Hello Guest it is July 14, 2020, 05:31:23 PM

Author Topic: Help with digitizing/probing  (Read 4588 times)

0 Members and 1 Guest are viewing this topic.

Offline m8298

*
  •  14 14
    • View Profile
Help with digitizing/probing
« on: March 21, 2013, 08:36:05 PM »
Hello, I am somewhat new to CNC and I am probably trying to run before learning to walk. I have a home-built 4 axis router table which I have been using for a little while now. I have the 2010 screen set and I have added the laser crosshair xy zero and a z axis zero plate. Now I want to learn all that I can about digitizing. I made the probe and I kinda have it working. I have been trying to use it with the 4th axis digitize wizard. One problem I noticed is it occasionally does a arbitrary point without actually touching. I put a stiffer spring inside the probe, thinking git was from vibration. But it still does it. I also noticed that when this happens it doesn't light up the probe led on the screen like it does when it actually touches. I am sharing port 13 between the tool zero plate and the touch probe(inverting the signal when needed.)  Any ideas what's wrong?  I would also like some idea on how to work with the resulting point clouds.  I would like to be able to take something I turned on my wood lathe, probe the surface on my 4th axis, then convert the point cloud to something I could work with to cut and decorate the surface of the wood turning.  Any help you all could offer is greatly appreciated. 

Offline ger21

*
  • *
  •  6,282 6,282
    • View Profile
    • The CNC Woodworker
Re: Help with digitizing/probing
« Reply #1 on: March 22, 2013, 08:57:49 AM »

It sounds like noise could possibly be triggering your probe. Adding a capacitor across the input should help, if noise is the culprit.

I've never done any probing, but here's how I'd go about probing your turning.
Rather than probing the entire turning, which will take a considerable amount of time, I'd probe a single pass along the length of the turning, to get the profile. I'd then bring those probed points into a CAD program, draw a line through the points, and revolve it.
Then you'd have a model of your turning in your CAD program, to do whatever you want with.

Be aware that this will probably require writing your own probe routine, and figuring out some way to bring the points into your CAD programl. Then, you''ll need to figure out how to program whatever you want to do without having to go over the entire turned surface.
This will probably not be a trivial task. It would probably be much easier to just have the CNC do the entire turning from start to finish. But I'm sure that's not really what you want.
Gerry

2010 Screenset
http://www.thecncwoodworker.com/2010.html

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html

Offline m8298

*
  •  14 14
    • View Profile
Re: Help with digitizing/probing
« Reply #2 on: March 22, 2013, 10:27:26 AM »
Thanks for the response Gerry. BTW, I still love the screenset.

I don't think I can use a cap. My tool zero plate is active low, and there is a pull-up resistor holding the signal high until the plate is touched. My touch probe, which is using the same input, is active high. If I use a cap the signal would remain high longer, right?  Should I do this a different way?

I have already been thinking about taking a single line of point on my rotary axis.  Then maybe extruding the results and maybe making a wrapped rotary design. At least as a starting point, until I could learn a better way.

Offline BR549

*
  •  6,932 6,932
    • View Profile
Re: Help with digitizing/probing
« Reply #3 on: March 22, 2013, 11:34:19 AM »
(;-) There are routines to probe on a lathe.  I have a routine to import the points to a dxf. You should be able to import a CSV point file directly into cad.

The tools are out there, (;-) TP
Re: Help with digitizing/probing
« Reply #4 on: March 22, 2013, 12:06:12 PM »
If you use a cap, yes it will stay high longer, but whether that is significant depends on how fast your probing move is.  One approach is to move until probe triggers, then withdraw a tiny bit and move much slower until it triggers again.  The significance of any delay the second time will be much less.

Offline m8298

*
  •  14 14
    • View Profile
Re: Help with digitizing/probing
« Reply #5 on: March 24, 2013, 09:20:15 AM »
I am a little confused.  When I am probing, the signal is active high (probe triggers when 5v is present.) The leads inside the probe hold the signal grounded until probe touches. So when I get these "phantom points", it must be seeing 5v.  I understand how this could be noise, but if I put a cap across the input, it'll stay high longer. If it stays high longer, won't the routine fault?  And how would I size the cap to try it?  I keep think I need something inside the probe to hold the signal low better, like pull down resistors.  I just don't know how to wire it.
Re: Help with digitizing/probing
« Reply #6 on: March 24, 2013, 11:58:35 AM »
Sorry, I misunderstood.  I though the spurious trigger was on your active low touch plate, but re-reading the OP I see it is on the active high probe.  This is probably noise from the x or y drives getting in through an earth loop.  The fact that you don't see it with the touch plate suggests that the z drive is not the problem.  Needs attention to single point earthing perhaps?

Within the probe, I assume that the signal is held low by a short to earth, so a pull down resistor will not help.  Since there is a spurious signal even when there is a low impedance to earth through this short, it indicates more inductive noise pickup from the stepper drive wiring.

Offline m8298

*
  •  14 14
    • View Profile
Re: Help with digitizing/probing
« Reply #7 on: March 24, 2013, 12:12:09 PM »
After reading different areas on this forum, I realized that I have never done the Windows XP Optimization and I had a bunch of startup processes running. After killing all of the unneeded processes I was able to run Art's 4th Axis Digitizing Wizard without a single error (hopefully not just a coincidence.) Now I have new questions:

The resulting file is a triplet text file with CSV data. I probed a round piece on my A axis using center of workpiece as Z zero. What I get are X, Y, and Z values that are all positive. How would that relate to a three dimensional round object?  Also, if I open the file in MachCloud, I don't see anything?  Are there any instructions for MachCloud or any of the wizards?

Offline ger21

*
  • *
  •  6,282 6,282
    • View Profile
    • The CNC Woodworker
Re: Help with digitizing/probing
« Reply #8 on: March 24, 2013, 01:58:28 PM »
Quote
The resulting file is a triplet text file with CSV data. I probed a round piece on my A axis using center of workpiece as Z zero. What I get are X, Y, and Z values that are all positive. How would that relate to a three dimensional round object?

Not sure what kind of relationship your looking for? All your getting is points in 3D space. There's nothing that says that the shape is round. As I said before, you probably just want to probe along a single line along the axis of rotation.

As for Machcloud, I don't know of any tutorials. I played with it once, and it took me a while to get anything to work. I think the difference was whether the point cloud file contained actual X, Y and Z letter. It would work either with or without them, but not the opposite. Can't remember which was which, though.
Gerry

2010 Screenset
http://www.thecncwoodworker.com/2010.html

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html

Offline m8298

*
  •  14 14
    • View Profile
Re: Help with digitizing/probing
« Reply #9 on: March 24, 2013, 02:24:01 PM »
I am not sure what relationship I am looking for. I just thought that a 4th axis digitize wizard would somehow mathematically calculate points such that the resulting point cloud would resemble a 3d object, round or otherwise. Because this object is round, I end up with duplicate sets of points for every rotation.  I guess I thought that there would be a way to calculate Y points based on degrees of rotation. How would I use a point cloud with only X and Z values for a three dimensional object?

I understand what you are saying about probing one line along a single axis, and i plan on trying that if I can find a wizard that will let me.