Yep, just like that. And a "mini Y" axis is a possibility too- the machine size is a little larger than you strictly need for the parts- say, a 700mm Y axis. That gives you room to cut off any uneven ends of the fresh material for proper feeding. Then you have an "A axis" that you "index" which is actually just feeding fresh material in. If you want to go to something resembling real production, processes like these are definitely worth thinking about. Since you've got an ESS controller, that would still leave you two spare axes and some IO for other things- automated clamping, feed pressure regulation, you name it. I have closed-loop NEMA 34s on my benchtop mill conversion. Super overkill, but delightful! I'd use spring (mechanical or pneumatic) tension to avoid as much manual adjustment as possible, if the idea is to run unattended. 1) Bear in mind you might get tearout if the fibers aren't supported from the spindle moving down. Different toolpathing (more plunging, less ramping?) may work, or compression cutting end mills, or whatever. Probably a few things to try, if it ends up being a problem. 2) The waste- unless it's the entire full width of the board would likely need separating from the main board so it can indeed be pushed off the end- or fall through, or whatever- and the main board re-dressed so it can be accurately indexed. This is less "a CNC gantry router" and more "a purpose-built custom board CNC processing mill" at this point.
Very interesting idea. You won't be able to cut all the way through and have parts drop out the bottom though unless you have a way to hold them in place until the bit is clear of the workpiece. On the final pass the bit would catch the side of the workpiece and push it around = broken bit and/or damaged workpiece. Would work very well with a laser. Alex.
Funny, I had the exact same thoughts, though a macro that engages additional clamping before removing the last two tabs or something wouldn't be a big deal. Or just leave it attached, zip across the board to dress it for the next feed, and then eject the whole thing for a small amount of finishing work. I wondered if a laser would work, but figured there are pockets or otherwise non-through-cuts involved.
yes, its very much turning into a custom machine! but i'm building it for a very specific purpose so, thats not really a problem. The 'manual' setup would be only to get an approximate 'guide' in place for the plank to 'ride' in on.. Once it hits the end, i'd almost cetainly use pnematics to clamp it in place.. Well actually i'd proably run it in reverse.. the clamping force woudl be provided by a spring, with pnematics to hold the clamps off. Kind of like pneumatic brakes. the tear out is a potentiall problem. but the waste board underneath is a bigger problem!
I had a PhlatPrinter. It was designed to cut foam but I cut wood all the time with it. The only downside with this arrangement was that you had to cut many passes to cut 3/4" wood. Something like 20 passes since you can't cut too fast and/or too deep. The accuracy of the machine, though, was excellent. I used a 8' foam board and measuring tape to calibrate the Y axis. Unless the rubber wheels press really really hard on the material the speed of the cut can not be increased or you'll lose position, do to slipping, with or without those expensive servos. The other option will be the toothed gear, as Rob suggested, but that will mar the material and will have to but cut off. BTW, since only the X axis moves (the Z axis move up/down) the PhlatPrinter had two rollers with a gap between them where the bit could move back and forth and you could cut through the material without hitting anything. Tabs, as they are usually used with a CNC, kept the parts attached.
Trying to acehive high accuracy, with the rollers pushing the wood through to create the Y motion is going to be way to diifcult. As described it seems it will be best to create an A axis that is used for feeding in.