Having gotten to the point of getting way covers on my Tree 325 CNC mill (converted from Dynapath Delta 20 to Centroid) so it would be unlikely to ingest any chips, I wanted to wire in the receptacle for the auto-tool setter and digitizing probe on the Centroid control. A small aluminum bracket for the receptacle will bolt onto the monitor arm using one of the 6mm bolts that holds that arm to the cabinet arm.
This is a small AMP connector that would be bolted into place with four .125" bolts. I could have just drilled a few holes by hand (probably with the small ones oversized for wiggle room) and then hacksawed off the chunk of bar stock, and then finished up with a quick rounding of corners with the big file. But I figured I was ready to try and make the bracket on the mill, especially as the tiny holes needed would be pretty easy to hand drill far enough off that the holes in the bracket wouldn't align with the holes in the connector.
I used the conversational programming (Intercon) in the control, as I need a lot of practice with Alibre/Rhino/Visual Mill before I can quickly draw 3D solids and generate tool paths. But this was a 2.5D part, and it seemed to work just fine.
It took me ALLLLLLLL afternoon, but I got it made. 5-6 hours of work got me a 1.75 x 2.75 x .3125" bracket. It turned out to be one of those "learning experiences".
One of the things I learned was that cutting slots with a 3/16" end mill can generate enough force to pull out any slack between the sides of the flat on the end mill and the set screw on the end mill holder. When you've only got the tool retracting to .100" above the stock (in un-slacked form), this makes for a "now why did it rip a gouge across there this time?" moment.
This in turn gave me a chance to learn how to use the facing routine in the control. It also convinced me to go back and add in a bit more clearance over the part before moving the endmill above it in the X or Y axis.
Considering that at this time I have made no entries into the backlash table, accuracy seems pretty good. There does seem to be some indication of the.0007" backlash in the X axis, but my dial caliper shows pretty much spot on or within .001" of the number I was shooting for on all dimensions. The .001" was on the X axis. The Y and Z are showing .0002" backlash.
The face of the plug seems quite flush to the surface of the bracket, and there's just a couple of thou clearance between the plug and its pocket.
There was a lot of duplicate cutting, as I'd go back and find something like I'd forgotten to allow for the angle on the end of the drill bits, so just the end of the point had broken through the bottom of the stock, or that I'd gotten a connecting radius on all but one corner. I'd edit the offending entry, and run things again from scratch, watching the tools move and making all the tool changes, even though it was basically just cutting air.
Towards the end I'd modify the program to delete the completed work and jump to the part of it I was changing (or had forgot to put in to start with).
But no drills or endmills were destroyed making this.
I do need to try to make some table enclosures so I can at least run the Trico Microdrop (if not flood coolant) to blow the chips out of the tight grooves. I don't know that I'd ever had an occasion to use anything smaller than a 1/4" end mill before, but the 3/16" sure ended up pushing a mound of chips down the groove before it when cutting the outside of the bracket from the bit of scrap plate I was using. But the plate was small enough that I didn't have a lot of room available to run a larger OD end mill for cutting it out and still clear the clamps.
I used the bolt hole circle and peck drilling cycles for the .125 and .250" holes, square and circular pocket routines for the pockets, and then plugged in the outline with cutter comp and a depth repeat subroutine. All corners are radiused, and none of the internal corners are the same radius as the end mill.
I even got my sweetheart to come and watch the machine cut metal. I couldn't convince her to stay for all the multiple passes needed to get to depth though.
This 2.5D stuff seems relatively straight forward, and once I get to the point where I know the things to watch out for (like those drills not breaking all the way through) I think things will go faster.
Sure, I could have done some of this a lot faster with a manual mill and a few X/Y coords for drilling the holes. The outside radii could have been sawed/filed/disc sanded to be "good enough". But if I'd had to have them exact, requiring me to use a rotary table to do the five outside corner radii, setting that up would have taken a lot of time.
And of course, if I need another bracket for some reason, I can just pull up the program and run it again instead of having to crank the handles all over again myself.