tool touch off issue

After successfully remapping M6 and reconfiguring to suit, I can now proudly touch off my tools automatically to my proximity switch based touch off "mushroom".
However, it turns out that the inaccuracy is about 0.2 mm, which renders it pretty much useless.
What precision are others able to reach?
Thanks,
/Chris

you'll find proximity switches will vary depending on the tool size , and the type of prox switch your using
you'll find a simple spring loaded pad , isolated from the machine as a base much better

would you be kind enough to post your config , i'm struggling with tool touch off at the moment ,
yours may help pin the problem down

Have you looked here...

softsolder.com/2010/04/14/emc2-ugliest-t...-probe-station-ever/

Rick G

Thanks for your responses! Gee, that was an ugly TLO probe! And far better than my luxury version! That convinces me to struggle on and not give up on automatic TLO measurement.

@cncbasher: I will document my equipment and how I did this in a separate thread. It was quite complicated for a noob such as me and I need to sort out what I did.

The "mushroom" I am using is a commercial product that has a spring loaded top. The tool pushes the top, which moves toward the proximity switch inside the "foot". The tool itself never comes close to the switch, so the size of the tool should not matter.

I can think of a few suspects for the big error:
1 Maybe a proximity switch is not such an exact device as I thought. But why then are we using them instead of mechanical switches? 2 Maybe my assumption that the tool length is captured at the very step (I am using stepping motors) that the probe pin changes value (without delay) is incorrect and that the slower the probe move, the more exact value? I am probing at full speed.
3 Maybe I have mechanical play despite a reasonably adjusted Z axis ballscrew? It might be time for a scientific approach to measuring any play.
4 My "mushroom" is made of aluminum. Maybe aluminum is a bad element for an inductive device?

I was thinking along the lines of 1 when I asked the question and the exactness of inductive proximity switches in general. Can they produce accurate enough results?

Thanks,
/Chris

I am probing at full speed.

As you can see in the link above the actual measurements are taken at a slow speed for greater accuracy.
Sorta like homing .

Rick G

as rick says approach the switch at a slower rate , i do it in 2 stages , a fast move to get me near , then a slower move at say 20% of speed to the actual switch .
inductive switches are better for say a fixed area where the sensed material both in area and density don't change , then you'll get reasonable fixed distances pretty well every time , but however a tool changes both in area and density ( think carbide or hss )
so repeatability will be hard , unless you all ways use say 4mm milling cutters .

you need a non metallic surface above proximity switches , aluminium will again alter the inductive field
and in effect de-sense it , you may find changing to a acrylic top , will make a big difference .

@cncbasher:
I have now documented my setup in this thread:
forum.linuxcnc.org/forum/38-general-linu...h-measurements#65597

Thanks,
/Chris