On Dienstag, 14. April 2020, 19:07:32 CEST Chris Morley wrote: > On 2020-04-14 9:10 a.m., andy pugh wrote: > > On Tue, 14 Apr 2020 at 03:58, Reinhard <[email protected]> wrote: > >>> Maybe G43.2 could be tweaked to take either a H number or axis codes. > >> > >> That would be very nice indeed! > > > > Take a look at the andypugh/G43.2-direct branch and see if that does > > what you expect. > > > > Looking through the code, it is clear that G43 / G43.1 and G43.2 all > > completely ignore tool diameter. This means that G43.2 is not actually > > useful for wear offsets on a milling machine (as far as I can tell) > > but is useful for a lathe. > > (Which is probably the intended use) > > Thats exactly what it was, a hack for lathe wear offsets.I think we > should try harder to get out of the work around idea and do things > properly.
Doing things right is a good intention! But relative offsets are not a wear replacement. It could be used in the same way, but the first goal is to fake a toolsize for roughing. I used that at lathe and stil use it on milling at work. If I remember well, Siemens has cycles where you can add offset for roughing. At work I need to create fake entries in tooltable, which is risky when others work at the same machine too. So having GCode that can apply a relative offset is safe and helpful. Another (big) challenge is all around wear offsets. This is completely different and here I support the idea of first thinking/talking about a good solution. If you want to implement support for wear offsets, its fundamental having a timebase to count worktime of each tool. Then a wear-offset becomes meaningful. At work they run performance tests for all tools, so the cutting volume, lifetime and price of each tool are calculated. Result is standard tools for every machine and the knowledge about lifetime of a tool. So if you know, that a certain tool has a lifetime of i.e. 100h it would be nice to know, how much time it already has been used for cutting. Reinhard _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
