[tkms002]

using arduino/proto screw shield and GRBL with UGS, how do you wire a touch probe and get it to work?
Thanks

[tkms002]

ok so apparently you just wire some kind of contact plate from pin A5 to ground to close this circuit.
I have also read you may or may not need a "pull up or down resistor" to not fry the arduino??
What is that about?
I know there is a probe helper on UGS but for now all I want to do is get the probe properly wired.

Thanks

[109jb]

Pull up resistor is a weak resistor that connects between the pin and 5V. A pull-down is about the same except the resistor connects between the pin and ground. What they do is insure that the pin doesn't "float" which can cause errant readings.

If you are running Grbl on an Uno, Grbl defaults to activating the built in pull-up resistors. So you don't have to install a pull up resistor if using a NO type switch. The contact plate deal is essentiall a NO switch so you would be ok just hooking between the probe pin and ground.

[tkms002]

I am running an UNO.
If I want to use NC limit switches what exactly do need to do with resisters and GRBL settings?
Thanks

[109jb]

There really is nothing mandatory to do on an UNO to use either NO or NC limit switches. The internal pull-up resistor is activated by default, so either NO or NC can be connected directly.

However.... The internal pull-up work but are very weak and you can still get false triggers especially if you have electrical noise. Common sources of noise are spindle motors, power supplies, and even the stepper motor wiring itself. You will need to pay careful attention to routing of all of the limit switch wires to try and route them away from possible sources of noise. You can also add an external RC filter as an additional combatant of noise. This consists of a resistor and capacitor that cleans up the spikes from the noise preventing then from causing a false trigger.

Here is a WIKI page from the Grbl site that does a good job describing limits. The same can be done for a probe since a simple probe is just a switch anyway.
https://github.com/gnea/grbl/wiki/Wiring-Limit-Switches

[109jb]

Thought I should add a little information from personal experience. I personally have used a couple methods.

1. On my first machine, it was already configured to use a PC based controller (LinuxCNC) with a breakout board, so it had a breakout board that had optocouplers for the limits. For this machine I just used an old parallel port cable to connect the arduino to the breakout board. So in essence I had optocouplers for the limits similar to what is shown on the wiki page I linked. No problems here.

2. On another machine I just connected the switches to the arduino directly. This worked ok for homing, but I experienced false triggers from noise during operation. So, I took the lazy way out and enabled the switches for homing but turned off hard limits and only used soft limits on that machine.

3. On the current machine I am building I will do the same as #1 because I plan to use this machine with both Grbl and LinuxCNC again.

[tkms002]

Thought I should add a little information from personal experience. I personally have used a couple methods.

1. On my first machine, it was already configured to use a PC based controller (LinuxCNC) with a breakout board, so it had a breakout board that had optocouplers for the limits. For this machine I just used an old parallel port cable to connect the arduino to the breakout board. So in essence I had optocouplers for the limits similar to what is shown on the wiki page I linked. No problems here.

2. On another machine I just connected the switches to the arduino directly. This worked ok for homing, but I experienced false triggers from noise during operation. So, I took the lazy way out and enabled the switches for homing but turned off hard limits and only used soft limits on that machine.

3. On the current machine I am building I will do the same as #1 because I plan to use this machine with both Grbl and LinuxCNC again.

On the machine you connected the limits directly to the arduino did you do anything to isolate or dampen interference like the WIKI page suggests short of the opto-couplers?

Like the addition of caps and resistors?

[109jb]

On the machine you connected the limits directly to the arduino did you do anything to isolate or dampen interference like the WIKI page suggests short of the opto-couplers?

Like the addition of caps and resistors?

On the one I connected the Arduino directly to the machine I had no optocouplers, resistors, capacitors, nothing.

[tkms002]

Does the touch probe wire need to be shielded or just a twisted pair?

[109jb]

Does the touch probe wire need to be shielded or just a twisted pair?

Shielded would be better, but on my machine an unshielded wire seems to work just fine for the probe. I attribute this to the fact that when probing the biggest cause of electrical noise, the spindle motor, isn't on.

[tkms002]

does arduino/grbl ignore any input from the probe when it it not actively probing?

[109jb]

does arduino/grbl ignore any input from the probe when it it not actively probing?

Yes and no. If your interface program has an indicator for the probe, it will show the probe state whenever it is triggered whether you are actively probing or not, but it doesn't actually do anything unless you are actively probing in a probe command (ie: 38.x).

[tkms002]

Yes and no. If your interface program has an indicator for the probe, it will show the probe state whenever it is triggered whether you are actively probing or not, but it doesn't actually do anything unless you are actively probing in a probe command (ie: 38.x).

so if there is an erroneous signal on the probe wires it will not affect a running program?

[109jb]

so if there is an erroneous signal on the probe wires it will not affect a running program?

As I already said, it will not do anything unless you are actively probing. If you program doesn't have a G38.x then it will only show the probe pin state, like an indicator light.

[tkms002]

thank you