False E-stop triggers. Need help finding and eliminating noise.

Anytime someone asks me about building a CNC machine I tell them absolutely do not. Buy a kit that already has the bugs worked out and a support system in place to troubleshoot. Learn from my mistakes!

My DIY machine has been pretty much unusable since installing home switches with my G540 system. I will run a program and get an error message in UCCNC that reads "Ext E-stop triggered". I am running a Bosch 1617 router with an 80-20 frame and gantry. I figured this false E-stop was a noise problem so I been trying to add measures to eliminate the noise. I believe the noise is coming from the 1617 because if I drop feeds and speeds along with depth of cut I can reduce the frequency of it. But once the router starts to chatter or plunge into a heavy cut, E-stop.

I have sheilded wire on all my stepper motors (minus about 2" between where I soddered the wire and the motor casing, is that a problem?) tied to pin 3 on the Db9 connector. Shielded wire on all my limit switches tied to the ground inside my electronics enclosure. I hadn't grounded my machine bed or gantry before so I did that this weekend running ground lines to several points on the machine and did continuity checks to make sure everything is grounded, especially the body of the 1617. After all this I still get E-stop triggers.

I am thinking maybe I have a ground loop? I dont really know what that is except for reading about it on various forums and hearing its a bad thing. I have all my electronics in a 8"x10"x6" metal enclosure (AC inputs, 48V DC power supply, 12V DC power supply, G540, aviation plugs for limit switches, etc.), can I be generating noise inside the box? If that is the case, how do you explain the 1617 bogging down and immediately getting a false trigger?

Any advice would be much appreciated. Here is a diagram I drew up on how I have wired things together inside my box. Maybe it will help see if I have something hosed up. https://imgur.com/KHX9mrb

Tonight I plan on taking limit switches out one by one to see if I can run a full program without any false triggers. If you have other recommendations on troubleshooting steps to locate or diagnose the noise source and or where the false trigger is coming from please let me know. Also, it might be pertinent that I am using a laptop running UCCNC and UC100.

Hi,

could you add pictures of your control box? Is the box connected to the router so that the box vibrates? Do you have floating Pins?


PS: This is how I ground the wire shields:

The latest test version of UCCNC has signal debounce settings to filter out noise issues.

Quote:

---

Originally Posted by ger21

The latest test version of UCCNC has signal debounce settings to filter out noise issues.

---

Thank you for this. I will go update asap. I had some success last night. If I remove the Z-axis limit switch entirely then I dont get any false triggers. Hopefully the debounce setting will help to eliminate the issue.

Quote:

---

Originally Posted by Jakeller449

Thank you for this. I will go update asap. I had some success last night. If I remove the Z-axis limit switch entirely then I dont get any false triggers. Hopefully the debounce setting will help to eliminate the issue.

---

A picture of your box would help :)

Quote:

---

Originally Posted by Jakeller449

Thank you for this. I will go update asap. I had some success last night. If I remove the Z-axis limit switch entirely then I dont get any false triggers. Hopefully the debounce setting will help to eliminate the issue.

---

A small cap across the input might help too.

A quick edit. Years ago I worked on an injection mold machine that would cause the contacts on rather heavy duty mill switches to bounce. Literally the contacts bounced even when held closed at the start of the cycle. It is possible that the additional vibration from running both axis is cause the switch to bounce even when it isn’t operated. This can be bad enough to overcome any filtering an interface might have. It is worth taking a look at the switch itself, it’s mounting and even the quality.

This is a distinctly different issue than electrical noise because there is a real change of state even if very brief. Now this doesn’t mean an electrical noise issue does exist just that we have a different but very similar issue. By the way even poorly mounted prox switches can cause similar issues but this is largely a mechanical switch issue.

Are you using a ground reference to the star point for all DC P.S. commons?
Al.