VFD Noise Issue - Help

[dsmoot]

I will attempt to state my situation as simply as possible. I am having an RS-485 communication problem when my VFD driven motor is started.

My power supply is a 3ph, 480V generator feeding a pump control panel which is driven by a C-H VFD. My communication link is an operator interface (modbus master) polling two slave devices (VFD drive and gas flow computer). When motor is not started, good communication between master and slave devices exists. When motor is energized with current flow, communication is lost with comm errors. Initially started with non-shielded CAT 5E cable and switched to shielded with no improvement. Total comm length is 10 feet - half duplex.

Incoming power is fed through breaker, harmonic filter to VFD input all located inside enclosure with VFD. All cabinet wiring single conductor (3 ph) between devices. VFD to motor 3 conductor w/o ground. Electrical contractor grounded motor locally to my dismay. Enclosure, VFD and flow computer all grounded with ground rod at panel. Motor located approximately 75 to 100 feet away.

After problem was initially discovered with some components replaced with no improvement the following things were checked.

1. Good grounding from panel devices to ground rod connection. Have not performed ground resistance check to soil. Typical Wyoming soil - very dry.

2. Communication shields between operator interface and VFD grounded at VFD ground terminal and VFD to flow computer grounded at flow computer. Continuity to ground rod from both locations good.

3. Have kept communication cables as far from motor leads as possible in panel given that motor leads and comm cables enter VFD but not at same entrance.

Any ideas, questions or suggestions. Any input appreciated.

 

[Skogsgurra]

Had that with a MiniMap comm link. Had to connect node grounds with heavy gauge copper close to communication cable to re-establish communication when VFDs were running.

Reason is that your grounds - how low ohm they ever are - are no good HF grounds so that your common mode range is being violated. Put a 10 or 16 mm2 (don't know what AWG that is) copper wire close to (strap together with) the RS485 cable and bond well at ends.

 

[DickDV]

Can't guarantee that it will fix your comm problems but you really need a continuous copper ground between the motor frame and the drive frame.

The more high frequency pulse current you force into the earth grounding system, the more likely you will have ground noise problems, often in unexpected places remote from the motor/drive location.

 

[buzzp]

I agree with DickDV.
Also sounds like your shield is grounded at both endsof your comms cable, generally only grounded at one end. Also, do you have any terminating resistors on the network?

 

[jraef]

"All cabinet wiring single conductor (3 ph) between devices. VFD to motor 3 conductor w/o ground."

You do not mention the output wiring method from the VFD to the motor. Is it in metal conduit? If not, it needs to be, or better yet use shielded cable. Even if it is in metal conduit, is there a significant distance from the VFD output terminals to the conduit entry? Any unshielded output wires will be acting like an antenna, the more exposure they have the worse it is.

Also, did your ground your CAT5 cable only at one end? A common mistake is to ground both ends of the shield. You WANT to do that on the power output cables, but you do NOT want to do that on the signal wires.

"Venditori de oleum-vipera non vigere excordis populi"

 

[weh3]

I would be concerned about the extra ground rods. Multiple grounding paths can create not only common mode noise from currents on the grounding conductors, but an unsafe situation that can defeat your GFP. The extra currents are especially likely if there is some distance between the service and the switchgear.

Make sure any grounded circuit conductors (neutrals) are grounded only at the service entrance through the main bonding jumper.

William

 

[dsmoot]

The VFD to Motor cable is a 3 conductor w/o ground and non-shielded. I do not believe the motor cable is in a metal conduit. It is not armored and not shielded. The conductors are exposed for approximately 8 - 10 inches once it enters the enclosure before it enters the drive. The motor is approximately 75 to 100 feet from the drive.

The RS-485 cables are grounded at one end only in each segment, although the operator interface to VFD shield is grounded at a VFD ground (not power ground but a ground terminal on the frame).

 

[hoy7t]

ALWAYS use shielded 3 conductor cable on ALL VFD wiring.
ALWAYS ground the shield on both ends.

You can use shielded cable for your modbus coms if you want to. it will not hurt. but the real issue is VFD cabling.

 

[ScottyUK]

Can you fit a galvanic isolator local to each node? Or use optical fibre? This is the other solution to the brute force solution of massive bonding conductors.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[jraef]

"The VFD to Motor cable is a 3 conductor w/o ground and non-shielded. I do not believe the motor cable is in a metal conduit. It is not armored and not shielded."

Methinks therein lies the root of the problem. You have one nice local FM radio station, broadcasting the audio track from the "Polar Bear in a Snowstorm" show.

Also, no ground conductor out to the motor? I take it your VFD has no Ground Fault Protection then? If it does, it isn't going to work.

"Venditori de oleum-vipera non vigere excordis populi"

 

[PowerfulStuff]

I'd go with the galvanic isolation - Black Box or RS Components have a good range. This will fix most conducted noise issues. The unshielded motor cable may be no good for the neighbours, but probably won't help with the noise inside the panel anyway.

 

[hacksaw]

Having been around a lot of motors in the field, I am surprised that you remark that the "Electrical contractor grounded motor locally to my dismay. "

All motors must be grounded. Isn't that part of the NEC?

 

[bigbillnky]

The situation with the grounding configuration you mentioned is very serious. Most people do not realize that a ground rod is a very poor means of grounding, and should a fault to ground develop at the motor, in all likelihood the overcurrent/shortcircuit protective device will not open adaquately to clear the fault. There is a big distinction between "grounding" and "bonding", and it sounds like your contractor needs to study up on these subjects.
Best wishes


Bigbillnky,C.E.F.....(Chief Electrical Flunky)

 

[mc5w]

You need a 4-wire cable from your VFD to your motor to provide an equipment ground. The grounding for the motor is a Code violation. Also put the cable in steel EMT, Rigid conduit, or liquidtight flexible metal conduit that has copper wires rolled into the interlocking steel tape. The aluminum kind of liquidtight flexible metal conduit does not have thick enough aluminum for the aluminum to act as a magmetic shield.

Aluminum is slightly magnetic with about 3.2 times the permeability of a vacuum. Rigid aluminum conduit is O.K. for your communication line as long as the power wires are in a steel conduit or wireway. If you drag a magnet through aluminum machine tool chips the smallest pieces will magnetize and stick to the magnet. Searching for a small steel part in aluminum machine tool chips will teach you this.

The better communication cables have aluminum foil over any copper shield because aluminum foil stops any magnetic field that leaks past the copper shield. This has been shown to work better than copper foil but not as well as the copper clad steel shield that is in outdoor telephone cables. However, copper clad steel foil is less flexible and more expensive.

 

[dsmoot]

I am very appreciative of all the comments provided. This has certainly given me many ideas for solving the problem and will be checking these items out on this drive. I have one other confusing input into this issue.

The contractor has installed a second of these panels in a location approximately 1/2 mile from the original. The panel was constructed exactly like the first unit. The contractor installed the panel in the same way including grounding the motor at the motor (ground rod) and grounding the panel with a ground rod at the panel.

The real head scratcher. Everything appears to be communicating on the RS-485 fine. Could this be a difference in ground conditions at this location vs. the other location? In no way am I trying to disprove any previous comments given but does this make any sense to anyone?

 

[ScottyUK]

mc5w,

I think you're confusing magnetic shielding and electrical shielding. Aluminimum makes a great electrical shield, as do most earthed metals in the form of a tube or braid. Magnetic shielding is damned hard work: mu-metal is one possible material. Aluminium isn't anything remarkable in terms of magnetic shielding, nor are copper, brass, or any of the other common electrical non-ferrous metals. Distance is the best solution if possible, then other countermeasures like twisted conductors and mu-metal screens can be applied where distance can't be achieved.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[jraef]

dsmoot,
Absolutely. The soil could be drier at the bad location, or it could have a different chemical makeup etc. etc. etc. That is the point behind bigbillnky's comment on ground rods not being the best.

I was at one remote jobsite with a portable rock conveyor system, and the ground was so dry and dusty that we had no ground with a 12ft ground rod. We ended up digging a 20ft hole with a back hoe, tossing in a steel plate with a cable attached to it, and had everyone "relieve themselves" into the hole before burying it again. Worked like a champ (albeit probably only temporarily).

"Venditori de oleum-vipera non vigere excordis populi"