Teachmotion probe data edit

I have a die that I probed in two stages and attempted to merge the two files. The line generated in the teachmotion and in the NC corrector file looks perfect. The problem is the tool will not follow the probed profile completely. It does fine for the first few inches and then diverts out away from the profile.
What am I missing when editing these files? I have occasionally had errors when calling an edited file up in teach motion, usually related to a length parameter. What files other than the probedata.txt and probeinputs.c do I need to be concerned with?

Hi GORDYKD,

I don't really follow what is happening. Please post the edited and other related files and describe what you are doing and what happens. Does the problem when teaching with Data Capture on? Is it you cant move up to the white line?

Hi Tom,
The die is a fairly simple shape, similar to a cooking pot. It has a radius at the end(pot bottom), a long straight section(pot side), and a flange at the other end(pot rim).
In trying to probe it, very small steps with significant pullback must be taken to capture the radius and the rim. If those same small steps are used to probe the side, it would take a couple hours to finish. It seemed logical to just add the straight section in through editing.
As simple as it seems, of course it didn't work right. There is obviously more involved in the background that I'm not catching.

The three files attached are the radius morrrad114.txt, the flange morrend114.txt, and the edited profile morrmrg114.txt .
What files are involved and how can a person just create a program for the profile if the desired dimensions are known?

Thanks

Kent

The problem is that when it gets to about X -15.00 the tool starts pulling away from the profile. The display shows the intended profile correctly and the tool will not get close to it after that point. It appears to be pulling out to the end major diameter.

Hi Kent,

The program assumes the probe points are all equally spaced in the one dimension. Displayed as "delta". Basically the position of that axis divided by delta gives the index into the table. If the table points are not equally spaced this doesn't work. You would need to edit the straight part with the correct number of points. Or change the program to work differently with different size deltas.

Tom,
Are you saying that the data points saved in "probedata.txt" are not actually used for the teach motion, there is a "table" elsewhere that it follows? I have not seen "delta" displayed anywhere, or ids delta just the "stepover". Where are the parameters stored? How do I just create a file that describes the profile I want to follow and have everything the teachmotion program is looking for? Several times when I tried editing the probdata file it would not read properly into teachmotion. There must be other data it compares as it loads. Also is there a way to apply an offset to the profile other than the entry as you start teachmotion? It seems we have some variation sometimes and a slight adjustment would be handy, rather than re-probing each time.

Hi Kent,

Quote:

---

Are you saying that the data points saved in "probedata.txt" are not actually used for the teach motion, there is a "table" elsewhere that it follows?

---

No. There is no other data. The file is read into a Table in computer memory by the program.

Quote:

---

I have not seen "delta" displayed anywhere,

---

You don't have this marked in purple?

Attachment 472436

Quote:

---

or ids delta just the "stepover"

---

Yes the distance moved between one probe and the next if that is how you are doing it.

Quote:

---

How do I just create a file that describes the profile I want to follow and have everything the teachmotion program is looking for?

---

As I said the program is looking for a list of xy points that are equally spaced in one direction and which determines the profile shape in the other.

btw we only provide TeachMotion as a free example that we hope someone will use as a starting point to develop something useful. It seems your Version has been modified as the axis ranges and orientation seem to be different. Did you do this? The source code is included and only a few pages long. You might read through it to get a better understanding of what it does.

Thanks Tom,
Now it's starting to make sense. The table expects a constant increment for the data, probably determined by the initial data points.
I will take a look at the source code. I have not modified it, but the machine builder may have. I also am going to try setting up a spreadsheet to create the data points for a known shape and see if
I can make it work.
It is definitely a huge benefit to have your expertise as a resource.
Thanks again.

Kent