Grounding (again): VFD and Control cabling differences or similarities

G

Thread Starter

Gerald Beaudoin

Control wiring practice dictates that the shields should only be grounded at one end for reasons already discussed at length in this forum. However, according to manufacturer's recommendations, the newer generation of shielded VFD cables have the drain wires grounded at both ends. Do we not run the risk of creating ground loops by grounding both ends in the case of VFD wiring? Why don't the same rules apply in both cases?

Thanks.

Gerald Beaudoin
 
C

Curt Wuollet

If they are _power cables_ ground loops are not as big an issue as conducting harmonics that would be radiated to ground. For small signal cables, the ground loop current could be large enough to induce hum and noise on a small signal. Within reason, we aren't concerned about the effect of
ground loop currents on power.

Regards
cww
 
P

Phil Corso, PE

Responding to Curt’s 26-Jul (23:50) comment that grounding the power cable shielding at both ends is not as big an issue as is harmonic radiation. I strongly disagree! Such a practice will negatively impact ground-fault protection!

Regards, Phil Corso ([email protected])
 
At the Chicago area Profibus seminar a couple months ago, the Profibus speaker had Powerpoint slides showing Profibus' recommendation for grounding shields at BOTH ends of the RS-485 cables, which are routinely run at 1.5Mb, that's M as in megabit/sec speeds.

It flies in the face of tradition.

David
 
C

Curt Wuollet

And I've seen a case or two where those grounds were lifted to fix a network problem. The problem is that automation networking is over simplified and the rules are most notable for the exceptions. Profibus hardware tends to be very reliable, except when it isn't :^) And it's got some of the least reliable connectors I've ever seen.

Regards
cww
 
C

Curt Wuollet

Hi Phil,
Yes, it will. But not doing it will often not meet emissions and radiation standards which is why I suspect they are making the trade off. Running these through grounded conduit can also do really strange things for both factors, but is often required per code.

My own favorite is simply high grade individual wires to the motor in conduit. The lower distributed capacitance helps with both issues, it's cheaper , and it's familiar to electricians, inspectors and techs and I don't have to get it from Stuttgart if I need to replace it.

I would only use shielded cable inside an enclosure, but many packaged cables do indeed ground both ends of the shield. The harmonics play hell with ground fault protection no matter what. The whole area is gray, fuzzy and a mess. But as long as they want to sell VFDs packaged, like a loaf of bread, to people who just want to bolt it in and run it, in the zany map of regulatory bodies they can only guess as to what wiring will be needed.

If you've ever tried to explain why the manufacturers wiring methods don't work but SO cord laid on the ground does, you see the problem right away. Of course, it'll kill anyone with a pacemaker that walks by, but explain that to a crusty old electrician.

Regards
cww
 
P

Phil Corso, PE

Responding to Curt's 28-Jul (10:34) comments... presumably you invoked the The Engineer's Code of Ethics in a letter to your client explaining the impact on safety!

Regards,
Phil
 
P

Phil Corso, PE

Responding to Bob’s 28-Jul (14:08) query... here are several whys:

a) With the shield grounded at both ends the GFP output is effectively short-circuited, thereby compromising proper trip action!

b) There have been cases, where, although properly grounded at the source end, inadvertent grounding elsewhere along the cable's path, has negated trip action!

c) And, sometimes the double-ground can result in unwarranted (some unexplained) trips!

Regards, Phil
 
C
That would be unethical. I am not an engineer, although I have played one in the real world.

Regards
cww
 
P

Phil Corso, PE

Bob, if you like, I can send you a case-history study covering the subject of double-grounding a shielded cable, i.e., at each end!

BTW, a simple test with a clamp-on ammeter clearly demonstates my contention that such practice is not only dangerous, but could result in false-tripping.

Regards, Phil Corso ([email protected])
 
To further add fuel to this thread- take a look at ABs System Design for Control of Electrical Noise-Publication GMC-RM001A-EN-P — July 2001. It states on page 8-2:

"Observe the following guidelines when bonding the motor power cable shield to ground. Bond motor power cable shields:

* At the motor frame.

* To the panel at entry to the cabinet (optional).

* To the drive (amplifier) chassis. If a connection point is not
provided, bond to the adjacent panel."
 
It seems that a number of publications suggest grounding the shield at both ends, and also having a separate ground wire in the sheath connected at both ends (of course).

I'm just rewiring a CNC Router and installing a new VFD and spindle motor. I've decided to rewire all of the power paths, and using much better quality cable than was originally provided. My machine just had 3 wires and a shield between the VFD and the motor, and seemed to rely on earthing through the bearings - frightening. I'm replacing this with 4 wires and a shield, and at the motor end branching the earth wire to a separate lug on the gantry chassis.

Here are a couple of links:

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

http://www.automation.com/library/a...-to-consider-in-cabling-a-reliable-vfd-system
 
J

John R. Goodman

We are obviously talking about two different subjects in the same discussion. In context the information I publish here is relevant to US NEC published by NFPA as well as 42 years of experience in the electrical trade as well as 25 years in automation.

1) Clearing a ground fault (electrical line voltage short circuit phase/phase, phase/ground) is a function defined in the US as an equipment ground. This is essential, required by codes and must not be ignored. And no, a ground rod through the floor cabled to the machine is NOT an equipment ground. It is a good way, if substituted for the equipment ground to kill someone! I know from PERSONAL experience if the equipment ground from the supply panel to the VFD for example, is not extended as an equipment ground all the way to the motor from the chassis of the drive to the motor (Bonding, creating a low resistance path back to the supply) there will be potential from the motor to ground (anything of a different potential) due to the fact a VFD is much like, in concept, a separately derived system (480 Delta to 208Y120 transformer with neutral).

2) Noise issues arise, although a different discussion and still joined at the hip with the equipment ground issue,must be handled in a manner that we NEVER EVER unland the equipment ground or fail to provide the green (green/yellow) wire along with the power conductors.

3) Also, the electrical system must be correctly wired. In the US the system ground (transformer neutral bonded to earth) and the service panel bonded 1 time to earth from the panel neutral and to earth is the correct way basically. The equipment ground is born in the service panel or at the actual service by bonding the two (neutral and newly created equipment ground) together and MUST remain separate from the neutral everywhere else! Neutrals grounded at every panel and/or neutral and equipment ground busses bonded together at every panel create NOISE out the ears, as well as a great way to kill people. (horrors!!)

4) Once we do our homework and make sure the electrical system connections in the distribution system are in order, THEN we can address noise. a) In panels we NEVER daisy chain bonds (equipment grounds). Each PLC chassis, Each Drive, any microprocessor based item or rectifier/converter gets a DEDICATED ground conductor from the component to the ground buss in that cabinet. The cabinet is then bonded (grounded by common understanding) to a good equipment ground with cable 2 sizes larger than required by code.

5) Each PLC needs an RF filter ahead of it (Corcom for example) CVS regulators (expensive) ahead of the 120V supply to the cabinet assures good voltage stability, improved surge tolerance and all but eliminates noise. "Pay me now or pay me much more later".

6) We need to understand the system ground is not for short or ground fault clearing. Among other things it is for rapid dissipation of PIT, peak impulse transients such as lightning close by (NOT direct strikes) utility activities (switching capacitors, etc)

7) We need to understand the EQUIPMENT ground is for preventing differences in potential which are the root in all malfunction of processor based equipment. The equipment ground does not PREVENT processor and networked system malfunctions, but does largely eliminate "gremlins" that go and come.

All in all BOTH the ground fault (clearing protective devices) and noise issues have to be considered. Low level arcing or other protective systems (Earth Leakage Circuit Interrupters, trip when equipment damage is an issue 30-100 Ma) or Class A personnel protection (Bender GFCI, Littelfuse trip at 5 Ma) will trip at low levels of leakage when personnel risk is in play.

All in all the electronics engineer cannot play in a vacuum. He must coordinate his safety (first priority) and noise reduction issues with the power engineer to produce the best noise free system possible with hazards of killing someone!

>Responding to Bob's 28-Jul (14:08) query... here are several whys:
>
> a) With the shield grounded at both ends the GFP output is effectively
> short-circuited, thereby compromising proper trip action!
>
> b) There have been cases, where, although properly grounded at the
> source end, inadvertent grounding elsewhere along the cable's path,
> has negated trip action!
>
> c) And, sometimes the double-ground can result in unwarranted (some
> unexplained) trips!
 
John, pls excuse the informality) you are correct!

While the thread covered many "grounding" topics, as does your treatise, I specifically addressed one... metallic-shielded power cables.

Regards,
Phil Corso
 
Top