VFD Com Problems

[richleva]

Hello everyone,

I need your advice please, all 3 VFD Benshaw RSi S4 are working perfectly until recently. VFD #1 (that is mostly use) has been faulting up on communication loss.

-I should also mention that our customer recently had a power outage
-all VFD's are configured the same
-this drive was also known to fault on Dynamic brake overload (not sure if this is relevant ?)
-we're using Modbus to communicate
-The VFD #3 has recently been powered off.
-The VFD #2 Is working perfectly

One thing that I noticed is that when the VFD drive is running, my PLC has problems pooling the drive, Using message instructions in RSLogix, I get Message errors ("message timed out in local processor"). However 10% of the time pooling the drive is successful.

But while the VFD is Faulted or stopped. The communication to the drive is excellent!! No errors?

Has anyone ever seen a problem like this? What do you think could cause this? Could there be damage to the communication hardware of the VFD caused by the power outage? Or interference?

If VFD #3 is powered off, can this affect the Modbus line since there all connected together?


Thanks everyone!!

 

[Ken Roach]

When communications to a VFD is unreliable during VFD operation, it generally does point to induced noise on the communication network.

What is the physical layer of your network ? You say you are using Modbus, but is it connected over RS-485, or Ethernet, or something else ?

Modbus isn't a native protocol for the ControlLogix, so there must be some sort of gateway or converter. What sort of device is that ?

Having a device powered on or off isn't supposed to affect an RS-485 network, but some devices might behave differently because of powered bias circuits.

In general when a device is powered down, the effect is just on the polling logic if the Modbus Master is waiting for that device to time out before moving on to polling the next device.

 

[richleva]

Hi Ken,

Thanks for the reply. Yes, we're using RS-485. There is a HRT-710, but I always thought that was only for the flow meters.

Keep in mind that I did not build or design this system, I'm not 100% sure to how everything is connected. I'm just trying to finish what someone started..

Thanks again Ken

 

[jraef]

When communications to a VFD is unreliable during VFD operation, it generally does point to induced noise on the communication network.

What is the physical layer of your network ? You say you are using Modbus, but is it connected over RS-485, or Ethernet, or something else ?

Modbus isn't a native protocol for the ControlLogix, so there must be some sort of gateway or converter. What sort of device is that ?

Having a device powered on or off isn't supposed to affect an RS-485 network, but some devices might behave differently because of powered bias circuits.

In general when a device is powered down, the effect is just on the polling logic if the Modbus Master is waiting for that device to time out before moving on to polling the next device.

I agree, if the comm problems only happen or get worse when the VFD is running, it's a noise issue. It could be common mode noise (coming in on the ground plane) or it might be EMI/RFI bleed over into the comm wires.

How are the comm wires run? Did they use shielded cable? If so, is the shielding properly grounded on only ONE end?

How are the VFDs wired to the motors? Are the cables in grounded steel conduit? If not, did they use shielded VFD power cable and ground the shields on BOTH ends?

 

[richleva]

Hey guys,

The first thing I did was power up VFD 3 again, and it worked!!
But the I believe there is still a problem since pooling the drive is not always successful. Probably a 60-70% success rate for all three drives? but enough to keep them going. I pool the drives every 4-8 Seconds, I set the com timeout settings to 30 Sec on all 3 VFDs. As long as the pooling doesn't fail four times in row...

Perhaps the EMI or inference is also part of the problem.

Thanks to everyone !