Hello, Rob Jansen wrote:
> Robert, > > Being new to g-code and cnc I've been trying to learn the best way to > think about the coordinate spaces and offsets etc (for a cnc mill). > > I'm thinking I should use a **fixture** plate with alignment pins > etc to > mount to the table. Some kind of pallet plate to hold the work > (plugs into fixture). > > > This depends on the type of work and the type (and size) of workpieces > you have. > I make mostly small parts (a few inches long with a max. of 10~15") > sometimes in small series (10 pc) which I then mill out of a larger > block by moving the home position using one of the user selectable > coordinate systems. And I am in the process of adding a rotation axis > which mostly be used for N-sided milling of small workpieces. > Production time is not critical to me so it does not matter if > mounting a new workpiece takes 10 seconds or 10 minutes > .... > Palleting systems can be very complex. The simplest one have one > pallet position and complex systems may have multiple pallet positions > with their own home coordinates, or an automatic pallet changer with > multiple pallet positions. These can be as hard, or harder, to control > as an automatic tool changer... The pallet system I'm envisioning I'd make myself (I'm assuming I'm using the term "pallet" correctly). The fixture (to hold the pallets) would be nothing more than an aluminum plate with alignment holes in it. This is for a classroom machine (Syil super X3) so we have many kinds of parts we want to make. In one case we'll be making 14"x5.0"x0.125" aluminum plates with a number of holes (at least two different sizes). When I think about the steps I'd use to set this up I keep coming back to a fixture plate and various pallets (for the various other parts) as being the most efficient. I definitely like the idea of not having to dial in a fixture (non-pallet design) everytime I clamp a new one down. I definitely like the idea of designing g-code procedure that assumes a master offset from the base-fixture (pallet/base-fixture design). Again this is all new stuff to me. I'm not sure about any *conventional wisdom* so I'm sorta making this up as I go along. > > My current thinking is that I would home x and y to the lower left of > the table. For z I'd use a spacer block off the table to the end > of the spindle. > > From there I'd use global offsets use a global offset (G92) to > align to > the fixture. From there I'm guessing one whould use the > users-selectable coordinate > systems to align ... > > > For a manual pallet changer manual homing may be satisfactory. Hmmm, so you're suggesting in this case that one would not home to the table but to a fixture? I'm assuming this is what global offsets are for (g92). > A pallet changer may use the same home position for all pallets (the > pallet is physically being replaced). But if you go for multiple > pallets on one machine your suggestions look like what I would try - > but I never used a pallet system before. > If your machine has a probe tool that is usable with an automatic tool > changer you may even probe for the proper home position - comes in > handy if your mounting system does not have a mounting precision that > is needed for your operation. > > Tool tables looked kinda fancy so I was going to defer using them > until > I had a better mental foundation of the system, using some kind of > programatic offsets instead. > > > Depends on your software. I mostly use Filou or DeskProto (before > anyone flames me - these are windows applications :-( ) and both > include their own tool table so tool offsets are calculated by the > software. But I also do some hand-G-coding sometimes and > miscalculating tool offsets cost me a mill-bit or two ... Based on my experience with trying to use cutter-radius-compensation and the apparent depth of topic of the tool-table system I feel a bit overwhelmed so I chose to simply code in tool compensation in my g-code manually. I planned to use this along with a calibration procedure written in the g-code with each (manual) tool change (using an offset block jig of some kind) and the quill feed and lock. It's kinda crude but it fits my way of looking at things until I get more savvy with cnc methodology. Thanks for your input. Robert W. ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
