Some of the most common problems I see are the ill effects of EMI (electro magnetic interference). AKA electrical noise. The adverse effects EMI can have on a system run the gamut and are too vast to list them all but I'll give you a few examples.
External E stop requested when no e stop switch or button is activated.
Limit switch triggered when the axis are no where near their limits.
Referencing or homing fails (most of the time a limit switch triggered error occurs while homing).
When my plasma turns on (you name it). When a plasma fires it produces an enormous amounts of electrical noise.
When I turn my spindle on (you fill in the blank, axis jumps, limit switch triggered, etc.)
I get asked almost daily how to solve this. This is not a software problem, it is a design problem and the only thing you can configure in the software is debounce (aka noise filtering) in general config. (also in some plug-in configs.). This is a work around at best to compensate for inferior design. It is to the electrical system what backlash compensation is to the mechanical system. It is a band aid at best. Just as band aids are meant for minor wounds debouce is meant for minor imperfections in design. Heavier bleeding requires stitches and severe noise issues will often require re-designing. Obviously the best time to try to avoid these problems is in the design phase of any project. Hopefully you knew this when you started your project. If not, it will not change the steps necessary to build a solid system and you likely have some rework to do. No sense sulking about that either, whats done is done. Best to dig in and get it sorted now. You have work to do and there is no silver bullet in the software to fix it for you. My intent is not to make you feel good or bad about your mistakes but rather to give you the information to fix them or hopefully avoid them all together.
Here are a few broad suggestions:
Physical separation or distance between low voltage, low current signals/sources/components and high voltage, high current signals/sources/components should be a goal throughout your design.
If low voltage and high voltage wires must cross, try to cross at 90 degrees. Running such lines parallel to one another is a no no.
Shielded cables are not optional in many instances. Shielded cables are more expensive but do you want it to work or almost work?
Google Faraday cage. Use this method when needed. Try to keep a conductive, grounded partition between low noise and high noise components.
If you must run signal wires outside of your low voltage cabinet I strongly urge you to use some type of optically isolated converter to take the signal from 5 volts to 24. This will usually give you 5 times the noise immunity right off not to mention the positive benefits of the optical isolation. I have used opto 22 modules quite a bit but there are all sorts available to choose from.
Star grounding.
Conditioners, not just for shiny hair!
These are a couple of great videos on switch bounce. The first also covers EMI a bit.
https://www.youtube.com/watch?v=Nj-Q8FQxHhUhttps://www.youtube.com/watch?v=4ZMiKMUec9oAll of this and more is covered in more depth at the following links. They are not very long and very informative. To skip this step would be shooting yourself in the foot.
http://www.cnccookbook.com/CCCNCNoise.htmlhttp://openbuilds.com/threads/electronics-101.693/http://www.machsupport.com/forum/index.php/topic,16323.0.htmlhttp://www.panduit.com/ccurl/371/589/journj-ar005_-en-p.pdfThis is a great link that covers a lot of ground in a single page.
http://www.autodev.com/files/downloads/controllers/bh01.02-adi-good-wiring-121906.pdfAnd a few other good ones I found.
http://www.controldesign.com/articles/2008/reducenoise0812/?show=allhttp://www.mccdaq.com/PDFs/specs/Noise-Reduction.pdfhttp://machinedesign.com/article/top-tips-on-squelching-electrical-noise-0619Thanks,
Brett
Guest
Show Posts
Seriously though, wait for confirmation for anything I said........... I'm beat.