Proximity limit switch problems

[irving2008]

Never thought about this before, but if you are using parallel connected HOME switches you MUST configure Mach3 to back off them (unless that's the default) else it'll never home the 2nd and subsequent axis as it'll never see the switch state change.


Also, series connected NC surely only works with the 2-wire versions.

[JAZZCNC]

Never thought about this before, but if you are using parallel connected HOME switches you MUST configure Mach3 to back off them (unless that's the default) else it'll never home the 2nd and subsequent axis as it'll never see the switch state change.

Yes that's how Mach works Irving Hits the switch then backs off then Zero's the Axis before moving onto next axis.

Also, series connected NC surely only works with the 2-wire versions.

For Limits ONLY then Series wired Prox switches work just they have slower reaction times with each switch but for the few needed, esp if like me you use traveling switches that look for a target so only 1 each axis then it's not a problem.
The Link Eddy gave explains about series wired prox switches.!! Sensors - Outputs & Wiring

[davrich]

I decided to upgrade my cheap proximity switches and makeshift wiring for some Pepperl & Fuchs ones.

Seemed like a good idea, until now!

The ones I’ve bought are PNP https://www.pepperl-fuchs.com/global...ils&prodid=514.

I’ve hooked them all up and they light up when metal is placed in front of them.

Problem is, in Mach 3 th fact that the switch is tripped doesn’t register.

When I add my old Chinese cheap switch, it works fine.

Seems that the issue is that my old ones were NPN and the new ones are PNP.

can anyone tell me whether this is easily fixable by some tweaks to the wiring or am I better off just getting NPN sensors?

If any more information is needed to answer the question I’ll fill the gaps.

Thanks in advance

[Doddy]

I decided to upgrade my cheap proximity switches and makeshift wiring for some Pepperl & Fuchs ones.

Seemed like a good idea, until now!

The ones I’ve bought are PNP https://www.pepperl-fuchs.com/global...ils&prodid=514.

I’ve hooked them all up and they light up when metal is placed in front of them.

Problem is, in Mach 3 th fact that the switch is tripped doesn’t register.

When I add my old Chinese cheap switch, it works fine.

Seems that the issue is that my old ones were NPN and the new ones are PNP.

can anyone tell me whether this is easily fixable by some tweaks to the wiring or am I better off just getting NPN sensors?

If any more information is needed to answer the question I’ll fill the gaps.

Thanks in advance

Sorry in a late reply to this - only found this thread by looking for information on inductive sensors.

For 3-wire sensors:-

NPN are low-side switching, so the sensor goes low impedance to ground in the presence of metal. PNP are high-side switching, so the sensor goes low impedance to supply-voltage in the presence of metal.

This has a bearing on how you wire these to your BoB, or whatever (for now, "BoB"). If the BoB's inputs are opto-coupled AND it presents both anode and cathode of the optocoupler input then you have a lot of flexibility, and either NPN or PNP can be made to work. For an NPN you'd connect the sensor output to the cathode of the optoisolator, and wire the anode to +5V (or +12V through a current limiting resistor). For a PNP, you'd wire the sensor output to the anode, and the cathode to ground (and if the sensor is powered by 12V, then probably have to include a current limiting resistor as well). More typically the BoBs (at least the couple I've come across) have the anodes of the optocouplers bonded to their own local 5V supply and you have to have low-side (i.e. NPN) switching - hence the NPN being the sensor of choice.

If you're stuck with PNPs and a low-side switching BoB, you *could* frig it a couple of ways. One to include a NPN transistor that is switched from the sensor output from the PNP - very easy, costs pennies and easily knocked up on strip board. Or, you *could* use the high-side switch to reverse-bias the LED in the optoisolator (it's naughty, but can be done) - connect the sensor output via a 470R resistor to the BoB input, connect a 220R resistor to ground (this allows the PNP to pull the cathode voltage up to reverse bias the optoisolator). In theory the cathode would sit around 4V with the PNP sensor "ON" and that's enough to stop the LED in the optoisolator illuminating. The only problem with this method is On is Off, and vice versa, that you have to invert the input in Mach3. You might have to play with the value of the 2 resistors, but keep the ratio the same (the series resistor from the sensor output having about twice the resistance of the external resistor from the BoB input to ground).

Something like



(Note: I've not tried this, but it ought to work)

One word of warning - if you removed the pull-down resistor, or disconnected the 0V to it, you're going to exceed the max reverse voltage on the optoisolator and destroy that input. Either don't do this!, or place a diode - anode on the BoB input, cathode to the BoB's 5V supply to clamp the input voltage to not much more than 5V

[john swift]

prompted by seeing Doddy's reply

I have 2 questions

1)
which BOB are you using ?
it could be one I have reverse engineered or
the BOB's circuit diagram available online

2)
what is the DC supply voltage you intend to power the
normally open PNP proximity switches ?

John

[davrich]

Thanks for all replies on this.

Sorry, I didnt receive notifications that anyone had responded, so will check my preferences.

I ended up buying some NPN ones, so will be selling the others on.

Thanks again