Hello Guest it is March 29, 2024, 02:20:43 AM

Author Topic: edge finding/2.5D probing  (Read 128741 times)

0 Members and 1 Guest are viewing this topic.

Re: edge finding/2.5D probing
« Reply #50 on: November 20, 2007, 04:06:13 PM »
Woah.....you need to signe her up on the forums and let her coax Art and Brian!  ;D ;D ;D

hey anything I can do to help. I use Rhino for my point clouds, but I'm not much of a modeler!

Mike

Offline stirling

*
  • *
  •  2,188 2,188
  • UK
    • View Profile
    • www.razordance.co.uk
Re: edge finding/2.5D probing
« Reply #51 on: November 21, 2007, 03:27:44 AM »
Hi Mike - this could get frustrating what with the time difference...

You could just run the g-code but it'd be better to run the app from your start menu first. By default all the settings are in mm so you might get a few surprises - like a feedrate of 3000 for instance.  :o So....

Run the app, set up your choice of params, run probe25D.tap. The probe will move to startX,startY at a height of safeZ then drop to probe@Z and then travel Y+ve till it hits your edge, it'll then wander round the edge putting points in your triple file. Note that you need to move/delete/rename the triple file each run because it's appended to - not re-written. BTW double-clicking on a file text box will bring up the open/save dialogs. Must write some destructions for it all ;D

Also - could you send Terry a copy of the zip - I'm having trouble getting it through...

Offline stirling

*
  • *
  •  2,188 2,188
  • UK
    • View Profile
    • www.razordance.co.uk
Re: edge finding/2.5D probing
« Reply #52 on: November 21, 2007, 04:19:36 AM »
TP - Re: your g4p0 trick - is that for use in VB as in after a "code G31..." or in a .tap or both? I tried it in a .tap and no difference.

G1 Z10
G1 X10 Y10
G31 Z0
G4 P0

Still crashed down to zero even though the object probed was at Z5 - strangely it records the point correctly - just doesn't ****ing stop! and crunch AGAIN!

So can you actually run a bed o nails routine and have it work ok? - I've never got one to work properly yet!... whether I use Art's wizard or Art and Brian's plugin. every time - crunch - I've tried versions of Mach from 1.84 to 2.6. - all the same.
Re: edge finding/2.5D probing
« Reply #53 on: November 21, 2007, 09:42:57 AM »
Hi stirling,

No problem sending the file to TP. I will do that in a few minutes.  Ok I know I'm going to look stupid, but I have before and will again! ;D How do I run the app from the start menu? How do I load it in Mach or rather call it so that it appears?

Mike

Offline stirling

*
  • *
  •  2,188 2,188
  • UK
    • View Profile
    • www.razordance.co.uk
Re: edge finding/2.5D probing
« Reply #54 on: November 21, 2007, 09:54:25 AM »
Hi Mike - yeah sorry didn't explain things too well. I'm assuming you've:
1) renamed the file I sent as probe.zip
2) unzipped it
3) run setup.exe
4) copied the gcode file and the macro as instructed

now please have a look at the attached (word) doc.
« Last Edit: November 21, 2007, 10:00:37 AM by stirling »
Re: edge finding/2.5D probing
« Reply #55 on: November 21, 2007, 10:31:10 AM »
Ahhhhh........I had not installed the exe file correctly. It will be a little while before I can get to the shop to test, but I will run a test sometime today and get back with you. Thanks for your help, and I will get this to Terry. He will be able to wring it out better and quicker than I! ;D

Mike

vmax549

*
Re: edge finding/2.5D probing
« Reply #56 on: November 21, 2007, 12:02:21 PM »
Ian now try adding in a delay G4 P1. First set your config page for G4 to be MSec not secs. I had to adjust what it took in delay time for it to work correctly. Processor depend I believe.

It is used in both G31 and CODE"g31"   I have seen where MACH will skip over code intructions IF it gets tied up to the point that the instruction overran the available cycle time and MACH was forced to continue. AND when it did, it was an instruction or two down the road and would miss some things.

In you case and mine , Mach was so busy trying to get the trip point data and record it to the file that it missed the STOP and retract instruction occasionaly . CRASH goes the tip.

SOme say that could not happen but in my case the G4 P1 solved the problem(;-) I have since probed thousands of cycles without mishap.

Just a thought, WHat processor and speed is YOUR computer??

(;-) TP
« Last Edit: November 21, 2007, 12:49:59 PM by vmax549 »

Offline stirling

*
  • *
  •  2,188 2,188
  • UK
    • View Profile
    • www.razordance.co.uk
Re: edge finding/2.5D probing
« Reply #57 on: November 21, 2007, 01:03:52 PM »
Terry, just tried it with a few delays and thus far no luck - I'll try some more experimenting tomorrow. BUT - how's this for wierd?

I just noticed that it fails to work in a pattern. I went back to Art's wizard and set it to probe 100mm in the X and 50mm in the Y with a stepover of 5mm. It turns out that whenever X is 0 or 5 or between 95 and 175 it works fine, any other value of X i.e. between 10-90 and 180-200 and it doesn't. i.e. it fails in "stripes" in the Y direction. Regular as clockwork  ???

Tomorrow I'm going to probe my whole table and I'm betting it spells something out - like playing a sabbath album backwards ;D

vmax549

*
Re: edge finding/2.5D probing
« Reply #58 on: November 21, 2007, 03:12:58 PM »
IAN it probably spells out GOTCHA(;-) That is weird. BUT I did notice it failed in a pattern, it would cycle for number of cycles and error, cycle for about the same number of cycles and error so a skip pattern could occurr.

(;-) TP

Offline stirling

*
  • *
  •  2,188 2,188
  • UK
    • View Profile
    • www.razordance.co.uk
Re: edge finding/2.5D probing
« Reply #59 on: November 22, 2007, 11:03:43 AM »
couple of updates: Couldn't seem to get anywhere Terry with the delays. But I've found another way of getting round the problem. If I make sure that probing is started on integer X and Y and stepover is integer then it seems quite happy. Of course this is only really practical if you're in mm. If I make the machine coords the same as work offsets then it's even more likely to succeed -  ??? ??? ???.

Did a probe of a good chunk of my table with Art's wizard and sure enough same old same old - failed in stripes in both X and Y - looks a bit like tartan!!!! That's how I came to figure out the integer bit - set everything to integer boundaries with no work offsets and it works every time!!!!

I've also added a bounded 3D probe vid to the web page for your entertainment  ;D http://www.razordance.co.uk/probe25D.htm

How's the software going? - got it probing yet?

Ian