Can anybody explain me in laymen terms, why a special cable is required from the output of the VFD to the motor? What happens if I don't apply a special cable? From my reading from web, I find that if I don't apply a special cable, Motor insulation damage occurs from high voltage. I do not know how the high voltage occurs. On further google search I find the terms dv/dt, high frequency, reflected wave etc. which are beyond me
VFD Cables
- Thread starter Charz
- Start date
- Status
We didn't - we used steel wired armour. But that was for us Brits. You may do something different..
KDough
Member
- Location
- San Jose, CA
- Occupation
- Electrician 42 yrs / Electrical Inspector 4yrs
VFD stands for variable frequency drive. With VFD’s I believe the voltage remains the same but the VFD varies the frequency to control the speed of the motor. I have never heard of having to use a special cable to the motor. I‘ve always used regular old building wire , thhn, thwn. Is this for a special type of motor with a VFD that is made specifically for that motor? Remember you can never go wrong by following the manufacturers installation instructions. Even the AHJ would be hard pressed to tell you you had to do something different.
drktmplr12
Senior Member
- Location
- South Florida
- Occupation
- Electrical Engineer
it isn't required. A properly connected aluminum conduit will work too. the damage to motors is due to harmonics, which are created by the VFD quickly opening and closing the rectifiers. the harmonics are trying to get back to the VFD inverter that created them. if they don't have an easy path, the voltages will remain elevated and break down conductor insulation. the VFD cable is an unobstructed electrical path back for harmonics to return to the source (the inverter) so the harmonics do not find a different path (they will).
a little more technical point of view:
VFD's use inverters to recreate a sine wave to manipulate and achieve speed control of the motor. to do that, they first charge a set of DC capacitors. then the DC current is discharged by using rectifiers (basically electrical gates that are open and shut very quickly at the carrier frequency) are operated to maintain a voltage to hertz ratio. this create harmonics that are injected into both the power supply and motor leads.
the cables help because they help manage the harmonics and leakage capacitance that occurs between all cables in proximity to each other. in most cases we can ignore this leakage for plant environments provided we provide separation between very low voltage (think 24VDC and 4-20 mA signal) and anything power related (120V 1-phase, 480, 4160 volt, etc). think of the motor leads as a large capacitor. the longer the run, the more conductor you have in parallel, the higher the capacitance and therefore voltage that can exist between them. harmonics make this worse and the elevated voltages will cause sparks across the motor bearings and shaft interface, causing fluting and pitting of the bearings, eventually leading to catastrophic failure.
Always use XHHW MINIMUM for even small VFD's because it is less susceptible to insulation breakdown than THWN/THHN. why?
THWN/THHN is a thermoplastic process where it is melted onto the conductor. it is a less than perfect process, leading to microscopic pinholes in the insulation. Not normally a problem for regular power and control. however, for VFD, these little holes will aid in the breakdown of insulation when exposed to higher voltages present in VFD outputs. cross-link polyethylene (XHHW) is an extruding process and has no pin holes and is therefore more resistant to insulation breakdown.
always use a DV/dt output filter for larger motors (50HP+), unless there is a good reason not to.
always use a sine wave output filter for larger motors far away from the VFD (~700 ft, depends on VFD manufacturer).
a little more technical point of view:
VFD's use inverters to recreate a sine wave to manipulate and achieve speed control of the motor. to do that, they first charge a set of DC capacitors. then the DC current is discharged by using rectifiers (basically electrical gates that are open and shut very quickly at the carrier frequency) are operated to maintain a voltage to hertz ratio. this create harmonics that are injected into both the power supply and motor leads.
the cables help because they help manage the harmonics and leakage capacitance that occurs between all cables in proximity to each other. in most cases we can ignore this leakage for plant environments provided we provide separation between very low voltage (think 24VDC and 4-20 mA signal) and anything power related (120V 1-phase, 480, 4160 volt, etc). think of the motor leads as a large capacitor. the longer the run, the more conductor you have in parallel, the higher the capacitance and therefore voltage that can exist between them. harmonics make this worse and the elevated voltages will cause sparks across the motor bearings and shaft interface, causing fluting and pitting of the bearings, eventually leading to catastrophic failure.
Always use XHHW MINIMUM for even small VFD's because it is less susceptible to insulation breakdown than THWN/THHN. why?
THWN/THHN is a thermoplastic process where it is melted onto the conductor. it is a less than perfect process, leading to microscopic pinholes in the insulation. Not normally a problem for regular power and control. however, for VFD, these little holes will aid in the breakdown of insulation when exposed to higher voltages present in VFD outputs. cross-link polyethylene (XHHW) is an extruding process and has no pin holes and is therefore more resistant to insulation breakdown.
always use a DV/dt output filter for larger motors (50HP+), unless there is a good reason not to.
always use a sine wave output filter for larger motors far away from the VFD (~700 ft, depends on VFD manufacturer).
Totally agree.
KDough
Member
- Location
- San Jose, CA
- Occupation
- Electrician 42 yrs / Electrical Inspector 4yrs
Thank You drktlmpr12 for that great explanation. I have never had anyone describe so clearly how a VFD functions. I didn’t ask the question but I learned something new today. Thank You again.
drktmplr12
Senior Member
- Location
- South Florida
- Occupation
- Electrical Engineer
- Location
- Placerville, CA, USA
- Occupation
- Retired PV System Designer
Another process which comes into play is that the wiring, whatever type and configuration can be characterized as a transmission line for high frequency harmonics, with a chacteristic resistance which depends on the distributed inductance and capacitance. Since the impedance of the motor windings will never match that characteristic resistance there will always be reflections at the connection to the motor. This in turn leads to a standing wave pattern on the wiring, with points where the peak voltage is at least double the nominal voltage. This can cause insulation breakdown in a pattern along the wire run which depends only on the carrier frequency rather thann the motor speed.
often the highest voltages in the pattern are at the nodes closest to the motor.
often the highest voltages in the pattern are at the nodes closest to the motor.
It’s utter nonsense.
As far as reflected waves:
Google transmission lines.
So if you launch a pulse down a cable at the end where the motor sits the impedance is very different. It’s a big inductor compared to the cable.
Physically it’s like making a wave in a pool with a big section and a narrow section. When the wave travels down either one when it hits the change some of the wave goes through but some bounces off the edges and returns in the opposite direction.
The ratio can vary but at the joint we have the incoming pulse, tge outgoing pulse, and the reflected pulse. If the two sections are identical then it just moves like nothing happened so no reflection. Worst case nearly all of the pulse is reflected. In this case we get up to twice the pulse height or voltage. So since the pulse is equal to the DC bus which is 145% of the AC RMS input it can be around 1400 V. At around 200-250 feet this is where it is for most VFDs.
Now the wave returns to the VFD, bounces back again, bouncing back and forth until it settles out. This causes a ringing pattern. BUT if the cable is long enough then a second pulse can occur before the first one dies down. So now we can get up to 4x the DC bus.
Now 600 V cable is tested per ICEA at 200% plus 1 kV or 2200 V. NEMA did a bunch of testing. The thinnest cable, #14 THHN, fails at 2850+ V. Now the highest rated motors fail at around 1750 V. So by the time you have cable failures you will have motor failures. So what is the benefit? Higher margins for the manufacturer.
EMC:
It is very true that VFDs can emit lots of RF noise and even interfere with controls. So be careful to use shielded signals. Ir put the cable in metallic conduit.
Bearing fluting:
This is two fold. The first is that all motors have bearing currents. If the cabling is say 3 phases and a ground in a cable tray in that order as power travels down the cable you can create a condition where the voltages are unbalanced because the capacitance between the cables and ground is different. This can create circulating bearing currents but in general although the problem is dubious at best, any typical install (MC, multi conductor test cable, XHHW ir THHN in conduit with a ground) has no issues.
Second is VFDs cause common mode currents by nature. These can cause non circulating so-called “EDM currents” in bearings as well as circulating currents both of which can lead to rapid bearing destruction.
As far as “shielding” running it in metallic conduit will do, Ir bundling or using multi conductor tray cable, or using MC all do exactly the same thing. Just make sure to run a ground.
There are various other bogus claims.
I retired from a large hospital/research center that had over 500 VFD'S from 1 to 1750 HP. All used type THWN common building wire. ( The 1750 HP drives were on 4,160 volts and feed with medium voltage wires ). As a rule all but a few tower fans drives were locstef within 100' of motors. In one 12 story building all drives over 40 HP were 18 pulse and they all were in 100' of motors.Motor life was bery good. Had a lot of standard NEMA motors ( not designed for drives ) that were converted to drives 20 years ago and still have the original motors. While moonlightning have come across some machines that had wooven metal jacket for drives. We PM'ed every drive & motor including IR scan 1 to 4 times a year and never found a motor running at a higher trmpetture then if feed with a starter. Most motors ran below 90% of full speed thus lower power and cooler temperatures. We set min speed at 25% and wish some motors in a very warm area that ran that slow would be fitted with an external fan.
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
The OTHER other issue is one of RF transmission. The output of a VFD can act like a radio transmitting antenna, "broadcasting" its noise all over the place to where it can cause severe interference with other equipment in the area. I once had some MPs track us down when doing a startup of a machine in the desert outside of Las Vegas because the 5HP VFDs were all wired with unshielded SO cord to the motors (because everything was moving). We were on a nuke test site and the MPs were 25+ miles away but were concerned for what the heck that signal was. They came at us with guns drawn... We had authorization to be there, but they didn't know what we would be doing with those kinds of RF signals. Shielding is important for that reason alone, although simply enclosing the wires in metal conduit will take care of that too. But elsewhere in the world they don't use conduit like we do, so the VFD mfrs always say to use shielded cable just in case. We would need to if running in cable tray or PVC conduit.
OTHER other other issue, albeit minor, is one of conductor geometry. Using VFD cable that is pre-twisted and jacketed creates a situation wherein the cable geometry, i.e. the relationship between one phase wire and the other, is consistent. This too has an effect on the cable capacitance and results in lower incidence of standing wave creation. Again, you can do the same thing by "triplexing" your cables before pulling them into conduit, but hardly anyone bothers to do that.
OTHER other other issue, albeit minor, is one of conductor geometry. Using VFD cable that is pre-twisted and jacketed creates a situation wherein the cable geometry, i.e. the relationship between one phase wire and the other, is consistent. This too has an effect on the cable capacitance and results in lower incidence of standing wave creation. Again, you can do the same thing by "triplexing" your cables before pulling them into conduit, but hardly anyone bothers to do that.
drktmplr12
Senior Member
- Location
- South Florida
- Occupation
- Electrical Engineer
this is a truly great story
The OTHER issue I’ll illustrate this way. A 900 HP motor had a serious bearing current issue, 90 A. After a lot of filtering and insulating AND adding VFD cable which helped but not for VFD cable reasons it was down to 18 A, of which 8 A made it to the motor and the other 10 A is created by cable losses. I’ve seen 2 A common mode on a small motor where it is entirely no circulating currents, 6 A with VFD cable at the drive, 2 A at the motor. VFD cable creates higher shunt currents nearly every time with little to no improvement in bearing currents. The issue is asymmetrical currents leading to bearing currents. Similarly it barely moves the needle on reflected waves. The issue is pulse rise time. With its lower characteristic impedance, Z0, it increases the impedance mismatch to the motor, increasing the reflection coefficient,
- Status