[Stevos758]

Just installed path pilot last evening. I was getting my ATC and stuff setup and just threw some code on the machine to see if it would run. Seemed to function ok with my old mach 3/tormach post. However a few of my codes were generating an Arc radius too small error.

I did some research and all I could come up with was to turn your cam tolerances to a finer amount, Say .05" to .005".

Now I will try this possibly later but I want to get some discussion going to see if anyone else is having this issue.

[widget_maker]

Yes, I had the same error using the available post program on hsmxpress.
Switched to another post and that solved the arc too small problem, but now I have the G80 issue when drilling. (check the drilling thread on the main tormach forum)

[datac]

I think I have an answer for you.

I have LOTS of older toolpaths that worked fine on my favorite control, but immediately failed on LinuxCNC (not PP). I found that in my case, it was an error with arc start and end points being inaccurate.

I too probed around for answers, because I sure as heck am not going to repost all my old files with a tighter resolution (and the program that they were created with may not even be able to do that!), but it was generally blamed on the cam program settings.

SO, I got into the LinuxCNC forum and asked questions. I pointed out that my other control (flashcut) had a tolerance setting for this for practically 20 years !

Forum gurus over there admitted that indeed, someone at some time hard coded what the tolerances were to be in the program, and one fellow took it upon himself to ask the higher ups why the program DIDN'T already have an adjustment for arc end point tolerance.

Low and behold, in the next higher up forum those guys work in, they very THOROUGHLY examined the entire issue, concluded on new default values, and installed a tolerance ADJUSTMENT in the setup INI file. I do not think it took them but a week or two to get the change in the lineup for upcoming releases.

So, to the point, perhaps my issue and yours are a little different, but if your errors speak of start and end arc radius, I believe your issue has been solved, but its probably in a newer version of LCNC than Tormach has released !

You can see that discussion here: LinuxCNC :: Topic: Arc Tolerance (1/3)

[nickfabb]

Just installed path pilot last evening. I was getting my ATC and stuff setup and just threw some code on the machine to see if it would run. Seemed to function ok with my old mach 3/tormach post. However a few of my codes were generating an Arc radius too small error.

I did some research and all I could come up with was to turn your cam tolerances to a finer amount, Say .05" to .005".

Now I will try this possibly later but I want to get some discussion going to see if anyone else is having this issue.

I am also having this problem. I end up having to manually fix some lines. I am using Fusion 360 with the Tormach Post. I'd be interested to hear if your get it resolved on your end...

[Stevos758]

I found out you can edit your post and change the radius from R to I J and now I am just getting a few errors that some radius are .0001 off. I haven't ran the code yet but I'll let you know how it goes when I do. I also found you can use the emc post and it seems to work fine.


Sent from my iPad using Tapatalk

[jcposada]

I found out you can edit your post and change the radius from R to I J and now I am just getting a few errors that some radius are .0001 off. I haven't ran the code yet but I'll let you know how it goes when I do. I also found you can use the emc post and it seems to work fine.


Sent from my iPad using Tapatalk

Can you please elaborate a bit? I attempted to run my first part with Pathpilot today and encountered this error.

[Stevos758]

Can you please elaborate a bit? I attempted to run my first part with Pathpilot today and encountered this error.

If you find your tormach post in the posts folder you can open it up with notepad ++, In there you can edit a line changing from R Radius calks to I+J calcs.

I ran some code that still gave me some warnings about a few arcs being off by .0001 but I ran it and did not notice any difference.

[jcposada]

Thank you. I tried the EMC as per some comment I saw here or on the fusion360 forum and that seems to be working fine (no ATC here)

[KSky]

I am also having this issue with GCode generated by BobCAD/CAM v27.

The error is
G-Code error in DriversSideFullJig #1.tap
Near Line 199
Zero-radius arc: start=(X-4.8796,Y-0.3566) center=(X-4.8798,Y-0.3561) end=(x-4.8798,Y-0.556) r1=0.0005 r2=0.0005)

The GCode is:
X-4.8796
G03 X-4.8798 Y-0.3556 I-0.0002 J0.0005

What's the right solution here? Have BobCAD/CAM not generate tight radius? Have PathPilot not be so picky?

How do I manually fix the GCode?

[shred]

What is the program trying to do there? If that part is what i think it is, there's no critical tiny radius anywhere on it. You could probably just delete the offending line but I think there's a setting in the BCC post file but don't know what it is for V27

[KSky]

You're correct in that there is no critical radius anywhere, but it seems BobCAD/CAM really likes to use them as most of my parts have them. I have support cases open with both Tormach and BobCAD as I pay for support from both.

[tmarks11]

So I noticed something interesting.

Ran some code generated by solidowrks/hsmworks. Ran just fine. I needed to make a final pass deeper, and I edited one of the lines (which happened to be a G3) to change the depth from 0.1 to 0.2

PP spit out the "arc radius too small". Seems like it didn't like changing depth that much in a circular cut. Strange. Put the Z.2 in the line before it, and everything was good.

[cordvision]

I get the very same error, if I run this code (see pic for error message):

G00 G17 G21 G40 G80 G90 G64
N540 G01 X123.524 Y-103.171
N541 X123.521 Y-103.172
N542 X123.464 Y-103.162
N543 G03 X123.456 Y-103.168 I-.001 J-.007


The problem is, that I could run the code anyways, but the visual toolpath representation won't display any of the toolpaths that are in the file that creates the error message. This is really annoying. Any solutions yet?