I've been bitten once again by the "project bug". I've got so many other projects designed, but don't feel like hand cutting parts, and I dread goofing around with chemicals to make circuit boards. Takes far too much work.
I've decided to throw my mind into constructing my first CNC rig. I've got a half of a 1 car garage, so space is essentially limited. I finished building up some very nice workbenches and cabinets (from IKEA), have a few power tools, but want to do better parts and have better accuracy over spray glue and hand cutting. As my current work status is "flaky", I don't have much money to put into this, but I'll scrape it up as I can while I'm in progress on building.
Since this will be a piece going on top of a rather sizeable table I purchased at a local store that was going out of business, the size will be somewhat limited. This is my first build too, so I'm going with trying to make it on the cheap. Shooting for $1,000 or less for the unit (sans controller as I have an old basket case PC to use) but I think I'm a tad over so far. Been trying to value engineer wherever possible.
Requirements: Needs to cut 3/16" plexiglass, some aluminum on occasion, MDF, plywood, particle board, as well as rout traces for the occasional circuit board. Cutting envelope is 24" (X axis), 18" (Y axis) and 3" (Z axis). I have access to a MIG welder and metal tools so I'm building this sucker out of tube, sheet and bar steel.
As I'm trying to keep costs to a minumum, I'm considering driving the X and Y axes with 1/4" timing belts while the Z axis will be a 3/8" - 12 ACME lead screw with dumpstercnc anti-backlash nut. Guides were going to be 3/4 extension heavy duty drawer glides, but I killed that idea when I realized the glides would stick out 2' on the back and one side of the unit making my footprint far more than I wanted. I'm now considering using (2) 1/2" precision shafts with Frelon lined bearings from McMaster for each axis. Head will be a Bosch Colt trim router (1HP adjustable speed). Holddown will be my shop vac sucking thru a piece of LDF with one face routed off and the edges sealed. Should be fine for occasional use. Longer milling runs I'll probably just screw the work down to save load on the vac.
With the timing belts, I don't expect much backlash. The lead screw for Z should be super tight and will keep me from having to come up with some sort of braking system. If I get any racking in the X or Y axis, I assume I'll add some cables (like the old drafting tables) at a later time to solve that. If the timing belts are less accurate than I like, I can convert to lead screws when I have more money. $250 or more for hardware per axis was far too much for me to swallow at this time. If I find something else in a surplus shop, I'll adapt to utilize it at that time.
I'm open to any suggestions on hardware or locations to get materials to get the costs associated with the XYZ drive down as much as possible. I'd also like the community's input on how much motor (stepper) will be needed to drive this at a reasonable speed with the given hardware, and any recommendations for locating steppers, drivers (thinking of using linistepper kits) and power (can I use a PC power supply?). The local surplus place is where I'll likely get my steppers unless I go for a kit that's cheaper. Will have to sort out how to identify steppers at the place though. Also, any and all criticism of the build is definitely wanted in order to keep my head on straight. If it simply won't work the way I think it will, please let me know how it is a problem and suggest alternatives.
The frame is about as beefy as I figure it needs to be for my application, and likely a little over-built. I engineer it how I think it should look. Looks good enough to me. The feet will have adjustment to level the unit out. As far as adjusting each axis, I've taken into account some areas where I can either drill slightly oblong holes for adjustment, or add shim stock where major components come together. I figure I've got the ability to true up most trouble spots on the rails and should be able to true up all three axes with some tweaking.
I think I'm pretty clear on setting up limit and home switches on the unit. Now when it comes down to calculating the drive ratios, steps and all that for the axes I'm probably going to have to come ask for a bunch of help. I'm ok with the mechanical and electrical bits, but when it comes to the computer behind the scenes getting it to go is a little out of my pay grade. I'm highly likely to be running EMC2 at the beginning. I have Autocad 2002 for CAD input. Now for CAD to Gcode I've not entirely made up my mind. Prefer to at least start with something free, even if it's fiddly as I have a low budget.
I miss the days when I had full access to (used to program) a KOMO 5'x12' machine with three routers, coolant and a multi-head boring machine. Nice piece of equipment.