[Uman]

If there is a tool offset diameter conflict between pathpilot and the CAM program, which value will be used during the cut?

[cbyrd]

Whatever is loaded in the tool table in cam is going to determine the tool path. You may break bits if they are really different.

The tool table in path pilot is only going to use the tool height offset. The tool diameter in path pilot is for the conversational.

Chris

[chuckorlando]

The exception to that is if you have an option in cam to allow the control to run cutter comp. I have never used it, not sure why I would, but I have seen the option in hsmworks.

[GLCarlson]

If there is a tool offset diameter conflict between pathpilot and the CAM program, which value will be used during the cut?

Some cited threads appear to be confusing LENGTH offsets with DIAMETER offsets (eg, reground tools) (OK, it's automatic cites, and computers are dumb- but watch out).

I have been searching without success for a way to enter diameter offsets in PP. PP claims to use the radius of the tool for D when D is not specified after a G41. I'm not sure it actually does, and PP complains about the G41 D(tool number) combination at least in some cases.

Does anyone know where PP stores the D(toolnumber) information? Or how to enter it? I have some programs that generate perfectly good code for M3 than I cannot make work in PP due to this issue- or something related.

[GLCarlson]

The exception to that is if you have an option in cam to allow the control to run cutter comp. I have never used it, not sure why I would, but I have seen the option in hsmworks.

It's one way to use sharp V cutters for threading, rather than form cutters (which Pathpilot assumes as a standard).

[chuckorlando]

Good Call

[GLCarlson]

Good Call

It would be, if I could figure out how to make it work!!! At the moment, I can't get any of my formerly-totally-reliable code generators to work under PP. I am imminently in danger of having to go learn how to code this manually so I understand it. Or regressing to Mach3 to get this job done.

[chuckorlando]

It has to be a weird code that PP needs.

[kstrauss]

What is the error message/problem that occurs? Post the code here and perhaps some of the gurus can point out the fix.

[GLCarlson]

What is the error message/problem that occurs? Post the code here and perhaps some of the gurus can point out the fix.

I'm using a commercial gcode generator (Advent2008) which seems very good. It can produce both I/J and R helical code, for pretty much any known thread (UN, ISO, NPx, or custom spec) using either sharp V or threadform cutters. It has very flexible control over start/stop point, lead-in/out, material speeds/feeds, internal/external, pretty much anything one can imagine. I've used it for 5 years or so, under Mach3, and it just runs. It comes with posts for Fanuc and multiple other machines. The Fanuc 0M post has worked well with the 1100 controlled by Mach3.

When I switched to Pathpilot, incremental (G91) code started in random places, and PathPilot complained -wouldn't run, actually) with tool diameter compensation reset (G40 D0) and with compensation turned on, generated threads of the wrong depth (G41 D(toolnumber) when sharp V correction was used. That is the remaining problem I'm working on- wrong depth of thread. It's obviously a function of some calculation in the code interacting with PathPilot. Absolute (G90) code had similar problems. Both incremental and absolute reference frames worked fine in M3.

Other errors (starting in the wrong place, commanding a G1 move without an F word, stuff like that) in the generated posts are pretty obvious and easy to fix in the post generator. Ditto, using G40 with a D word in the safety block (G90 G0 G17 G40 D0 G54 etc) generates an error in PP, and must simply be a rule violation in the LinuxCNC implementation because without D0 there's no problem. I'm fairly close, I think, to having a PathPilot post for the Advent2008 threadmill gcode generator.

All of this is to explain that I'm not an expert in G code, don't have all of the works in the Advent executable, and may be chasing something that isn't part of the problem. But- it appears that however Tormach is implementing diameter compensation in PathPilot, it is different from what happens in Mach3, and that's what I suspect is causing the wrong depth of cut ultimately.

I can and am tempted to simply hand code the current problem. I am talking to Tormach about this, since one reason for chasing it is the very limited number of threads in Conversational (and, at the moment, no way to add more- there's a mechanism that doesn't seem to work, another point of discussion with Tormach). The other reason for pursuing this is that Conversational presumes a threadform tool; there's no published way to use a sharp V cutter with Conversational, and Conversational has some other quirks as well. Essentially, if you have the right tool and need a thread it has in the table, it works great. Otherwise, not so much.

Right now, I don't think I have enough of the pieces pinned down to post code or even define the issue better for the forum. I'm reluctant to waste our experts' time with an ill defined, idiosyncratic, and possibly limited value problem. When I can definitively show code that works in M3 and blows up reliably in PP, I will post it. Or maybe post a solution. This has become a bit of a quixotic crusade.

If anyone can point me to details on how LinuxCNC handles tool diameter compensation -or better yet, something beyond the Tormach published info about how PP handles the problem, that would be a big help.

[MFchief]

What happens if you simply put in the TPI, major and minor diameters that you want?

I did that for a 3-48 internal thread and it worked OK. It did reset the 48 TPI readout to 48.1 for some reason, but since my thread was only 3/16" long, I didn't think it would matter. I did use a single point threading tool from maritool, which I belive is what you mean by a "form tool"veresa a sharp point vee tool.

Terry

[GLCarlson]

You're absolutely right. You can simply put in the size you want, and Conversational does work. I was a bit too grumpy about Conversational. Long day.

What I can't do is add to the list of stored fonts in the User folders, or use a sharp V cutter. Maritool sells form tools (and are a terrific supplier, too). I did try fiddling the diameter of my tool, and got a sharp V cutter to work, kinda-sorta, but it was very much cut and try. Not an analytic solution ("do the math and it works").

And I do like the ability to use the Advent program for NPT etc. I admit I'm a bit manic on the topic just now.

[MFchief]

Yep I get it, the sharp vee cutters are what I started with on mach 3 and they worked pretty well. Definitely cheap also.