Long cables from VFD to motors - Shielded or not? 5

[JuanBC]

Hello,

I wonder if anyone has a nice "guide for dummies" that cover the advantages/disadvantages of using shielded/unshielded cables when feeding motors through VFDs

I always thought that shielded cables were a must but when having long distances, manufacturers recommend unshielded cables

Thanks in advance!

JBC
.......
"The more I read, the more I acquire, the more certain I am that I know nothing"

 

[controlsdude]

Google is your friend
vfd manufacturers guide to long motor leads

https://library.e.abb.com/public/fec1a7b62d273351c12571b60056a0fd/voltstress.pdf

http://literature.rockwellautomation.com/idc/groups/literature/documents/in/drives-in001_-en-p.pdf

https://mobile.yaskawa.com/delegate...D.05&cmd=documents&documentName=AR.AFD.05.pdf

http://www.wrp-llc.com/uploads/VFD and Motor Insulation Stress.pdf

just read PDF till your blue in the face and then ask your questions.

 

[Marke]

Hello JuanBC

Have a look at

http://www.vfd-emc.co.nz/index.php

and

http://www.no-bs.co.nz/index.php

The important factor is that the PWM waveform with fast edges causes a high frequency current to flow between the stator winding and the stator frame. This current is driven from the DC bus by the output switching devices, and it needs to get back to the DC bus to complete the circuit.
One path is back to the frame of the VFD and from there through the EMC capacitors back to the DC Bus. Another path if via the entire electrical ground network, back to the AC supply via every bit of equipment connected to the supply, and then back to the DC Bus via the input rectifiers.
To minimize the spread of this high frequency current, and thereby minimize the interference from it, the best solution is to provide a very low impedance path between the frame of the motor, and the frame of the VFD. This low impedance path should be at around 150KHz to be effective. That requires considerable surface area alone the entire path. The use of a screened cable is promoted because that gives a low cost return path with a typical impedance at 150KHz, that is around 1/50 of the impedance of the standard phase (and earth) conductors contained within that cable.
If it is not practical to use a screened cable, then use a conductive surface of a similar surface area as a return path. I have used Al strips, Cu strips, galvanized steel pipe, anything conductive, BUT YOU MUST NOT HAVE ANY LENGTH OF circular conductor in the path. It will affect the effective impedance of the whole path.
It is not necessary to have the conductors enclosed in a screen, rather to provide a very easy path home. Any restriction on the way has to be avoided.

Mark Empson
Advanced Motor Control Ltd

 

[JuanBC]

Hello,

Thank you both for your guidance! I've read the pdfs you recommend and "reach" the following conclusions:

There are (3) three main problems that may occur when feeding motors through VFD:

· Voltage overshoot (may damage/destroy motor's insulation)
· Electrical Discharge Machining -EMD- (may damage motor's bearings)
· Cables from VFD to motor are a source of electrical interference known as "EMC". EMC can interfere with radio and communications equipment

1. Voltage overshoot: Happens when a fast-rising voltage pulse is applied to the stator coils of an AC induction motor.
This Vpeak can be as high as Vp = Safety factor x 2 times x DC bus voltage = 1.1 x 2 x sqrt(2) x Vrated. ; Vrated is the network RMS value

If the motor is spec. as "inverter duty" it can withstand this peak so there is nothing to be worried about. (Note that the value of "Vp max" on motor's datasheet should be higher that the value calculated before).

If the motor is spec. as "general purpose motor" different types of "filters" need to be installed on the VFD output in order to limit Vpeak.

The shape of the pulse that the motor receives depends on the interaction between the VFD, the motor power cables, and the motor.
The longer the cable, the more important the phenomenon.
The use of shielded cables increase Vp?

I do not understand what should I do with the values of "min. rise time" and "max. dV/dt" that motor's manufacturers give in their datasheets
With what data should I compare these values?

2. Electrical Discharge Machining (EMD): Can be caused by electrical discharges within the bearing itself and this can result in severe bearing degradation and failure over time.
Insulated Bearings, Shaft Grounding brushes and Insulated Coupling between motor shaft and driven load mitigate EMD.

3. EMC: Using shielded cables reduce EMC but... increase Vpeak? Increase EMD? I do not understand the negative impact of using shielded cable


JBC
.......
"The more I read, the more I acquire, the more certain I am that I know nothing"

 

[Marke]

Hello JuanBC

The use of screened cables results in increased capacitance per unit length on the output of the VFD and this will increase the capacitive current on the output of the VFD. This can increase the EMC.
The use of screened cables reduces the velocity factor of the cable and so standing waves set up by the high frequency components are shorter for any given frequency. In effect, this can result in the maximum transient voltages being reached with shorter cables than with unscreened cables. A rule of thumb applied by some, is that at 150M of screen cable, the peak voltage at the motor will double. Others will claim that today, that figure is much less.
Most describe the requirement for screened cables as a means to stop the cable radiating, but at the wavelengths we are talking, the cable on its own, even unscreened is not a major radiator, however if there is not a good High Frequency return path as an alternative to the screen, then the whole structure and installation becomes part of the radiator. Hence, poor installations without screened cables can exhibit strong radiation characteristics at 150KHz.
The slower the rise time of the waveform, the lower the frequency component of the switching noise and the longer the wavelength of the standing waves set up along the cable length. This will reduce the peak voltage amplification for cable lengths less than a quarter of a wavelength. So for a given cable length (less than a quarter wavelength) the peak voltage should reduce ans the voltage rise (dv/dt) is slowed down.
EDM is influenced by the high frequency common mode voltage. Reducing the high frequency common mode content will reduce the EDM in most cases.
EMC will not only interfere with communications equipment, it can interfere with any electronic apparatus on the same electrical supply. I have seen many flow meters damaged and/or giving incorrect readings due to EMC. Last week I came across temperature metering using PT100 in the motor that was very eronious due to EMC.

Mark Empson
Advanced Motor Control Ltd

 

[itsmoked]

Other things to note Juan;

The "inverter duty" aspect is much more important with 480V or higher motors because the insulation systems used start topping out in the higher voltage motors where over voltage peaks more easily exceed the insulation system.

At the lower voltages (~240V) most newer motors have no issue with being driven by VFDs as modern insulation systems can handle it without issue.

All VFDs allow the user to set the modulation frequency, often offering 10 or more different frequencies. Lower frequencies tend to make the motor noisier but lower frequencies are MUCH easier on both the motor and the VFD. This helps reduce the standing waves and it considerably lowers the heating in the VFD's switching transistors. It also reduces the EMC since there are less cycles and sharp edges.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[edison123]

Keith - Is it possible to change the modulation frequency on the fly? i.e. start with one frequency and then change it midway without powering down?

Muthu

http://www.edison.co.in

 

[itsmoked]

Hi Muthu.

I have never seen a drive that can do that since the DSP running the Clarke-Park transforms and creating the PWM output generally can't switch something that fundamental in operation but perhaps someone figured it out. I haven't seen all that many brands. Usually you just have to issue a stop command change the modulation parameter to the desired freq and hit GO again. You don't have to power cycle the drive or execute a reset.

Running drives to run aquarium features I'd change the modulation frequencies a lot trying to find one that made the least disturbing sounds. Often that was somewhere in the upper 2/3rds of the range. I didn't worry about the how high the freq was because the motor side cable lengths tended to be less than four feet. In that app it turned out to be completely hopeless to use any motor that wasn't totally enclosed. The smallest opening would let out skin-crawling room-evacuating noise at most frequencies and merely unacceptable noise at choice settings.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[flexoprinting]

I am having a problem understanding Mark's statement,

To minimize the spread of this high frequency current, and thereby minimize the interference from it, the best solution is to provide a very low impedance path between the frame of the motor, and the frame of the VFD. This low impedance path should be at around 150KHz to be effective. That requires considerable surface area alone the entire path. The use of a screened cable is promoted because that gives a low cost return path with a typical impedance at 150KHz, that is around 1/50 of the impedance of the standard phase (and earth) conductors contained within that cable.

1. So the ground lead within the shielded cable is inadequate due to surface area or skin effect and will not carry all the current?, the shielding of cable is used as a supplemental carrier?

2. How could one test (measure) for this situation as a possible cause of say control issues, would you use a scope on ground of entire system to see magnitude of stray DC pulses?

Chuck

 

[Skogsgurra]

Flexo
Stray currents go from stator windings to motor frame via winding-to-slot wall capacitance. There are also stray currents from cable conductors to cable shield to screen or armour. These currents seek ground (one could also say that they want to go "home" it is a good pedagogical picture, but it may obscure the understanding).
The impedance in the path from motor frame to ground determines if the HF voltage drop in the return path will be huge or small and you could either measure frame voltage against "Mother Earth" (if you can find her) or against the driven machinery. I mostly measure against driven machinery because it is the voltage across the bearing that is critical and the voltage difference between motor frame and driven machine usually is dropped over the DE bearing.
If you want to avoid bearing EDM and also reduce EMI, it is good to create a low-impedance path from motor frame to driven machine. Especially if the latter is a huge thing with good ground connection and tubing for water, gas, reject or products attached to it. You can't find better ground anywhere even if you buy a ton of copper plate and bury it. The reason for that is that the cable needed to connect to the immense ground plate still has around 1 microhenry per meter and the PWM edges (100 or 200 ns with 600+ V) create spikes with anything between a few volts and up to 20 V amplitudes. Often with ringing that makes the EMI worse.
Bonding the motor frame to the machine with a straight tinned Cu braid will create an easy path to machine (as close to Mother Earth as you can get) for the stray currents and is the number one remedy when you have EDM in large motors.
Your question about where to measure? I always measure between motor frame and driven machine and then add one or two braids* to get peaks down to one or two volts. I then check voltage across the DE bearing to see if it is “low enough” and that there are no discharges through the oil film.
*A note on braids and copper wires: The braid is effective as long as the length/width ratio is below around 3. For long braids, the impedance is about the same as it is for a wire. I therefore use naked Cu wires placed as wide apart from each other as possible as soon as distance between motor and machine is one foot or more. They are easier to find, easier to connect and clamp and cost less.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[itsmoked]

Gunnar; "Low enough" means what? ~2V?

How do you measure across the DE bearing? Probe to spinning shaft and end bell?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[jraef]

Is it possible to change the modulation frequency on the fly? i.e. start with one frequency and then change it midway without powering down?

Several VFDs mfrs do something similar, but it is cooked into the firing algorithm, it's not something the user can adjust on the fly. Rockwell calls it their "Reflected Wave Voltage Protection", I've also heard it referred to as a "Standing Wave Reduction Algorithm" and even a "Whisper Wave" setting, etc. etc. But rather than changing it based on speed, the algorithms usually change the carrier frequency WITHIN each sine wave, thus changing the way it interacts in the circuit with the line capacitance, impedance etc. In all of them that I've seen, the only option you have is to turn the feature on or off. The reason to turn it off is for when you are going to permanently set the carrier frequency high, often for performance issues when looking for extremely fast responses to step changes in load (i.e. servo-like performance), or where the VFD capacity is too closely matched to the motor current requirements (increasing the carrier frequency also increases the switching losses in the IGBTs of the drive, so you must de-rate the drive if using a high CF).


" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden

 

[Skogsgurra]

That is a complex question. For standard four/pole motors up to a few hundred kW, a couple of volts peak value is OK. If there is a large capacitive load, like a huge fan, with 10 or 20 nF extra capacitance in addition to the rotor-frame capacitance, even a couple of volts can cause damage.
Then, there are cases where bearing temperature is high. That usually means that the oil film break-down voltage is low and that makes you feel safe. But it also means that there is a metal-metal contact that makes the induced shaft voltage drive a circulating current through the bearings. Which is a bad thing. There is no easy answer. Too high is definitely bad and very low should rise some suspicion. If bearing temperature is high and the machine can be stopped, then let it cool down and measure again to see what the voltage is.

Always use a scope for these measurements. A DMM, even the faster ones with up to 50/100 kHz BW, miss the microsecond wide induced shaft voltages.


Yes, measure from shaft end to ground. End bell is good. OSHA? You bet. So we use a KCA (a Kjell Carlsson Attachment), google EDM KCA. It has been OK:d by safety officers - even in the US.
For NDE-DE measurements, use a KCA plus a probe in the NDE center bore.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[flexoprinting]

Thank you Gunnar for that jewel of information! and thank you JuanBC for asking that question.

 

[edison123]

Thank you, Keith and Jeff. Learnt something new today.

Eng-tips: Come for the motors, stay for the drives.

Muthu

http://www.edison.co.in