From: Jeff J. <jo...@su...> - 2014-05-30 13:42:20
|
As a long time lathe operator and shop owner I hate to see the wheel re-invented. Operators are taught and used to T#### the first two digits are the tool number and the second two are the off-set number. I am not saying that your thoughts are not more flexible but it goes against 40 years of standard cnc lathe practice. Last two digits call Offsets from wear or adjustable off-set page with a column for X and a column for Z. I know you understand it but I cannot understand the need to go against convention. My oldest commercial control on the shop floor is a Cincinnati Milacron 900 1983 Version and it operates this way and my newest commercial control is a Fagor 8050 2006 Version and it too uses T####. I also have three turning centers running LinuxCNC utilizing the tool patch. I would really to see this developed further with separate pages for fixed offsets (geometry) and assignable offsets (wear) as I have eight turning centers and to re-train operators how to run lathes does not appeal to me. Please explain to us lathe users why we have to move away from what has been a standard and works very well. Jeff Johnson jo...@su... Superior Roll & Turning 734-279-1831 -----Original Message----- From: Sebastian Kuzminsky [mailto:se...@hi...] Sent: Thursday, May 29, 2014 4:45 PM To: EMC developers Subject: Re: [Emc-developers] RFC: Support for multiple tool offsets (aka wear offsets) On 5/29/14 14:37 , John Thornton wrote: > If I understand this say you have T100 as -0.001" in diameter you can > do > G43.2 H100 to make your tool effective diameter 0.001" smaller for > G41/42 paths? That's how I read it. And you can say "G43.2 H100" twice to make it 0.002" smaller. Seems handy. > > JT > > On 5/29/2014 1:59 PM, Chris Radek wrote: >> I think the general solution to the specific problem of wanting >> separate geometry and wear offsets for lathe tools is to allow the >> gcode to request any number of tool offsets simultaneously. >> >> I've implemented that in a very simple way in the cradek/multi-tlo >> branch on git.linuxcnc.org, and here is the basic documentation for >> this feature, as implemented: >> >> To use tool 1 with the sum of offsets 1 and 10, program: >> T1 M6 >> G43 (or G43 H1: replace any existing offset with T1's) >> G43.2 H10 (also add in the T10 offset) >> >> You can sum an arbitrary number of offsets by calling G43.2 >> more times. There are no built-in assumptions about which >> numbers represent geometry and which are wear, or that you >> should have only one of each. >> >> Note that having multiple offsets active interacts with "touch off" >> in the UIs (aka G10 L10/L11) in a possibly-unexpected way, because >> those write only to the loaded tool's offset. (This was also >> previously the case when you are in a situation like T1M6 G43H2, or >> G43.1 mode) >> >> Frankly I don't know how that can be fixed without loss of >> flexibility. I think if you are using multiple offsets maybe you >> ought to be editing the tool table directly to keep them straight. >> >> I'd welcome any comments about this scheme, except please note that I >> already know this is not how Fanuc does it. I do not like the >> restrictions and assumptions built in to their scheme and I believe >> this implements a superset of it. (But to our remap experts: could >> you use remap to emulate it?) >> >> Chris >> >> --------------------------------------------------------------------- >> --------- Time is money. Stop wasting it! Get your web API in 5 >> minutes. >> www.restlet.com/download >> http://p.sf.net/sfu/restlet >> _______________________________________________ >> Emc-developers mailing list >> Emc...@li... >> https://lists.sourceforge.net/lists/listinfo/emc-developers > > > ---------------------------------------------------------------------- > -------- Time is money. Stop wasting it! Get your web API in 5 > minutes. > www.restlet.com/download > http://p.sf.net/sfu/restlet > _______________________________________________ > Emc-developers mailing list > Emc...@li... > https://lists.sourceforge.net/lists/listinfo/emc-developers > -- Sebastian Kuzminsky ---------------------------------------------------------------------------- -- Time is money. Stop wasting it! Get your web API in 5 minutes. www.restlet.com/download http://p.sf.net/sfu/restlet _______________________________________________ Emc-developers mailing list Emc...@li... https://lists.sourceforge.net/lists/listinfo/emc-developers |