Work with probe

myval wrote:

The just touching the spindle on my vice I would like to avoid if possible. Have you seen what damage can 2KW servo do? I actually have and it does not really play nice and friendly game. 

When I touched the nose of the spindle to the vise I was using a pendent and manually bringing the Z down one click at a time.  At the same time I was manually turning the spindle to feel when it started dragging on vice.  There was really no drama at all unless I dropped the pendent or something. I do understand the concern though.

Yeah I do it carefully too but I know myself I would rather keep it automatic and keep me and my short circuit brain out of equation as much as I can.
I think I will try the touching the table solution first (when I put the machine back together and finally source new PC) I think I can do less damage that way.

To calibrate your 3-d probe I would mill a piece to a known height, dimension, and location. Then probe that, and adjust your probe settings until you get probing results that match the known piece.

That is not going to give me much since you always touch on the bottom of the ball. That can not give different result then the height of the block.

I was suggesting milling the sides of the block as well, not just the top. So that you can calibrate the sides of the probe.

Right ok, for sides I have precision ground calibrated cal rings. We are fairly well metrologically setup. The only problem was not so calibrating the Z height of the probe but the basically Z position in space in relation to any other tools.

The idea was to mill the calibration bock with a tool that had been touched off on you tool touch probe, so that its height was known. After milling the 5 faces of the block, then all 5 faces positions would be known. Even if you don't know the precise diameter of the milling tool, it can be calculated by measuring the finished size of the block.

I don't think having a precision ground block helps you, unless your probe is mounted in the spindle and you know it's tip is concentric to your tooling, because you wouldn't know the blocks precise location,

Ok I actually do something similar to get the exact tool diameter. I machine circle with roughly measured tool then measure the exact diameter result and tweak the tool diameter in my tool library.
The precision ground ring actually is for calibrating the ball end dia on the probe, touch it in x+ x- y+ y- and get exact ball dia. My probe is mounted in the spindle but this works for probe mounted anywhere really.

You just gave me another idea (or at least it just clicked in) why don't I use this technique to work out my probe Z offset issue?
I can surface a piece then probe it and this will give me the probe to the tool Z difference (really what I was worried about from the start).
Since I will have the tool touched off on the tool setter this will give me exact difference between probe and tool (tool setter trip point actually) and any other tool touched off on the tool setter will have the same mathematics (complex math +- few mm) applied to it.
And I can even move mathematically my 0 point up so every tool length will have the same sign. Hey welcome back brain.

yeah I think by brain just klick into gear. If that is the way you guys kept saying I finally got it.

 

Hello all.

The actual fork for probe screen next gen seem to be dead (i haven't the admin stuff...).

So i have actually created a new project on github, your welcome for test comment and propose change if you like it.

Some little change done to GUI, but stuff completly revised :

-Possibility to swap from 3D touch probe to Touchplate with a simple tickbox.
-Manual tool change with auto length use now built in python stglue remap. (i think about going back to a ngc version)
-You can bypass the auto length stuff with a simple tickbox.
-Config is done for use separate input for Touch Probe (or Touchplate) and Tool setter and the status is checked before each move.

Of course status is something like beta, use with precaution.

github.com/alkabal/PSNG_V3

Br

Interesting post