Odd 7i76 error
- bottleworks
- Offline
- Senior Member
-
Less
More
- Posts: 73
- Thank you received: 0
15 Apr 2014 01:22 #45945
by bottleworks
Odd 7i76 error was created by bottleworks
I'm getting some odd errors on my good mill. I'm getting random comm errors with the 7i76. The machine appears to be functioning normally. This may be related...I'm also getting run time errors sometimes which I'm also trying to sort out. I've attached a screen shot of the comms error. Ideas?
Please Log in or Create an account to join the conversation.
15 Apr 2014 01:58 - 15 Apr 2014 02:00 #45948
by PCW
Replied by PCW on topic Odd 7i76 error
These sorts of errors indicate an electrical noise problem.
It may be they you have noise coupled into the DB25 cable with nearby power cables,
or there is a ground loop so enough HF noise current is being conducted via the DB25
ground that more than ~1.5V noise is being developed from the 7I76 end to the PC end
of the cable.
For coupled noise the best solution is re-routing cables so the power cables
(especially motor and power line cables to/from VFDs and servo drives) are as
far as possible from signal cables.
For ground loops you need to make sure no large noise sources connect directly
to the 7I76s ground or common pins. Sometimes line filters on VFD power lines
or common mode chokes on the motor leads may be required on large noise generators
like VFDs and servo drives.
Another possible ground loop issue is having the PC ground be connect via a distant (non machine)
ground.
Run time errors are probably unrelated but need to be be investigated
( make sure you capture the error to see how big the timing error is )
It may be they you have noise coupled into the DB25 cable with nearby power cables,
or there is a ground loop so enough HF noise current is being conducted via the DB25
ground that more than ~1.5V noise is being developed from the 7I76 end to the PC end
of the cable.
For coupled noise the best solution is re-routing cables so the power cables
(especially motor and power line cables to/from VFDs and servo drives) are as
far as possible from signal cables.
For ground loops you need to make sure no large noise sources connect directly
to the 7I76s ground or common pins. Sometimes line filters on VFD power lines
or common mode chokes on the motor leads may be required on large noise generators
like VFDs and servo drives.
Another possible ground loop issue is having the PC ground be connect via a distant (non machine)
ground.
Run time errors are probably unrelated but need to be be investigated
( make sure you capture the error to see how big the timing error is )
Last edit: 15 Apr 2014 02:00 by PCW. Reason: clarify
The following user(s) said Thank You: bottleworks
Please Log in or Create an account to join the conversation.
15 Apr 2014 04:05 #45952
by andypugh
VFD manufacturers tend to state that you need an input filter (but they rarely shout about it). Some VFDs have filters built in.
They are not very expensive: www.ebay.com/itm/251503495598
(Just search eBay for "Rasmi", as that is all they seem to make, though other manufacturers exist)
The filter prevents noise from the VFD travelling back into the mains and into other devices.
Be careful if you have RCD devices on the power feed, some of the filters have more earth leakage than the breaker are happy with.
Single-phase versions also exist if you are running from single phase.
Replied by andypugh on topic Odd 7i76 error
Sometimes line filters on VFD power lines
or common mode chokes on the motor leads may be required on large noise generators
like VFDs and servo drives.
VFD manufacturers tend to state that you need an input filter (but they rarely shout about it). Some VFDs have filters built in.
They are not very expensive: www.ebay.com/itm/251503495598
(Just search eBay for "Rasmi", as that is all they seem to make, though other manufacturers exist)
The filter prevents noise from the VFD travelling back into the mains and into other devices.
Be careful if you have RCD devices on the power feed, some of the filters have more earth leakage than the breaker are happy with.
Single-phase versions also exist if you are running from single phase.
The following user(s) said Thank You: bottleworks
Please Log in or Create an account to join the conversation.
- bottleworks
- Offline
- Senior Member
-
Less
More
- Posts: 73
- Thank you received: 0
15 Apr 2014 04:47 #45956
by bottleworks
Replied by bottleworks on topic Odd 7i76 error
Fantastic. This is definitely something I can look into, diagnose and isolate.
Just for clarification on those RFI filters: it's for mains side, not motor side, correct?
Just for clarification on those RFI filters: it's for mains side, not motor side, correct?
Please Log in or Create an account to join the conversation.
15 Apr 2014 04:50 #45957
by andypugh
Yes. A common-mode choke on the motor wires close to the VFD can also help, but addresses a different issue.
Replied by andypugh on topic Odd 7i76 error
Just for clarification on those RFI filters: it's for mains side, not motor side, correct?
Yes. A common-mode choke on the motor wires close to the VFD can also help, but addresses a different issue.
Please Log in or Create an account to join the conversation.
- bottleworks
- Offline
- Senior Member
-
Less
More
- Posts: 73
- Thank you received: 0
09 May 2014 14:53 #46749
by bottleworks
Replied by bottleworks on topic Odd 7i76 error
So, just to update this....
The run time delay was caused by the fglrx driver. Apparently, I installed it when I first installed LinuxCNC on that machine. I don't know why I would have done that.
The comms error was caused by interference by the VFD. The error would usually get triggered when the VFD was stopping. The source of the interference appears to be some foolish ground bonding. I've got some test leads clipped on different spots and I'll leave it like that for a month before I will feel comfortable that it's the sole source of the issue. So far, all signs point to it being the cause. Thanks for the help!
The run time delay was caused by the fglrx driver. Apparently, I installed it when I first installed LinuxCNC on that machine. I don't know why I would have done that.
The comms error was caused by interference by the VFD. The error would usually get triggered when the VFD was stopping. The source of the interference appears to be some foolish ground bonding. I've got some test leads clipped on different spots and I'll leave it like that for a month before I will feel comfortable that it's the sole source of the issue. So far, all signs point to it being the cause. Thanks for the help!
Please Log in or Create an account to join the conversation.
Moderators: cmorley
Time to create page: 0.140 seconds