I retired from a large campus that had over 500 VFD'S. Never wasted money on shielded cable on there 480 volt output side of any drives. Used plain Jane THWN copper wire for every 6 & 18 pulse drives. Most motors were plain Jane NEMA motors & the rest high efficiency motors. Had some 40 year old motors that were on drives. Even had two 4,160 volt 1,750 HP chiller motors on drives with regular medium voltage cable. Most drives were less then 40' from motors.
Considering the cable lengths were mostly under 40' how many cents do you reckon you saved skimping on the cable? As a service tech doing warranty inspections and repairs for an HVAC manufacturer (northern Europe) I've replaced bearings (or the whole motor) on hundreds of under 2 year old 1-20kW VFD driven motors. The bearing grease is usually burnt into a black tar or a crumbling solid and the races are pitted or cloudy. Many have been wired with an MCMK cable with 3 phase conductors and a shield where the shielding copper foil and strands have been twisted at the ends into a single PE and EMC conductor. A version with 3 phase wires, a designated PE of the same gauge or thicker and a separate EMC shield would have cost between 5 to 10 cents more per meter. Some have had a proper cable but the grounding contact of the shield inside the metal glands hadn't been done properly. Often a safety switch between the VFD and the motor had been the culprit. Bearing replacement on a large motor can cost a 1000€ while the unit is out of service for the day. I think VFD motor drive cables should be the last thing to skimp on. I tell that to all our customers and recommend improvements on new sites. Many times I've known right away that unless the cabling is corrected we'll be back in a year or two.
Update: Just last week I was called back to one site I wa specifically referring to 6 months ago. To no surprise it was to investigate a noise in a VFD driven 3kW fan motor. The noise was a low rumble that's typical for a bearing damaged by EDM. The heat from the arcing has burned away the lubrication and the balls roll dry against the micro-pitted races. That kind of damage usually escalates very rapidly and the bearings have a tendency fail catastrophically suddenly ejecting the balls ripping apart the cage. I've seen a quite a few lately and everyone had problems in grounding, shielding.or both. Now they want a comprehensive report on the cause and mechanism of the problem and instructions on how to improve the cabling. There are around 40 units where the the VFD to motor cabling is done using three phase conductors and a twisted shield serves both protective and functional purposes. All of them have a disconnect switch between the drive and motor, too. The electrical contractor isn't going to be happy and might take the case to court. Be my guest. The Schneider ATV212 VFD wiring instructions specify a separate PE conductor. It's mostly a safety feature that has to be able to carry high currents at lower frequencies. They highly recommend using a continuous EMC shield between drive and motor when operating multiple units in the same space. It's function is to be a current carrying conductor for the high frequency instabilities, harmonics and common mode currents a VFD inevitably produces. The shield's purpose is to offer a low impedance path for the HF and RF currents back to the source, the inverter or the rectifier. Every discontinuity in impedance, ie.g. twisted strands screwed to a terminal in a safety switch, reflects part of the current back to the motor. Such a cable radiates EMI to it's surroundings. I'm building a current transformer for my scope so I can demonstrate the effect and the remedy.
this is a brilliant presentation, thank you. The only thing I would add is that conductors need to be SIZED TO PASS THE PEAK CURRENT density, not the average value. This helps transfer the dV/dT spikes into the motor without dissipating energy in the serial impedance of an undersized sized cable. (ppl talk about reflecting or bouncing waves when energy can't go through!) This is is true with any Pulsed Modulation: "pass the peak not the average"! 🤞
@6:58 "in this scenario..." what is different in this scenario such that grounding on both ends is the recommended way to go vs. only grounding at one end only in some other kind of scenarios?
Great information, but I note Belden does not recommend the use of 360 degree shield terminating cable glands. With a shieled terminating gland, significant shield currents can be released, for example at enclosure ingress, exposing sensitive equipment to noise currents that would be contained if an isolating cable gland such as a Crouse CGB, ADE or CMP TC or A2 Gland. Pete C
I don't know where you read that, but 360 bonding of cable screens at BOTH ends to the equipment chassis is almost always best for EMI. The old advice to bond only at one end was typically to avoid the potential for excessive currents to flow through the screen and damage them, but there are ways to address this concern.
very informative, but one question - doesn't the motor act as as smoothing inductor (reactor) and smooth out the square wave at least somewhat in the entire circuit?
You're asking, will the motor’s inductances do the same job as an output load reactor? The answer is absolutely not. The output load reactor is placed next to the drive. This slows down the rise time before it even gets to the cable. This helps in many ways and if you'd like to know more, I can tell you. You might not believe this but smaller motors have more problems with pulse width modulation then larger motors when you are running long cables. If you'd like to know, just ask.
Geee, build the vfd power pack on the motor back-end or very close to the motor, then no more long cables... sorry belden...:) or go back to the one common line shaft or adjustable gearbox
no help lot my project, i build DIY cnc machine and just buy 2,2KW VFD and spindle motor, i used VFD input 230V single phase, output have 230V 3-phase to spindle motoer, i not know what cable need use, i think need be 4x2,5mm2 shielded cable inverter to motor, have this right? wall socket not need be shielded cable ? normal power to VFD 3x2,5mm2 cable. need cable be 2,5mm2 if VFD and motor have 2,2KW or what cable need be, curret have motor 8Ampere, or can i use 1,5mm2 cable (lot cheapen than 2,5) europe normal 230V main power.
I retired from a large campus that had over 500 VFD'S. Never wasted money on shielded cable on there 480 volt output side of any drives. Used plain Jane THWN copper wire for every 6 & 18 pulse drives. Most motors were plain Jane NEMA motors & the rest high efficiency motors. Had some 40 year old motors that were on drives. Even had two 4,160 volt 1,750 HP chiller motors on drives with regular medium voltage cable. Most drives were less then 40' from motors.
I'm pretty sure the MV Cable confers the same advantages as VFD cable because it's also shielded
Considering the cable lengths were mostly under 40' how many cents do you reckon you saved skimping on the cable?
As a service tech doing warranty inspections and repairs for an HVAC manufacturer (northern Europe) I've replaced bearings (or the whole motor) on hundreds of under 2 year old 1-20kW VFD driven motors. The bearing grease is usually burnt into a black tar or a crumbling solid and the races are pitted or cloudy. Many have been wired with an MCMK cable with 3 phase conductors and a shield where the shielding copper foil and strands have been twisted at the ends into a single PE and EMC conductor. A version with 3 phase wires, a designated PE of the same gauge or thicker and a separate EMC shield would have cost between 5 to 10 cents more per meter. Some have had a proper cable but the grounding contact of the shield inside the metal glands hadn't been done properly. Often a safety switch between the VFD and the motor had been the culprit.
Bearing replacement on a large motor can cost a 1000€ while the unit is out of service for the day. I think VFD motor drive cables should be the last thing to skimp on. I tell that to all our customers and recommend improvements on new sites. Many times I've known right away that unless the cabling is corrected we'll be back in a year or two.
Update: Just last week I was called back to one site I wa specifically referring to 6 months ago. To no surprise it was to investigate a noise in a VFD driven 3kW fan motor.
The noise was a low rumble that's typical for a bearing damaged by EDM. The heat from the arcing has burned away the lubrication and the balls roll dry against the micro-pitted races. That kind of damage usually escalates very rapidly and the bearings have a tendency fail catastrophically suddenly ejecting the balls ripping apart the cage. I've seen a quite a few lately and everyone had problems in grounding, shielding.or both.
Now they want a comprehensive report on the cause and mechanism of the problem and instructions on how to improve the cabling. There are around 40 units where the the VFD to motor cabling is done using three phase conductors and a twisted shield serves both protective and functional purposes. All of them have a disconnect switch between the drive and motor, too. The electrical contractor isn't going to be happy and might take the case to court. Be my guest.
The Schneider ATV212 VFD wiring instructions specify a separate PE conductor. It's mostly a safety feature that has to be able to carry high currents at lower frequencies. They highly recommend using a continuous EMC shield between drive and motor when operating multiple units in the same space. It's function is to be a current carrying conductor for the high frequency instabilities, harmonics and common mode currents a VFD inevitably produces. The shield's purpose is to offer a low impedance path for the HF and RF currents back to the source, the inverter or the rectifier. Every discontinuity in impedance, ie.g. twisted strands screwed to a terminal in a safety switch, reflects part of the current back to the motor. Such a cable radiates EMI to it's surroundings. I'm building a current transformer for my scope so I can demonstrate the effect and the remedy.
Thank you Al. You are more important to us than you will ever know.
Excellent video except you kept saying XPLE instead of XLPE towards the end of the video. Thanks for the explanation fantastic information
Fantastic, well explained video!
this is a brilliant presentation, thank you.
The only thing I would add is that conductors need to be SIZED TO PASS THE PEAK CURRENT density, not the average value.
This helps transfer the dV/dT spikes into the motor without dissipating energy in the serial impedance of an undersized sized cable. (ppl talk about reflecting or bouncing waves when energy can't go through!)
This is is true with any Pulsed Modulation: "pass the peak not the average"!
🤞
Thank you Al,
Excellent introduction to cables.
Best Wishes,
Aaron
@6:58 "in this scenario..." what is different in this scenario such that grounding on both ends is the recommended way to go vs. only grounding at one end only in some other kind of scenarios?
Very instructive. Thank you for posting.
Great information, but I note Belden does not recommend the use of 360 degree shield terminating cable glands. With a shieled terminating gland, significant shield currents can be released, for example at enclosure ingress, exposing sensitive equipment to noise currents that would be contained if an isolating cable gland such as a Crouse CGB, ADE or CMP TC or A2 Gland. Pete C
I don't know where you read that, but 360 bonding of cable screens at BOTH ends to the equipment chassis is almost always best for EMI. The old advice to bond only at one end was typically to avoid the potential for excessive currents to flow through the screen and damage them, but there are ways to address this concern.
Thank you for the information, boss
Should be XLPE (Cross Linked Polyethylene)?
Very good information and explanation. Thank you.
very informative, but one question - doesn't the motor act as as smoothing inductor (reactor) and smooth out the square wave at least somewhat in the entire circuit?
You're asking, will the motor’s inductances do the same job as an output load reactor? The answer is absolutely not. The output load reactor is placed next to the drive. This slows down the rise time before it even gets to the cable. This helps in many ways and if you'd like to know more, I can tell you. You might not believe this but smaller motors have more problems with pulse width modulation then larger motors when you are running long cables. If you'd like to know, just ask.
Sort of. What the motor inductance DOES do is smooth the motor CURRENT so that it approximates a sinewave, despite the VOLTAGE being a PWM mess.
Geee, build the vfd power pack on the motor back-end or very close to the motor, then no more long cables... sorry belden...:) or go back to the one common line shaft or adjustable gearbox
what would i start with to learn about this general area? something more specific than "electrical engineering"
Slide says LGBT instead of IGBT at 0:48
maybe all along it was just a lowercase L, and we've all been misreading it
" I " for Insulated
" L " is for someone on the crew has a sense of humor.
ROFL.
It’s the newer type Lesbian gay bipolar trans-sister
Super video😊
no help lot my project, i build DIY cnc machine and just buy 2,2KW VFD and spindle motor, i used VFD input 230V single phase, output have 230V 3-phase to spindle motoer, i not know what cable need use, i think need be 4x2,5mm2 shielded cable inverter to motor, have this right? wall socket not need be shielded cable ? normal power to VFD 3x2,5mm2 cable. need cable be 2,5mm2 if VFD and motor have 2,2KW or what cable need be, curret have motor 8Ampere, or can i use 1,5mm2 cable (lot cheapen than 2,5) europe normal 230V main power.
Just to clarify no special cabling necessary on input side?
Thus the comment drive as close to motor as possible.
Volunteer fire Department
Was he even speaking English? #iamdumb