Looking for Good Limit Switches

I’m looking for high accuracy limit switches. I know there was a guy making some (Mr Beaver I think was his name) but they don’t seem to be available anymore.

The switches I got from Carbide 3D a few years ago just don’t have the accuracy I’m looking for. Sometimes over a mm off in any direction.

Edit: I’ve found that if I re-home my machine each time I change a tool it is much more accurate. I assume this means the machine either loses track a bit over a run, or that it gets bumped out of place a little while changing a tool. Whatever the case, re-homing makes these things pretty darn accurate.

Mr Beaver used precision proximity switches. I think others have purchased them from 3rd parties.

https://www.beavercnc.co.uk/post/2018/07/25/fitting-your-hdz-precision-proximity-switches

2 Likes

The stock switches should be much more precise than that. Are you sure it’s the switches giving you that error?

2 Likes

I believe the OEM S3 homing switch is an Omron D2QW-C003H. According to the datasheet, it has an operating position tolerance of +/-0.3 mm, but I imagine that the repeatability of a given switch is better than that.

2 Likes

I just ordered and received these recently. They seem to be nearly identical to the ones Luke (Mr. Beaver) had.

2 Likes

@The_real_janderson
I haven’t used those, but wouldn’t you need a 5v input? Or do you connect to 12v or 24v?
I found these.
https://s.click.aliexpress.com/e/l6i7iobq

Yeah, I missed that. Oh well, shoot, I guess I’ll hook them up to a separate power supply. The frustrating thing is that they provide no wiring diagram and I couldn’t find one online. How do people wire these up?

Mr. Beaver provided a splitter that tapped one of the +5v pins and splits it to three. I lost mine since it took me 6 months to actually install them, but I made mine from a bunch of breadboard jumpers I had laying around.

Something like these, with both sockets and pins would work for making your own splitter.

EDIT: Here is Mr. Beavers page for installing his.

3 Likes

I’ve used similar for other purposes. Red or brown (I can’t tell) goes to +V, blue goes to GND (against convention), and the black goes to your signal input.

1 Like

Ah perfect! Thanks for the instructions. It looks like I either need to get the 5v ones or power them from 24v.

Thanks @neilferreri! That makes no sense given standard DC electronics. Black wires are always ground. That makes sense why the signals were super strange.

1 Like

I installed low cost reed switches into my XXL and never looked back. No power required, just figure out how to mount the and a magnet and forget it.

1 Like

Are the proximity sensors really more accurate/precise/repeatable? The data sheets say they sense 5mm ±10%
Does that mean any given detection could be 1mm from any other?

1 Like

I would merely be content to suppress rehoming after a run in Carbide Motion. I felt like they didn’t understand the significance in the Motion feature request thread (and then they closed it). Maybe it doesn’t matter for wood, but the speced 10 thou on the OEM switches is pretty apparent when multiple runs with a tool change are needed (my switches seem significantly worse - I will need to test).

3 Likes

Please report your findings!

I need to do some testing as well on my Nomad.
I have proximity sensors, but have yet to tackle the wiring.

2 Likes

Thanks for that info!

Where do you find ‘operating position tolerance’? I looked through the linked pdf and didn’t see that. Is it a standard metric for something like a concept approaching ‘precision of zero at rest’ when motioning through the switch (consistency, repeatability, etc.)?

I am sounding really dumb, but I would like to be able to compare positional accuracy (actually precision since we are talking repeated procedures) across different switches/homing methods. I feel it is almost not worth the investment I made in an ATP-5 aluminum plate bed if can’t reduce the possibility of getting up to 0.3mm of positional drift when homing (especially in the z-axis)…

I found it.

Actually, I think that is a reference to the location of the switch junction inside the switch housing, as it is listed as 8.4mm+/-0.3mm, not ‘junction action’ drift. For example, if the mechanical junction action (closing or opening the switch) for a given switch occurs at 8.2mm, then it will always occur at 8.2mm. Let me know if this isn’t correct…

I am way more interested in the possibility of any drift in the location that the junction action occurs (for example: one time the switch triggers at 8.23mm, and the next time it happens at 8.18mm, and the next time it happens at 8.20mm, etc.). I don’t mean gradual drift due to wear, I mean more of a random drift (mechanical positional tolerance?), or even an environmentally triggered drift (like from changes in temp/humidity/possibly corrosion). Is there any metric for this?

I think you are right about the +/-0.3mm tolerance on the Omron switch being purely a dimensional tolerance. I couldn’t find a repeatability spec, so the best way to determine that would be to actually measure it from a sample switch. Snap-action switches are often mounted in a way that they can be adjusted, although the Shapeoko switch mountings are not adjustable. I don’t think they need to be, though. You just want the home position to be the same every time, or as close as possible to the same.

1 Like

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.