator sleeve.
BTW I thought I had a copy of the IEEE541, alas I only have the 841.
"[FONT=Arial,Bold][FONT=Arial,Bold] 5.3.2 Adjustable speed drive (ASD) Supply
....
i) Shaft-to-bearing voltages and/or currents resulting from common mode currents flowing throughstray system capacitances to ground via the bearings. These currents are induced from the ASD’s
high rate of change ([FONT=TimesNewRoman,Italic][FONT=TimesNewRoman,Italic]dv[/FONT][/FONT]/[FONT=TimesNewRoman,Italic][FONT=TimesNewRoman,Italic]dt[/FONT][/FONT]) of output voltage"[/FONT][/FONT]
Last edited:
I've come across a few motors with dv/dt failures. A bunch were 800kW on the same site in a matter of weeks. As one guy put it, 800kW motors are not normally considered consumables.
- Status