VFD/Motor Taking down CANOPN network

[KCRider]

We have 3 Square D VFD's (10, 15, and 40 HP) on a CANOPN network connected to a PLC. Whenever we start the 10 or 15 HP motors, our CANOPN network drops out. The drives begin to start up then fail. The 50 HP VFD/motor does not exhibit this behavior.

If we disconnect the leads from the VFD to the motor and then start the VFD, the network continues to operate. This is the case for either the 10 or 15 HP motor.

If we disconnect the CANOPN cable that connects the drive to the network, then start the VFD in HAND with motor connected, the CANOPN network drops out.

We've considered everything from line reactors to shielded motor cabling. The problem is that the 50 HP unit does not exhibit this behavior. All 3 VFD's are in the same enclosure. The motors are within 15 feet from the VFD's.

The hidden clue is that the CANOPN network still drops out even if the network cable is detached from the VFD. Any help would be much appreciated.

KCRider

 

[Skogsgurra]

Get an EMC guy to have a go on it. It is no good to just 'consider' what the problem can be. You need to get someone with experience and the right equipment to do the job.

The size of the drive has very little with the level of emitted EMI to do. So draw no conclusion from the fact that the 50 HP drive gives no problems.

This kind of problem is so common that most of us have to deal with them at least once each month.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ScottyUK]

The size of the drive might well influence the carrier frequency used by the drive.


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

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

 

[FreddyNurk]

You can, if the carrier frequencies of the drives are all close enough, get what appears to be a 'beat' frequency as all the clocks are close enough, but the carriers move in and out of phase.

That said, its only likely to present an issue if you only see the problem when more than one drive is running.

In terms of trying to fix the problem, Gunnar is right. The only thing I can think of that would get close is a scope and a current clamp to see just what noise is floating around. Of course, the scope by itself won't solve the problem.

A former employer had an issue with overcurrent on a particular conductor, it melted the insulation on a few units before the problem was finally identified. Turns out the current at the carrier frequency was about 4 times the 50Hz fundamental, and it wasn't that apparent until the scope came out.

 

[MAGTiger]

Larger drives are more likely to have internal bus magnetics. than smaller ones. I suspect the 50hp has either a DC bus choke or internal 3 phase reactor. The smaller drives probably have no bus magnetics. You should be able to determine this through the drive spec or manufacturer. If that's the case add a line reactor to each of the smaller drives. I'd go with 5% since you already know you have a problem and don;t oversize the current any more than necessary.

Neil

 

[Skogsgurra]

It is very seldom the harmonics on the input that cause problems with busses and other communication systems. It is almost always the HF from the PWM going to the motor and then reactors on the input don't help.

No use to speculate. Get someone with the right equipment and experience to sort it out for you.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ozmosis]

KCRider
I agree with Gunnar that this is a very typical RFI (Radio Frequency Interference) issue.
High Frequency noise will be polluting the serial communication signal and the likely issue will be the installation and probably ineffective bonding of the motor, VFD and any other connection you have in the vicinity of the VFD.
By the sound of it, the bonding is not working too well and the 'noise' is going into the serial comms connection and affecting your canbus waveform.
Put a scope on your canbus signal and look at the noise on the signal.
The VFD acts like a major emitter of HF noise, they all do. The trick is to ensure the majority of this HF noise goes to ground through low impedance conductors; braid, flat straps, 360deg termination of the shield, metal gland nuts on the motor if the screen is terminated in the motor gland, segregation of power cables from control cables, and so on. The VFD manual you have will/should explain all this.

Do not start spending money on expensive RFI filters (yet). Get your installation right first or you will be wasting your money.
I'm guessing you are in the US with your use of HP and the "SqD".
The laws and regulations on EMC (ElectroMagnetic Compatibility) are not as stringent as in Europe and Australia but the problems are the same..

 

[KCRider]

Thanks everyone for some really good suggestions. At this point, I am inclined to agree with Gunnar's and Ozmosis' assessments of the situation. The smaller Square D VFD's are not as robust in their EMI attributes as the larger unit, so RFI noise is probably the culprit.

Disconnecting the CAN-bus from the smaller VFD's had NO EFFECT on the network dropping out. In other words, the network continues to drop when the drives are energized even with the CAN-bus cable disconnected! - more support for the RFI theory.

We'll be taking more readings on Monday, and I'll report back on what we found.

Thanks again.
KCRider