Limit Switch Troubles - HAL Debounce

ctrl-frq wrote: I have not been able to run any G-code yet..... all my engraving files have been done in Signlab, which has an obvious IJK start/end radius problem.

I've never heard of Signlab, what is the IJK radius problem? It might be possible to parse through the files and correct it...

JT

Signlab, Engravelab, Profilelab.... all at this point, overly priced products from Cadlink. www.cadlink.com/

Been around for decades. Kind of like Cad/Cam would be if Corel Draw did it. VCarve comes close, but sadly VCarve STILL does not do certain things Signlab can do.

I believe my IJK error "issue" is covered in the Lcnc user manual under G2, G3:

It is an error if: When the arc is projected on the selected plane, the distance from the current point to the center differs from the distance from the end point to the center by more than (.05 inch/.5 mm) OR ((.0005 inch/.005mm) AND .1% of radius).

Signlab is not the only program that compiles G-code and produces this erroneous radius, and there seems to be quite a few people who get the error in Lcnc. The answer so far that I have come across for some is to try running with G90.1 (or G91.1, I forget which) rather than just G90. Right away, responses blame the cam program and not Lcnc.

I've got 20 years of specifically engraving files that I have run perfectly well with FC... and this machine is an engraver.......
www.google.com/search?q=vision+1624&biw=...CgAw&ved=0CAgQ_AUoAw

I was just hoping to come across a SETTING for the above Lcnc statement as to what constitutes an error or this will be a problem. I believe that my error is kicked by the second part of the statement of the ".0005" AND 1%. because the adjustment in FC is called "Arc RADIUS Tolerance" and I believe defaults to .01". I can tell you that whatever this amount of error is generated by the post in Signlab is not enough to actually SEE, even if I engrave very, very tiny letters (thus the .0005 and 1% potential culprit). Its never been an issue !

Bottom line is this is not my only engraver and parsing my existing files will not be a good resolution!

Not being a sign maker I have never heard of any of those programs. It's a shame they can't describe an arc to better tolerance. Maybe the arc tolerance is configurable in Signlab, if so setting it to a tighter tolerance might be the answer.

G90.1 and G91.1 have nothing to do with the arc offset tolerance.

Is the G code 2.5 D or full 3D paths? I'm just curious.

Do you have a short sample of a G code file that Signlab generates?

There is no setting that I know of to make LinuxCNC accept arcs with out of tolerance offsets.

JT

Well, for years, I did not use it only as a "Sign Program"... While it was designed specifically for that, but I would draw any of my mechanical needs in Cad, the import it and use Signlab as a Cam program. In many ways, it still does things a lot of other Cam programs fail to do, but I would never recommend it as a Cam program these days.

My problem paths in general are all 2.5d, and I have attached a snippet of something really simple that opens and runs fine in FC, but gives error in Lcnc. Now, unfortunately, in gmoccapy, I can not COPY that error message in order to paste it somewhere, but it is essentially complaining about a start and end point being some percentage off from each other.

I did not look, but I assume that yes, I might find a setting in Signlab to tighten things up in this regard... its just never been an issue ! I've been using Signlab for things since probably 1994 -95ish, and Flashcut since mid 1996. I ran Signlab files on Ah-ha controls before that, and before that, some proprietary controls.

I probably wouldn't know anything about such an error if it had not been for Flashcut occasionally popping up a warning dialog when the file must have had a similar error just outside the default tolerance. When I saw it in Lcnc, I immediately knew what it was complaining about.

There obviously IS a setting for this in Lcnc because the manual describes exactly when they will cause it to trigger the error we see on screen. Its just that the developer figures he knew what was necessary or best. As such, the end user just does not have access to such a setting unless of course he is a coding genius and can find the references and alter them.

It's not an end all problem for me, just that anything older has to be run on a different machine. Some of these files were created without G2-3 arcs and all short line segments. In the beginning, Flashcut ran thru short line segments faster than arc commands, so we just created files accordingly (that has obviously changed).

I did have troubles with the gmoccapy interface tonight.... the visual toolpath display would fail to show me the path of the file that I had loaded. In fact, I never got it to show the toolpath segment I have attached.... I opened it in FC and it displayed exactly what I figured it would.

No worry.. I have a LOT to learn ! There are many things I am used to having that I can't find... YET at least. But I am homing reliably now !

Should I start a thread asking specifically about the IJK error and the default Lcnc settings that we can not get at ??

Thank YOU for your help though.

Ah, but we (anyone) can get at any setting because LinuxCNC is open source.

Another thread might be good to discuss the source code.

JT

ctrl-frq wrote:

Should I start a thread asking specifically about the IJK error and the default Lcnc settings that we can not get at ??

BigJohnT wrote: Ah, but we (anyone) can get at any setting because LinuxCNC is open source.

Another thread might be good to discuss the source code.

JT

Try this one

www.linuxcnc.org/index.php/english/forum...-arc-tolerance#54186

regards