Plasma Cutter causes communications lost with 7176.

I have a 5i25 to a 7i76, with the I/O all working now. Thanks to those who helped me. The thread is gone to reply.
Anyhow, everything works great right up until the Plasma Cutter torch turns on. The two wire trigger is controlled via a N.O. contact of a 24Vdc relay from Digital out 0. Once this event has occurred, I loose ALL I/O and the software starts pumping out many communication errors and halts. If I reload, the I/O comes back.

Anyone have an idea of what this might be?
Thanks

I just ran it again. There are two errors: =4 (too many characters) and =13 communications loss??? This is in the EMC software.I pops up three times each.
The program carries on tracing all the paths. I have to stop it manually..
When I did it this time, I completely isolated the Plasma cutter from the table, EXCEPT the relay, which only connects the troch trigger wire. But should be completely isolated from the control system.

I created an opto-isolator to isolate the mesa card from plasma you can order the PCB from here

oshpark.com/shared_projects/flj0c7qC

If you get EMI induced sserial communication errors with a 7I76/5i25 system i usually means one of two things:

1. You are inadvertently coupling noise into the cable between the 7I76 and 5I25
The fix is to route the 5i25/7ii76 cable as far away as possible from noise sources (power wiring, any plasma wiring etc)

2. You have a ground loop and high frequency current is running from the 7I76 to the 5I25/PC ground. One way to find out if this is the case is to disconnect the 7I76 outputs
and re-connect one at a time and see where the conducted noise is coming from
(you may need to manually enable the plasma source for this test)

Fixes for this kind of issue are to either lower the cable impedance between the 5I25 and 7I76 (use shortest cable possible, put PC in same box as 7I76, use a heavy ground strap between the PC and the 7I76 box) or alternately, raise the high frequency impedance of noisy signals connected to the 7I76 (series 47 ohm resistors or ferrite beads can help here)


In either case the EMI is severe enough to generate more than about 1.5V of noise between the 7I76 and 5I25/PC

That sounds logical. I have done everything except that. Was my next step, but did not want to start going down the rabbit hole.
Got too late in the day and getting tired. I stop working on this stuff tired.
I am going to remotely control the relay (move it far away from the 7i76, and keep the torch wires far away from the control box.
By the way... I believe you helped me with my I/O problems a couple weeks back. That thread has disappeared from my view. It was the card (7i76). I did have everything set up correctly. Bought new ones.
I really appreciated the help you provided.
Thank you

I keep loosing submissions in this forum for some reason???
I tried isolating the relay. Did not work.
I tried controlling the Plasma cutter separate from the control system. Did not work. I manually operated the plasma cutter while the motion control did it's thing. Same issue.
I am now going to common earth ground ALL devices. I am in a rural areas so all the grounds are to an earth plate.
I have three switch mode power supplies and the ground for the plasma cutter that I am going to common ground to the plate as well as the ground neutrals for the Vac.
Any concerns you may want to share?

I take that back. I am NOT connecting the DC and AC grounds together, but rather ensuring that the DC and AC circuits individually are properly common grounded.

I do not understand what you mean about the heavy ground strap.
The 7i76 is Din rail mounted and not "apart" of any box. How would you connect that?
I also don't understand what to do with the 47 ohm resistor. What.is it in series ?
thanks

First would see how the noise is getting into the 5I25 --> 7I76 cable

As I mentioned before, one way to do this is do disconnect all 7I76 outputs
and see if you still lose communications when the plasma generator starts.

If you find that some output connection(s) cause the problem you can often add series impedance (the 47 ohm resistors or a ferrite bead) in series with the output that is a problem

If still have an issue with the outputs disconnected .the noise may be coupled in via power supplies (its generally best to use cable 5V rather than an external 5V supply for the 7I76 5V if you have a noise issue) or the noise could be due to routing the 5I25 --> 7I76 cable near cables that radiate
noise

The issue seems to be a power source problem. I completely isolated the Plasma Cutter. No physical connection whatsoever. Not even touching any part of the CNC table. As soon as I pull the trigger, all comms are lost with the I/O.
I am using cable power.
I moved the 7i76 away from any power sources (more than a foot away).
All switch mode power supplies are commonly connected on their -ve terminals.
The only connection between the Plasma Cutter and the Control System, is the power lines. I wonder if an isolation transformer for my 120Vac side would eliminate the affects of the Plasma Cutter? I think I need a way to filter the noise from the Cutter. An input filter/terminator?