Well like so many people after seeing Dave DeCaussin's ATC on you tube I had my ďI want that!Ē moment.
The thing that put it over the top for me was being able to regrind the r8 to work with the BT30. I settled with using the 4 ball gripper design and Belleville washers for the drawbar pressure. My quill is locked and I use the knee for the Z axis so that made things a little easier. All of the drawbar parts are A2 that I hardened and tempered. The drawbar has about 1000 lbs of pressure. Iím using a 4Ē Bimba flat multi power cylinder. I needed the multi power as I donít have a ton of travel on my Bellevilles. Iíve been using the drawbar setup since July and have had zero problems. I took it apart about 2 months ago and it looked perfect.
When I started working out my design for the ATC I was going to have the carousel on the end of the table and have the X axis bring it over to the spindle. I had a job come up where I really needed some sort of changer so I knocked together a quick tray type. One thing became immediately apparent was that I was running the crap out of the X axis. So out went that idea and I settled on a more traditional ďfly it inĒ design. Over the course of time I settled on a 3 stage system. One stage down to line the forks up with the tool. one stage to engage horizontally and one stage to bring the tool down out of the spindle. There are proximity sensors at the end of each travel. The M6 macro is written so that it operates in a follow the leader fashion. When the first stage proximity activates it triggers the next stage. This works really great and is very safe. I opted to use slides from cncrouterparts. They are inexpensive, well built and really versatile. They are designed to work with 80/20 extrusion and 1/4Ē CRS flat bar. Iíve used them in a couple of other projects and I was really happy. The carousel is driven by an extra servo motor I had and is designated as a C drive. Spindle indexing was supposed to accomplished with the new Delta VFD-M I purchased but unfortunately it didnít play nice with my 2 pole motor. I ended up going with a shot pin arrangement until I can afford a new motor.
Terry (BR549) wrote really slick macro for me that makes it all work. He has the spindle stop then turn slowly to the index signal. It then kills the spindle and shoots the pin. Another neat thing he did was instead of having a fixed position for the tool change he uses the ďtool change positionĒ DROs on the setup screen. This allows me to position the tool change at the most efficient spot for each setup.
Iím a big proponent of working out designs in 3D before committing to actual construction. The first thing I did was model my mill as closely as I could. I included the table and a couple of Kurt vises. You can download 3D files of the vises directly from Kurt. This allowed me to run the table through itís motions and get a better idea of what kind of envelope I had to work with.
Then I proceeded to model the rest of the components as I worked out all the motion and clearance issues. Although his kind of modeling takes time I find for me the end product is so much cleaner. Plus I build it a few times in the computer so when itís time for actual construction I have a good idea of what Iím doing. Most of the purchased components offered 3D files so this made things a lot easier.
Even 3D files of the pokeys were available.
The build was fairly straight forward. Most items were made from .5Ē thick 6061. I have a small anodizing line so I anodized everything. The extra holes in he parts are for mounting things if necessary. When I prototype I put extra drilled and tapped holes all over the place so that if I have to add something itís not a big deal.
I opted for a traditional fork design with tension spring to retain the tools. It took a few designs to get it to slide over the balance holes drilled in the tool holders.
I also engraved the tool slot numbers in the edge of the carousel. Not sure how well this will work when I get the cover on.
Using the 80/20 extrusions and cncrouterparts slides made designing things like brackets much easier to build. They have 3D models available for all their components.
Iíve done a few thousand tool changes in an actual machining environment and only had a few bumps and those were caught by the proximity system.The things Iím going to change are pretty minor. The main vertical movement is too short. The tools collide with my splash guard mounted on the table. I have plenty of height and the higher I can get the tools the less chance chip contamination there will be. Coincidentally I didnít model the splash shield in my original model and now Iím paying the price. I also have a bit of tidying up to do but Iíll do that after the vertical mod.
Having an ATC has opened up a whole new set of things I can bid on and has already paid for the price of the parts in the few jobs Iíve done with it. Plus it was a gas to build and that in itself made it worth it.
Here's a couple of videos. The first one is a normal change with a 6000 rpm spindle speed and spindle indexing. The second one is an earlier macro before the indexing was hooked up but it shows the speed of the change.
Thanks for looking