QTDragon_hd with auto toollength measurement

Hi,

This is driving me nuts. I used to use gmoccapy but since there is no support on it anymore I changed to QTDragon and QTDragon_hd.

Looks great, even the probing routines are useful with my topcom.cz/dig3dlpm/ probe. Material heightsetter topcom.cz/digzm/ works too. For a toolsetter I use a cheap Chinese one nl.aliexpress.com/item/32931537443.html

Problem is this, in gmoccapy I had some M6 remap routines to automatic measure and adjust the tool table for a given end-mill. It worked great and automated a large boring part of the milling on my machine.
I wanted to use the same or something similar in the QTDragon screens.I can not get it to work. There is no documentation that is of any use, the examples I find are outdated and / or do not work. The gmoccapy way just doesn't work in the Dragon screens since there seems to be an internal hal_manualtoolchange module.

I see that the author of QTDragon is aware of the problems, I just do not see any solution or a working example where the M6 T* command will do the following if the current tool is not the same as the  new tool:

1. display a dialog box stating a tool change is necessary, showing the new toolnumber
2. cause the spindle to move up
3. move the spindle to a toolchange postion
4. let's the user click a button in a dialog box to proceed to nest step
5. move the spindle to the toolsetter position
6. do a rapid move to a safe distance from the setter
7. move slowly towards the setter till contact
8. retract and measure again but now even slower
9. adjust the tool offset table with the value obtained
10. move spindle up
11. and continue with the gcode program

This is how it worked in gmoccapy and I was very happy with it. Would like to use the same workroutine with the modern qui.

Is there anybody who got it working and is willing to share the ini files and ngc scripts?

I'm using QTDragon under LinuxCNC 2.8.2 and my setup works pretty much as you describe.
See my posts in this thread  It may help.
John

 

JohnnyCNC wrote:

I'm using QTDragon under LinuxCNC 2.8.2 and my setup works pretty much as you describe.
See my posts in this thread It may help.
John

 
 

Ok, thanks. Looks promising. I will try tomorrow and let you know. Thanks for your response.

Can you point out what docs are out of date?

Chris

cmorley wrote:

Can you point out what docs are out of date?

Chris
 

Well for starters the text under 'probing' is not helpful at all.
linuxcnc.org/docs/html/gui/qtdragon.html#_probing It only mentions the possibility of tool length measurement (ATLM). I hoped/expected to see more guidance there for beginners like myself.

And further there are numerous examples scattered over the forums that do not work with QTDragon. I tried them all and no luck at a working solution for remapping M6 for automatic toolsetter offsets


Maybe this has to do with the definition of ATLM. Some examples only do a touchoff and do not adjust offsets, others go a little further.
Would be real nice if all the old not working examples would get a warning so people do not waste their time on those solutions. The level of frustration gets very high when there is no place to find help with something that should be very common by now, even for hobby machinists like myself.

I had a working installation of gmoccapy. Now most things do not work anymore and I had to start almost from scratch. If things changes so much that the 'older' way does not work anymore there is something wrong in my opinion.

Very grateful for the superb LinuxCNC software that is available here, I am not blaming anyone and no complaints about what is free to use. But some working examples for new stuff (is QTVCP new stuff?) would be very nice. It just doesn't work to tell a machinist to read through the intricacies of the hal layer just to get something working again that used to work perfectly. But that is just my opinion.

Was the on screen probe help dialog not helpful?

I can't really fix examples all over the forum - but I can document a working example in the qtvcp docs if one is shown to me.

toolchange routines are hard to document as it seems every integrator does things a little differently.

Chris