Centroid Community CNC Support Forum

I found some precision limit switches at Automation Direct. They claim 5 micron accuracy but have only 5mm travel. Would these switches be suitable for setting X home position so I don’t have to take a skim cut and measure every time I power up the lathe?

items needed to answer
link to switch
report from your system so i can see steps per rev and turns ratios
tolerance you hope to hold
tolerance of the machine tool

Like Keith said, need more info.
I have found that hall effect and proximity sensors give good repeatability.
The best repeatability comes when you use an axis motor with an encoder and its drive's ability to output that index pulse.
Use a mechanical switch to rough home, then CNC12 will move and look for the marker pulse input. Set home there.
Provided the machine is mechanically sound (Good thrust bearings, good axis motor to screw coupler or solid timing pulley/belt setup) its much more repeatable and accurate in my opinion. Infact, this is how Centroid Servo systems work. (and most others)
What you do have to be careful of is whether or not that index pulse output can drive an input without any external circuitry in order not to damage the encoder. DMM DYN2 and DYN4 have this feature engineered into them and can sink an Acorn input without external logic.

Marty

I use these inductive proximity switches on my CNC router. My homing routine for each axis works like this:

• move to switch at 50ipm
• move opposite direction until switch clears at 5ipm
• move to switch at .1ipm

I have found this give very accurate homing, well within the accuracy/repeatability of my machine.

I am using Clearpath servos (low res versions...can’t recall the steps/rev but think it’s 800?) so, unfortunately, unable to output the encoder signal. I didn’t know DMM added encoder output capability to the Dyn2 drive. I have an unused Dyn4 drive but don’t have a motor that would suit this application.

These are the switches I was looking at. 5 micron repeatability and 5mm stroke length.
https://cdn.automationdirect.com/static ... tls5um.pdf
https://cdn.automationdirect.com/static ... ightls.pdf

I have the Clearpath mounted to a THK linear slide for my lathe X axis. I believe it’s 6 turns/inch. Tolerance of my lathe is pretty sloppy. If I could repeat to a couple thou, that would be plenty good enough. I would imagine these switches should help but may have to slow the home speed so as not to crash into them. 5mm is the longest stroke available and that does not sound like much.

ScotY wrote: ↑Tue Dec 17, 2019 1:18 pm I am using Clearpath servos (low res versions...can’t recall the steps/rev but think it’s 800?) so, unfortunately, unable to output the encoder signal. I didn’t know DMM added encoder output capability to the Dyn2 drive. I have an unused Dyn4 drive but don’t have a motor that would suit this application.

These are the switches I was looking at. 5 micron repeatability and 5mm stroke length.
https://cdn.automationdirect.com/static ... tls5um.pdf
https://cdn.automationdirect.com/static ... ightls.pdf

I have the Clearpath mounted to a THK linear slide for my lathe X axis. I believe it’s 6 turns/inch. Tolerance of my lathe is pretty sloppy. If I could repeat to a couple thou, that would be plenty good enough. I would imagine these switches should help but may have to slow the home speed so as not to crash into them. 5mm is the longest stroke available and that does not sound like much.

Proximity switches should do well for you then Scoty
Marty

With six turns per inch on your ballscrew, each turn is 0.1667". Divide that by 800 counts per turn and you have 0.0002" per encoder count. That should work for most people.

I've noticed that many users put their proximity sensors in line with the target. I've always mounted my sensors perpendicular to the target so that the sensor would never smash into the target. Normally, I just use a bolt for the target that has had its head shaved on the lathe so that the surface of the target is flat. On rare occasions, I have used a rare earth magnet glued to the face of the target to allow more distance between the sensor and the target. The target is mounted to a movable bracket so that the desired home/limit position can easily be set. The secret for me was to always use a constant speed when approaching the target the first time, then I would back away from the target and approach it again at a very slow speed. That gave me results as repeatable as the mechanics of my router would allow.

ScotY.

The switches you are talking about is probably the same switches I use on my mill.
They are normally closed 5 or 6mm plunger switches and VERY accurate. Made by Metrol.

I can home my Z axis many times and measure position with a test indicator each time and always get back to the same position within my measurement capability of 0.005mm.

Expensive but GOOD.

Here is the one I use.

One more thing.
I have by mistake run into the switch and bent the bracket without damaging the switch.
The one pictured above has 2.8mm stroke but also comes in 5mm stroke version.