Noise Powerflex 753

Sounds like you have noise interference...

..."EtherNet/IP option module for network communication and how to use the module with ... parameter [Drive Ref Rslt], which is parameter 883 for PowerFlex 753 drives and ... environment, this product can cause radio interference"...

https://literature.rockwellautomation.com/idc/groups/literature/documents/um/750com-um008_-en-p.pdf

Actually, the problem seems to be a reluctance to pick up the phone to the original equipment manufacturer, and talk it through with someone there.

Rockwell has comprehensive instructions on bonding, grounding, shielding etc of both drives, leads, signal cables, etc.

AC drives are VERY NOISY. Follow the instructions to a "T".

I have used fiber in problem areas.

Double check the shield wires on your cables.They should be grounded on ONE END ONLY,otherwise,you will get lots of noise,however there are certain cases where a driven shield is used.Check your specs.

You might try an input line reactor, 3% should be fine. I’ve gotten in the habit of always installing them, as it helps the Rockwell Powerflex drives carry through power dips and utility upsets the same as an electromagnetic motor starter.

You may need to install a shielded output cable for your motor feeder. It really depends on how the motor leads are segregated from your network wiring. I’ve never had an Industrial Ethernet/IP problem traceable to VFDs, to be this bad, it points to a severely poor installation practice.

And this experience with various installs of maybe 70 or 80 Pflex in groups and singly, no shielded motor leads, do have input line reactors, minimal segregation of network & power, cable tray with divider, nothing over the top. Aluminum conduit, even.

Practice in the USA is typically steel conduit, and if the joints are well made provide fairly decent "shielding", but can also provide ground loops.

Other jurisdictions have other rules - and each system has their quirks.

I suggest a DV/DT filter for your application These are LRC devices,not simply a choke. .In addition to suppressing noise,it can increase motor life.Many companies sell them,so Google for your best source.

It's most likely a problem with the routing of your DeviceNet cables, probably too close to the VFD output conductors.

That would have the deviceNet "system" interfering with itself.

Surely you meant the PLC control cables too close to the VFD output conductors.

What sort of Ethernet cables are you using for the PLC?

CAT 5, 5E , 6....

Unshielded, shielded, individual pairs shielded or overall shield.

Try just changing the existing cable for one with a better spec. and re-routing it.

If you need to change to shielded cables check that the RJ45 sockets provide shielding contact at at least one end.

A bit of a segue: there seems to be an obvious product missing from all the sites I've just checked: an RJ45 coupler with a male shielded connector at one end, a female shielded connector at the other end, and, a separate earth connector (or pair of earth connectors to daisy chain several together). The shielded male would only extend the shield as close to the original socket as possible.

He said he is using DeviceNet, not Ethernet. DeviceNet is a different industrial network, so I'm not sure why you would say it would be interfering with itself? Some types of DeviceNet cables are poorly shielded (some not at all), so they can be interfered with by the output cables of VFDs very easily.

"We have a PF753 on one of the motors, control is done by devicenet, In the same control room we have quantum PLC that we use ethernet iP to our scada system.

The problem we have is that only when the motor is running we lose communication with the PLC, the problem is intermittent. At lower Hz problem is less. We have a sine wave filter on the load side of the PF753. The PWM is set at 2KHz,"

My understanding is that DeviceNet is being used to control the VFD.

But that the thing being interfered with is the Quantum PLC which is controlled via Ethernet Internet Protocol.

Have I got it wrong?

I see now, but it’s ambiguous. I interpreted it to mean when the drive runs, the drive looses its connection to the PLC. The Op actually worded it poorly. He is the only one that can clear it up.

Don't worry - he never came back.

He actually means lost communication with their SCADA, when drive runs. A good bit of detective work to get that far.