edge finding/2.5D probing

[vmax549]

HUM I just checked and on my machine if I run the G31 and trip the probe the Digatize LED lights up on the trip.?

 It will light up on any trip of the probe does not matter what process you are running, or just sitting there and you trip it with your finger.

Be VERY carefull if you run the probe in any function other than the G31. Crunched probes get expensive(;-)

Like I said before the G31 in the bed of nails approach is a piece of cake. Using it it a CMM type probing is very difficult with the current G31 cycle and NO tip compensation.

Alot of times I just drive it around with the hand pendant and have a huge led setup on the CMM probing page. I just drive it until I get a trip on the LED and back off the trip. I then apply the comp by hand/calculator and record the position on a piece of paper. I then use macros to do other functions such as center of circle with diam. Center of two points x,y z and a axis. COR rotation of the vise or COR of part by two hole method, etc.

NOT always classy looking but functional.

If the G31 had tip COMP it would be easy to construct a plugin to probe the majority of 2 d objects with a little hand intervention from time to time.

(;-) TP

 

[stirling]

I've taken a bit more of a look at the digitize LED lighting/not lighting. It seems that sometimes it does and sometimes it doesn't. If you have a moment, try this code in the VB Script window. I'd be grateful for your findings on your system.

code "G21"
code "G31 Y100"
While IsMoving()
Wend
MsgBox GetOEMLed(825)

arrange for the probe to be tripped by some obstacle (i.e. not your finger) at somewhere around Y50 mm. Try it with different feedrates. On my system at feeds around 1000mm/min the LED always lights fine but when you get down to lower feeds around 100mm/min, although the probe trips - i.e. it stops! - the LED doesn't light.

[vmax549]

If there is extra movement in the probe mechanism you can get that. When moving slowly you can actually bump the trip switch but when it stops it losses contact and the led goes out. When moving faster there is always a touch of overun that will keep the switch in contact.

It would be nice if the G31 cycle would be movement to trip switch then back off switch untill it untrips THEN record position.

(;-) TP

[stirling]

If there is extra movement in the probe mechanism you can get that. When moving slowly you can actually bump the trip switch but when it stops it losses contact and the led goes out. When moving faster there is always a touch of overun that will keep the switch in contact.

yes - I've done a bit more experimenting and I agree. It seems to be all about tweaking the motor tuning to make sure you get a good trip! - reduce accel and up feedrate etc.

It would be nice if the G31 cycle would be movement to trip switch then back off switch untill it untrips THEN record position.

Indeed it would, but I think I've found a way to get round it.

[vmax549]

Actually we need to build better probes that are more dependable(;-)

If we build the mechanical side to be as dependable and acurate as we want the software side to be we will have it whipped(;-)

(;-) TP

[stirling]

Agreed.

Bit of an update on progress. My cheap (well actually it cost me zippo ) and cheerful probe is at least allowing me to experiment with the software side of things. So far I have a working wizard that's producing some reasonable toolpaths. At the moment the main issue is the probe ocasionally "sticking" when I retract from a touch - but hey, I'm learning lots so I'll stick with it for a while longer...

[vmax549]

OK Stirling, Art has fixed up the G31 code so it writes the tip COMPed  position to the points file. SO now we have a true touch position available for use.    (;-) TP

[stirling]

Hi TP - Nice one - thanks Art.

Which version/revision do I download to get it please?

Update on my progress then. I have a 2.5D edge finding/profiling wizard that works fine apart from an issue with what I can best describe as suspected communication sync problems between the Mach thread and the VB thread which occasionally trips it up. I've discussed this with Art and he thought this was probably down to the G31 command as well. Maybe his G31 fix will have resolved that also - I'll need to test it out. (see thread http://www.machsupport.com/forum/index.php/topic,4456.0.html)

The other problem I'm having at the moment is with my setup. I seem to maybe have a dodgy sig/gnd which I'm trying to resolve over on the zone (see http://www.cnczone.com/forums/showthread.php?p=356082#post356082).

Hopefully in a few days, time allowing I'll have an up and running routine for beta testing if anyone's interested in giving it a whirl.

[vmax549]

Stirling the same thing may be causing the problem in both instances(;-) A very POOR  contact at the switch. If the switch has a high resistance at the contacts you can get phantom signals from it as it attemts to make contact. Normally a small capacitor across the two leads will help it clear the contact and make connection. That is the problem with running with a 5volt source there is little energy available to help maintain contact. As a test try raising the debounce setting in config. it may give the switch enough time to settle before mach deals with it.


THe probe tip COMP is in V2.56 BUT there is a small glitch it is corrected in version 2.57 so wait for the .57 version to be posted

When you get ready to test your plugin let me know I always like to see other peoples work. I usally get to learn something from the example(;-) TP

[stirling]

Stirling the same thing may be causing the problem in both instances(;-) A very POOR  contact at the switch. If the switch has a high resistance at the contacts you can get phantom signals from it as it attemts to make contact. Normally a small capacitor across the two leads will help it clear the contact and make connection. That is the problem with running with a 5volt source there is little energy available to help maintain contact. As a test try raising the debounce setting in config. it may give the switch enough time to settle before mach deals with it.

Yes - I thought of this, so I actually took the probe right out of the equation and just shorted the input pin to ground with a jumper - same problems still happened. Tried messing with debounce and no change.

I've got a thread running on the zone http://www.cnczone.com/forums/showthread.php?p=356625 where we're looking at the possibilities of an unstable (floating) earth. It almost seems like earth is managing to occasionally come up to logic 1.

THe probe tip COMP is in V2.56 BUT there is a small glitch it is corrected in version 2.57 so wait for the .57 version to be posted

I still haven't quite understood how to know when a new release/version/bug fix is in the offing - maybe you could let me know - thanks.

When you get ready to test your plugin let me know I always like to see other peoples work. I usally get to learn something from the example(;-) TP

When I'm happy it isn't likely to trash your probe