Debugging Limit Switches

I had a EMI problem when I got my XYZ probe. Now its not the probes fault it was mine. When I first built the machine I was aware of the emi of limit switches so bought opticouplers to isolate them. I had not had one problem with interference. When I first got the probe after I was done zero'in my part I would place the probe on the table next to the machine. Well I started having trips here and there. I would have job here and there not finish. Then the machine would finish a job. Then on one job it couldn't even travel a inch and then it would say it had to re home for safety. Then I looked over and realized I had thrown the probe on the table and the little magnet had attached its self the PSU of my 500w spindle. Ahhh! So now I place it far away from any power supply. I put it goes in a drawer on the workbench. I have not had a emi trip since. Just thought I would share, something small like a habit of where I placed a piece of equipment was causing the trip. And I do believe it was tripping through the probe cause I still have the opicouplers on the limit switches.

 

Good catch. And a note to self to remember to remove all USB cords not in use.

@DarkPenguin nice glad you found something that worked.

 

Oh man if you want to watch some YouTube videos with static, google " static electricity gas station explosion" Before I touch the pump I always touch ground to discharge the static. What eye opener.

 

Can you still home with soft limits only? I switched to optical switches and had it almost all set up and running after I expanded my machine (workbee clone) to a 1510. Everything was/is great except for my Z switch (named after my youngest son, who is notoriously contrary) which insisted on being ON (according to Control troubleshooting) when it wasn't triggered. When I triggered the switch, it showed OFF in Control, meaning the switch was operating albeit backwards. By the way, this was still when that switch was an Openbuilds Extension roller switch with the filter etc. Convinced that the internal circuitry was corrupt, I clipped out the old roller switch and soldered in a new one, this time jumping the wires in order to make it NC, even though it was still wired in parallel. IOW, the switch was now ON until it was triggered. This fooled Control and everything worked peachy but required that I set $5=1 for the system to operate. Figuring everything would sort itself out when I replaced it with the optical switch (I had to print a custom mount and trigger finger before I could do that) imagine my horror when after running the 12 feet of wire through two cable chains and the final sheathing that the optical switch LED glowed bright green when I plugged it in, indicating it was triggered. Furious at the thought that I, despite meticulously connecting, soldering, and double shrink wrapping each and every joint of the four meter-long dedicated (but unshielded) cable, had shorted one of those connections, I pulled all the wiring back out, and ran a brand new three conductor cable the whole length with the same result. I tested the end stop with a short cable plugged into the Z limit socket and it worked fine. Apparently, the twelve feet of unshielded wiring was picking up enough EMI to trigger the switch? I've never had any EMI problems with the smaller version of this machine before despite running my router cord with the all the other cables? Will an optoisolator on the limit switches fix this?

 

Best solution to emi problems is always to identify and remove the source. If you have mains (or worse - vfd or spindle) wiring close to and parallel to low voltage wiring you have created a capacitor - which is a low resistance path for AC current. Physical distance is the best cure, shielding can only do so much.
Alex.