I have a PDF copy of the Fanuc oi-mc manual and I read the section 14 on tool length compensation and it does not say anything about storing the actual tool length from zero (home position). It says that all tool length compensation values are offsets from programmed position to actual.
There is a G37 tool measuring cycle that measures tool length probably via a tool probe, but the end result is a correction to the tool offset table, the actual value from home( zero) position is not recorded. I can send the manual to you (or anyone else) if you would like to see it. It is 4 megs in size and 900+ pages so it is very detailed. Dave (Dave911 or the IRC) On 4/22/2010 6:10 PM, Wes Johnson wrote: > Hi everyone. This is going to get wordy. We were discussing on IRC how EMC > handles tool touch off for the tool table (length). > > I attached the figure. If you cannot see it, go to my forum post here: > http://www.linuxcnc.org/component/option,com_kunena/Itemid,20/func,view/catid,21/id,2653/lang,english/ > > So here goes > > Tool length offset in EMC2 > > I am hoping that I can make this more clear. EMC is functioning correctly. > However, the functionality does not conform to the industrial CNC control > standard. The methods described below are for a vertical mill, other > machines and configurations will work in a similar manner. > > For the purposes of this discussion: > Absolute Coordinates = the position of the machine with respect to its > home position for any given axis > Relative Coordinates = the position of the machine with respect to the > currently active work offset (G54, G55,....G59.3) > > The current EMC method for setting up tool length offsets by "touching off" > works as follows: > > 1) A tool is loaded into the machine (M6 T1) in MDI mode. > 2) That tool is jogged to some know location on the machine (usually the > location of the desired active work offset). > 3) Using the "touch off" commands in Axis, the machine records the current > position relative to the active work offset and enters this value into the > tool table as the tool length. The value in the tool table ensures that the > tool height offset for the current tool with be correct for the active work > offset. > 4) Additional tools are added using the same method. > > Why this is good: > -If the tools are touched off relative to the work piece, then having G54 > active and touching off tools relative to G54 causes no problems, and saves > a step (more on this later). > -This also allows one work offset to be designated for tool setting (G59 > could be used for the location of an automatic tool setter for instance) > > Why this is a problem: > -If tools are setup on the table or some fixed machine position, and the > G54 Z coordinate has been changed since the tool table was originally > loaded, any new tools added to the table will now "touch off" relative to > the new G54 Z coordinate, and the tool lenght will be wrong by the amount > that the G54 Z coordinate has changed. > > > The method used by most industrial controls (Fanuc, Haas, Fadal, etc.): > > 1) A tool is loaded into the machine (M6 T1) in MDI mode or using a built in > utility. > 2) That tool is jogged to some know location on the machine (usually a > sliding gauge block or tool length probe). Using the tool table in the > control, the active tool is selected, and a measure command is used to > record the current ABSOLUTE Z axis position which is then entered as the > tool length by the control (you can then add in the height of the gauge > being used if you like). Tool 1 would have a "length" of -220 according to > the figure. > 3) Additional tools are added using the same method at the same gauge > location. Tool 2 would have a "length" of -210 according to the figure. > 4) After the tools are set, one tool is chosen to set the work offsets by > "touching off" (any tool will work as long as its length offset is active). > (See included figure) > > Why this is good: > -Tool length measurment is done independently of any active work offset, > changing offsets will not impact the ability to add or change tools in the > tool table. > -As long as tools are measured from the same know location on the machine, > the tool setting method will always be the same. The operator does not need > to know which work offset was active when the tools were originally set. The > tool table only needs to know the DIFFERENCE in length from one tool to > another, so this method is prefectly adequate. > -If a tool breaks, a replacement can be set up without concern for the work > offsets. This is important if the tools were set using the top of the work > like one would do with EMC now. If machining were to be done on the top of > the work and a tool broke, a new tool could not be set using the top of the > work as a reference, because this surface was machined away. > > Why this is bad: > -The only problem is that this is not the current EMC method, and changing > this behavior may inconvenience some users. > -Most popular controls can be changed to set tools like EMC by changing a > control preference, this is rarely done. > > > Where do we go from here? > > It would be great to have the Axis interface enhanced to add an option to > touch off the tool lenght using ABSOLUTE coordinates. This could be a simple > radio button available when the "tool table" touch off is active. This way > users could set tool lengths using RELATIVE or ABSOLUTE coordinates. There > are a number of possible work arounds for this issue, but a simple > enhancement to Axis would eliminate that need. > > The current EMC method works, but is very confusing to those migrating from > more popular controls, and is very poorly documented. Setting tool lengths > relative to the current work offset is unexpected behavior for a CNC > control. If we wish to expand the use of EMC, it seems only logical that we > would want to standardize its behavior with the industry standard (I have > never seen another control setup to work like EMC currently does). I have no > doubt this can be done without losing any current functionality. This is > just a new option. > > Thanks > > -Wes > > > > ------------------------------------------------------------------------------ > > > > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users > ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
