Noise on EtherCAT from servo drivers? Grounding issue?

[badhabit]

Hi,

I'm testing out the electronics setup for a new machine i'm building. Currently i'm testing the axis electronics. I have my servo-drives attached to the drivers, and the drivers connected together via ethercat. This is then connected to a Beckhoff module and then on to my control pc(LinuxCNC) via EtherCAT. I've used short cat 6A S/FTP cables for best possible shielding of the ethercat bus. I've grounded the drivers, the beckhoff modules and the computer via 2,5mm2 cable which again is grounded via the wallsocket. All components are layout out along the table, with about 2m from the motors to the drivers, about 5-10cm between the drivers, about 1,5m from the drivers to the Beckhoff modules and another 1m to the computer. However, when the drivers "kick in"("Voltage enabled" via the controller) the entire EtherCAT bus gets scrambled and loses connection for a couple of seconds, here is a screenshot of the bus just before and just after the drivers gets enabled:

(1st: all is ready for enabling drivers, 2nd: just after voltage-enable, all devices are lost, 3rd: after a couple of seconds, "something" is found again)

The drivers are 3-phase 2,2kW CTB-brand and my guess is that the voltage-enable causes a small EMP which causes the problems. If i only run 1 driver there is no issues...2 drivers works most of the time...3 never works... and i actually have a 4th 7,5kW driver for spindle i haven't even connected yet :-/ It doesn't matter which drivers that i put onto the bus(i thought maybe one was bad, but they all act the same)...

This is going to give me a headache if i can't even power it up on the table...then it will never work when i put everything into a box :-/

So questions:
- How does large VMC's handle this? Do they stage the "voltage enable" so not all drivers turn on at the same time?
- Are drivers seperated in the larger setup's by shields or something?
- Any ideas to what i might try or what i have missed?

Any input will be greatly appreciated!

/Thomas

Similar Threads:

• Grounding Issue
• Just IN- LinuxCNC with EtherCAT Master and digital servo drives
• Stepper motor&drivers,servo motors&drivers,motion control card CHINA factory
• Need Help!- Grounding or noise

[mactec54]

Hi,

I'm testing out the electronics setup for a new machine i'm building. Currently i'm testing the axis electronics. I have my servo-drives attached to the drivers, and the drivers connected together via ethercat. This is then connected to a Beckhoff module and then on to my control pc(LinuxCNC) via EtherCAT. I've used short cat 6A S/FTP cables for best possible shielding of the ethercat bus. I've grounded the drivers, the beckhoff modules and the computer via 2,5mm2 cable which again is grounded via the wallsocket. All components are layout out along the table, with about 2m from the motors to the drivers, about 5-10cm between the drivers, about 1,5m from the drivers to the Beckhoff modules and another 1m to the computer. However, when the drivers "kick in"("Voltage enabled" via the controller) the entire EtherCAT bus gets scrambled and loses connection for a couple of seconds, here is a screenshot of the bus just before and just after the drivers gets enabled:

(1st: all is ready for enabling drivers, 2nd: just after voltage-enable, all devices are lost, 3rd: after a couple of seconds, "something" is found again)

The drivers are 3-phase 2,2kW CTB-brand and my guess is that the voltage-enable causes a small EMP which causes the problems. If i only run 1 driver there is no issues...2 drivers works most of the time...3 never works... and i actually have a 4th 7,5kW driver for spindle i haven't even connected yet :-/ It doesn't matter which drivers that i put onto the bus(i thought maybe one was bad, but they all act the same)...

This is going to give me a headache if i can't even power it up on the table...then it will never work when i put everything into a box :-/

So questions:
- How does large VMC's handle this? Do they stage the "voltage enable" so not all drivers turn on at the same time?
- Are drivers seperated in the larger setup's by shields or something?
- Any ideas to what i might try or what i have missed?

Any input will be greatly appreciated!

/Thomas

You most likely need an EMI Power Filter can you take a photo and what drives are you using

You should not be powering the drives direct from a wall socket, this needs to go through a MCSB the the drives are feed from this circuit breaker the EMI Power Filter would be mounted after the Circuit Breaker

All the Drives Etc need to be mounted on a Ground Plane ( Metal Plate ) this metal plate would have a Ground Bus where all Grounds attach at this one Point ( Star Ground )

More information needed on the servo drives do you have a link

[badhabit]

You most likely need an EMI Power Filter can you take a photo and what drives are you using

You should not be powering the drives direct from a wall socket, this needs to go through a MCSB the the drives are feed from this circuit breaker the EMI Power Filter would be mounted after the Circuit Breaker

All the Drives Etc need to be mounted on a Ground Plane ( Metal Plate ) this metal plate would have a Ground Bus where all Grounds attach at this one Point ( Star Ground )

More information needed on the servo drives do you have a link

Yeah, i'm going to add protection circuit etc when i'm done testing, but right now i just need to verify that the setup works. My wall-socket already have an 3-phase 16A/400V "MCCB" so i don't think i need anymore for testing...

I'm also going to mount everything inside 1-2 Rittal metal enclosures with a sturdy and removable backplate....i've been thinking about putting the drivers in their own enclosure...but right now i'd like to just "dry run it" on the bench until i know how much space i need in the case(s) etc...i do see that a solid metal plate would provide better grounding than a single 2,5mm2 wire...if i keep having issues i'm going to buy a plate of some sort and mount things on for testing....but i would have liked to see it run without this much grounding needed, it worries me a bit...

I've attached the manual for the drives...

Can you recommend any good EMI filters for 3-phase for around 20A/400V/50hz? I'm guessing that i need just one for the entire enclosure, or should i have one for each driver? Anyways, i don't quite understand why i need such a filter to avoid noise on the ethercat bus...i thought they were mostly to be "nice" to any other electrical equiment attached to the mains nearby...

Edit: something like this: https://docs.rs-online.com/b3a9/0900766b813e2a34.pdf or https://docs.rs-online.com/b541/0900766b8142ceb8.pdf ?

Thanks for the input!

/Thomas

[mactec54]

Yeah, i'm going to add protection circuit etc when i'm done testing, but right now i just need to verify that the setup works. My wall-socket already have an 3-phase 16A/400V "MCCB" so i don't think i need anymore for testing...

I'm also going to mount everything inside 1-2 Rittal metal enclosures with a sturdy and removable backplate....i've been thinking about putting the drivers in their own enclosure...but right now i'd like to just "dry run it" on the bench until i know how much space i need in the case(s) etc...i do see that a solid metal plate would provide better grounding than a single 2,5mm2 wire...if i keep having issues i'm going to buy a plate of some sort and mount things on for testing....but i would have liked to see it run without this much grounding needed, it worries me a bit...

I've attached the manual for the drives...

Can you recommend any good EMI filters for 3-phase for around 20A/400V/50hz? I'm guessing that i need just one for the entire enclosure, or should i have one for each driver? Anyways, i don't quite understand why i need such a filter to avoid noise on the ethercat bus...i thought they were mostly to be "nice" to any other electrical equiment attached to the mains nearby...

Edit: something like this: https://docs.rs-online.com/b3a9/0900766b813e2a34.pdf or https://docs.rs-online.com/b541/0900766b8142ceb8.pdf ?

Thanks for the input!

/Thomas

I also would not rule out the EtherCat side of it being faulty also, this should work for testing like you have it, I would not be buying Chinese Drives for EtherCat, there English documentation looks good though

[badhabit]

I also would not rule out the EtherCat side of it being faulty also, this should work for testing like you have it, I would not be buying Chinese Drives for EtherCat, there English documentation looks good though

yeah, there might be a reason why they cost 1/3 of their "western" counterparts :-/ However, support from the manufacturer has been good! Anyways, I just ordered an EMI filter: https://www.mouser.dk/datasheet/2/41..._A-2075941.pdf and some big ferrite core for the outputwires on each driver...i'll see if that helps...but i kinda doubt it'll do much about the ethercat bus :-/

Still, if anyone has any input to the EtherCAT disruptions it will be greatly appreciated!

[mactec54]

yeah, there might be a reason why they cost 1/3 of their "western" counterparts :-/ However, support from the manufacturer has been good! Anyways, I just ordered an EMI filter: https://www.mouser.dk/datasheet/2/41..._A-2075941.pdf and some big ferrite core for the outputwires on each driver...i'll see if that helps...but i kinda doubt it'll do much about the ethercat bus :-/

Still, if anyone has any input to the EtherCAT disruptions it will be greatly appreciated!

EtherCat is normally noise immune, it's good that the manufacture is trying to help, have you Grounded / Earth all the drives and motors

[badhabit]

I solved it. It was defently some sort of grounding issue causing lots of EMI which killed the EtherCAT bus when the drives was enabled. I put all the "noisy stuff" inside a a good Rittal AX enclosure with a proper backplate and all the issues went away...

[mactec54]

I solved it. It was defently some sort of grounding issue causing lots of EMI which killed the EtherCAT bus when the drives was enabled. I put all the "noisy stuff" inside a a good Rittal AX enclosure with a proper backplate and all the issues went away...

That's good to hear, most systems like this need to have a very good Ground system in place and mounting of drives Etc.