Faulty Limit switch warning??

[Barefootboy4]

For some reason I'm getting a warning on my Mach 3 which says "limit switch activtated" . It's a faulty warning. No limits are being hit. All conections are good. Does anybody have any ideas or heard of this happening on Mach 3??
It stops the program run. The machine has been running fine for months now with NO problems. Could the vectric 3D be a problem?
Any ideas out there??

[ger21]

Noise on the switch lines. Some have fixed it by installing capacitors at the terminals.

[ZOB]

Had same thing happen to me, it was my Z limit switch went bad, replaced it and no more false triggers. bring up the last screen in MACH 3 that show the input LED's and tap/play with the switches.

[Barefootboy4]

You could be right about this. It just started for some reason. The machine has been great. I'll check all limits and get get back to you on the results.
Thanks for the idea. I found if i run the program with the off and just run air it works fine. If I mill wood with the router it has a limit alarm. Your idea of a bad limit makes total sense. The vibration could cause the problem
Thanks again
Steve

[brtech]

It sounds like noise.

First up is shielded wire. If your wire is shielded, then do you have a pullup resistor on the circuit? Your BoB may provide this, or you may need to add one. It goes between the input and + of the power supply for the switches. Then set some debounce in your general configuration, 1000 is a good start. Then capacitors (try .1uf between the input and ground).

It CAN be a bad switch, for sure, but you should have shielded wire and a pullup resistor anyway.

[Barefootboy4]

I have had problems before so I ran a two conductor Shielded wire with a seperate ground and grounded each end of all 6 limits. Still having a problem. I also changed our 1 limit on the Z axis. Does anyone have a drawing for fixing this problem. I'm not grasping this "PULL UP RESISTOR " idea.What kind is it, Where does it go. I think this will solve the problem if I do it correctly. Ity all works fine until the 3.25 HP porter cable router loads up. The power cable for the router runs in the track with everything else. Could this be a problem?
thanks for the help.
PS: Should I do my limits in a STAR wired instead of a series setup?

[Al_The_Man]

Are you going into the parallel port direct with the switches?
Did you just ground the shield or did you ground the common conductor of the switch?
See this thread.
http://www.cnczone.com/forums/phase_...se_issues.html
Al.

[brtech]

Start with shielded wire. Do that first.

Separate wires to each switch probably didn't help you. It didn't hurt, but it did nothing.

An input needs to be in one of two states ("high" or "low") with fast transition between them. "Low" is close to 0 volts. "High" depends on the input. In most BoB's, it's a couple of volts, but it could be 3/4 of the power supply voltage.

One way to get a good low is to connect one side of the switch to ground, and the other to the switch input. That's great, and most switches are wired that way.

But now, what makes the input go high? The switch is connect/no connect. Typically, the switch is normally closed, which means in this case it's normally connecting the input to ground. That is fine. What happens when you trip the switch? The input is connected to ........ nothing!

Not good. You want to connect it to some positive voltage.

So, what you do is, connect the input to the power supply voltage through a resistor. A value of, say 220 ohms is pretty good.

Now, when the switch is connected (not tripped), you have the resistor trying to pull the input to the power voltage ("pull up"), but you have the switch pulling it down to ground. Since the switch is metal, the wires are metal, and ground is really ground, the switch "wins". What this means is that there is a current through the resistor, which has one end connected to power + and the other connected to power - as long as the switch is closed, which it is, normally.

When you trip the switch, the connection to ground is opened. Now the pullup resistor is pulling the input to power +, and nothing is pulling it down. Voila, a "high", and a good high it is.

So that's what the pullup resistor does, provide a good "high" voltage when nothing else is providing an input, meaning when the switch is tripped.

WARNING: although most BoBs allow inputs to go as high as the main power supply, SOME of them only allow it to go up to 5 volts (although I thought I read a spec sheet that limited the input to 12 V, but I can't find that now). That means the pullup has to go to a +5V supply, not the motor supply for those BoBs.

You connect a 220 ohm resistor from power + (or +5 if your BoB won't allow an input to go over 5 volts) to the input. You can do that at the BoB. One resistor per input, no matter how many switches.

Many BoB's have the pullup resistor built in. Check your documentation.

[Phife]

Just add 400-1000 to the debounce setting in Mach3. Should fix that issue right up.

[brtech]

Adding debounce fixes the symptom, not the problem. It seems to work, and very often does, but because it doesn't fix the problem, you can get in trouble at a time you least want it (in the middle of a big job for example).

Better to fix the problem. The problem is noise; get rid of the noise.

Debounce is valuable for a different problem - mechanical switches don't make clean transitions from off to on or on to off. Debounce, with a small value, is the appropriate solution for that.

[Barefootboy4]

Guys, Thanks for the ideas. I'm going to pick up a resistor tomorrow and create a commom STAR ground for everything and install the resistor and try it out. Thanks for the help and I will get back to you. Just to make sure I don't do damage I have a Keling C10 breakout. This has a standard 5VDC input to the Board. You are saying to add a SEPERATE power supply to the limit circuit?? with the Pull down resistor. This is correct?? A drawing would be great!!
Steve
PS. I tried the debounce to 1200 and it helped but did not fix the problem.

[ger21]

You can crank the Debounce up to 10,000 or even more to get up and running. What it does, though, is delay the activation of the switch.