Wiring limit, home, an E-stop switches

[gumball]

I have my machine running with a plain Jane breakout board, problem is I keep getting transient signals.

If tightened up every thing I can and still get a quick flash on the diagnostics screen, so I don't think loose connections are the problem.

Of course the darn thing just stops and when I hit reset it happens again.
I used unshielded 2 conductor and separated it as far as I could from the motor wiring, still get stops...

Will a optical isolation board (expensive)stop the transients or should I rewire with shielded wire(cheaper).
If I have to rewire I can make sure the wires don't go near a motor wire.  And maybe use roller switches instead of the cheesy switches I am using now...

[ostie01]

Hi, I think you will need sheilded cable for low voltage signal.

Try to hook up a small capacitor (.o1uF) between faulty pin and ground.

Also, try to increase debounce setting in the General Config page.

Enter 2000 as a value and see if it makes any difference.

Jeff

[gumball]

I changed the debounce value to 2000, I ran a test program and it worked without a problem, thanks...
I will also rewire the switches with shielded wire and I''m going to move the break out board a couple feet away from the driver and computer.  I think having the breakout board sitting next to the driver box is also part of the problem. Thank again.   

[HimyKabibble]

Input signals, like home/limit switches, probe inputs, etc. should work very reliably without having to put huge de-bounce values in IF you properly shield the wires (use shielded cable, with the shield connected to ground at the BOB end ONLY!), AND put a stiff pull-up resistor at the BOB end (220 ohms or so to +5V).  Being near the motor drivers should make no difference if things are properly shielded and grounded.  Mine are all inches from a 3HP VFD, and four servo drivers, and I run my encoder cables alongside the motor cable for their full 12-foot length.  Never a problem.  It's all in the grounding and shielding.

Regards,
Ray L.

[Fastest1]

Ray, doesnt your resistor do the same as a debounce setting? I was under the impression that once all the wiring was cleaned up and properly drained possibly no resistor or debounce would be needed. I am not sure why I thought/think the resistor has a time value also.

[Jeff_Birt]

I was under the impression that once all the wiring was cleaned up and properly drained possibly no resistor or debounce would be needed.

Nothing could be further from the truth. No matter how well you plan your wiring, shield it, etc, etc there will always be some noise present. That is why there are so many different ways to deal with it. Some things, like mechanical switches will always need some debounce because the contacts 'bounce' as they open and close.

I always like to describe the situation as building lines of defense. Your wiring layout, and choice of shielding is the first line of defense. You then move on to any filtering or isolation that your I/O boards provide, and finally your left with any software settings there might be. All of these items are important, none of them are mutually exclusive.

[HimyKabibble]

Ray, doesnt your resistor do the same as a debounce setting? I was under the impression that once all the wiring was cleaned up and properly drained possibly no resistor or debounce would be needed. I am not sure why I thought/think the resistor has a time value also.

No, the resistor and de-bounce serve completely different purposes.  The resistor lowers the impedance of the circuit, requiring much stronger interference to be able to influence the signal level.  The stiffer the pull-up, the higher the current required to pull the line down, and the faster it will pull itself back up.  That higher current also helps keep the contacts in the switch clean.  De-bounce is a crude means of removing noise, and slows down the signal, adding time-delay from when the switch changes state to when the software "see"s the change.  A de-bounce of 2000 adds a LOT of delay.  Adding a capacitor on the input has more or less the same effect as de-bounce.  With noise, it's always better to prevent it from getting into the system in the first place, rather than try to remove it after it's already there.  I've never had to use de-bounce at all.

Regards,
Ray L.