Confusion over HomeAll wiring stage

[NovaStar]

Hi all
I have assembled an open-build Lead CNC1515, I purchased CL57Y drivers with NEMA24 encoders (closed loop system) Initially reading documents and watching videos I decided NPN NO proximity sensors would be the way to go.
I'm now starting the wiring phase of the build, and looking deeper into the specific documents required.
So my understanding going off the schematics for the Acorn (ACORN_rev4, LIMIT/HOME NPN PROXIMITY SENSORS) is that it shows 3 sensors connected to 3 inputs (IN1-3) .... so where is the 4th slave? I'm I assuming correctly that the Y-Axis is connected to IN4 as per my drawing?

Then I read this document (Centroid Acorn CNC12 Axis Paring and Squaring) page 13 where it shows a different schematic configuration for the same sensors .
Which is correct??
I also read in that document that its advised not to use NO sensors as they may fail due to EMI/RF interference ... I've used double shielded cables for everything as well as ferrite beads on each individual wire as well as grounding every shield to eliminate EMI, So I believe I can disregard the that statement in the document?
Also I didn't run any bench tests first as was suggested on some videos .. should I be hooking up the Acorn to a PC at this stage and running the software as I go or just finish the wiring completely firsts then powering up??

[jpmsteadi]

Its my understanding that the benefit of NC limit switches is that if something breaks somewhere it alerts you by instantly tripping as opposed to something broken that then doesn't show up and more things could break.

My understanding that the diagram you posted is for individual homing switches whereas the HomeAll or LimitAll inputs are meant to have the switches all connected to one input either in parallel (for NO switches) or series (for NC switches, like my machine).

I am not as familiar with the squaring of a slaved axis as my machine is a mill with only one motor per axis, but I would follow the setup lined out in the axis paring and squaring guide, using the diagram for the NO switches.

[NovaStar]

Thanks jpmsteadi ... so would it not be better to connect individually to an input rather than in series or parallel , that way you have a lot more accurate information from each sensor.
I do get your point on something breaking and you're not aware of it until its too late .... but being a solid state device, not much chance of that happening unlike mechanical switches

[jpmsteadi]

Very true on the solid state switches not breaking. I think my concern is more a cable getting caught and cut or whatnot.

I think the idea is that since the acorn and CNC12 can home each axis at a time and it knows which axis it is moving it isn't actually having any more accurate information as it is only using the switch on each axis as it is moving that axis. I don't think one or the other ends up being more accurate than the other. The added plus is that it frees up some more inputs for touch probes or tool touch offs.

[NovaStar]

Sorry to bump this up ... but would love some advice from experts regarding the second half of my question;

Heard or read hardly anything about the importance of EMI, ferrites or the use of shielded cable;

I also read in that document that its advised not to use NO sensors as they may fail due to EMI/RF interference ... I've used double shielded cables for everything as well as ferrite beads on each individual wire as well as grounding every shield to eliminate EMI, So I believe I can disregard the that statement in the document?
Also I didn't run any bench tests first as was suggested on some videos .. should I be hooking up the Acorn to a PC at this stage and running the software as I go or just finish the wiring completely firsts then powering up??

[tblough]

There are many reasons that NC sensors are much better than NO sensors on machine tools. Susceptibility to EMI/RF noise is only one. If a wire breaks or comes loose on a NO sensor, how do you know until that circuit fails to perform? On a limit switch, this is usually demonstrated with the machine smashing through the inoperative switch. With a NC limit switch, a broken wire would cause the machine to think the machine was at the limit and prevent movement until the problem was corrected.

As with any recommendations in the documentation on here on the website, you are free to ignore them if you wish. You'll just probably regret it later.

[carboncymbal]

There are many reasons that NC sensors are much better than NO sensors on machine tools. Susceptibility to EMI/RF noise is only one. If a wire breaks or comes loose on a NO sensor, how do you know until that circuit fails to perform? On a limit switch, this is usually demonstrated with the machine smashing through the inoperative switch. With a NC limit switch, a broken wire would cause the machine to think the machine was at the limit and prevent movement until the problem was corrected.

As with any recommendations in the documentation on here on the website, you are free to ignore them if you wish. You'll just probably regret it later.

I absolutely agree with this. Additionally, addressing interference issues isn’t a binary actions. No one can say “I did x, y, and z, so I won’t have interference issues”. I would argue that it isn’t that simple and that any countermeasures move you to a lower risk position on a sliding scale, but nothing is a 100% guarantee.

I can’t see any compelling reason to use NO switch’s. Even if you have already purchased them I would buy new NC switches.