All, I recently moved which required me to dis-assemble my standard OX machine. I'm in the process of building workbenches, installing shelves, re-wiring the lighting and outlets in my new shop. At the same time, I'm trying to plan the rebuild of my OX. I'm 99% sure that I will be eliminating the belts and pulleys and installing linear rails and screws while also beefing (OX-ing) up the Z. I'm looking for recommendations of successful rebuilds from belts to rails and screws. I need advice on the plate design, screw location and type. P.S. The reason I'm looking to switch is mostly precision. I mainly cut balsa and ply for RC aircraft and I'm a little OCD-ish about parts fitting perfectly. I've found that when cutting a fully nested 24" x 36" sheet of 1/8" plywood parts things stretch a little from 0,0 to the the extents of the cut area....just a very little. But noticeable when the parts are removed from the sheet, cleaned up and ready to glue. Also, I attempted to cut 1/8" aluminum motor mounts and have very little success, seemed like the gantry was all over. Josh
I'm planning the same on my 1000x1500 OX. Up to now here's what I tought of: X axis - C-Beam X axis with a 2080 glued/screwed directly to the back - Linear rails (20mm profile) mounted on the front of the C-beam. - 16 mm lead screw mounted on the top of the X-axis Y axis - I will leave it with belts for now, as the project will be on 2 or 3 phases, but I will plan the new plates to use 25mm profile linear rails Z axis - Ultra low profile 20mm linear screws mounted on a C-Beam Electronics - Will surely use the new, spectacular, unnamed for now, all mighty driver/controller from OpenBuilds (wink wink @MaryD , no pressure) I already upgraded the router for a 1.5KW spindle with a standalone watercooler. I will add more braces to the bed.
These guys seem to have decent prices - HGW25 Hiwin Flange Block Linear Guides | Motion Constrained | USA's Top Hiwin Linear Motion Guide Rail and Block Store
Have you considered using a rack and pinion setup instead? I'm rebuilding my own CNC router and am moving from a lead screw setup to R&P because I wanted speed as well as precision. I've got C-Beams for linear motion + CNC Router Parts pro rack and pinion for the driver. My custom steel plates work perfectly and I'm pretty close to final assembly/testing. If you do want to go a R&P route, I'm happy to share my experience and ideas/gantry-plate-design. I could probably have an extra set of gantry plates cut out if you want as well (they cost me $55 per set or so).
Where do the C beams come in? What are you using for linear motion? Speed AND precision would be awesome!
The CNC Router Parts setup has a 3:1 reduction for their R&P driver (at least with NEMA 23, but there is a NEMA 34 option as well with a different ratio): PRO Rack and Pinion Drive, NEMA 23 | CNCRouterParts I used M5 screws and fender washers to mount their rack to C-Beam: PRO Gear Rack, 39" (1 meter) | CNCRouterParts The rack sits inside the C-Beam and the gear moves along it. I then attach their driver system to my custom steel plate. It has a solid connection point with a bearing for pivoting. And then a second connection point with a spring/screw for tensioning the rack to the gear to prevent backlash. You can see some plastic plates I was using for fit testing in the attached files. The rack is inside the C-Beam which gives it a bit of protection from any debris coming out of the machine. And it can be mounted on the top facing down to prevent accumulation.
I should add that one disadvantage is that I had to attach a 20x40 beam on top of the C-Beam (making a 40x100 combined beam). This is because the flange of the rack needs that space to prevent it from interfering with the wheels. The other disadvantage (so far) is that everything gets heavy quickly with steel plates, steel racks, and the gear driver assembly. Heavy is good when cutting hard materials. But it also might make my motors strain more. I'll be posting a build thread at some point soon and talk about how things go when I get to the final assembly.
Regarding improved precision, I've worked up a new calibration procedure that I will follow for the new router. I'll have to let you know how it goes. (1) Use tinyg instead of grbl for the controller. This will let me parameterize and home each axis/motor independently (including having separate parameters/homes for the two Y-axis motors). (2) Measure and calibrate each axis independently without cutting. To do this, clamp a piece of aluminum near one end of the travel. Then use the electronics to move as far as I can away from that aluminum. Using a combination of gauge blocks and a caliper it is reasonable to very accurately measure this distance to about 500mm-700mm down to the nearest tenth of a mm or better. I can use this to get the precise parameters to set the movement. I will also be measuring and calibrating each Y-axis independently from each other before adding the X-axis. (3) Use a laser square to accurately square the table and accurately align the X-axis gantry along the rails. Then set my two Y-axis homing stops to always come back to this position. (4) Use a dial indicator to precisely capture the runout of my spindle/bit and use that as the bit size. (5) Level the bed at the end to ensure that there are no Z-axis surprises. After all of that, I hope my new machine will be as accurate as the underlying physical parts are capable of handling. -D
Clever design! I like the use of the gear inside the cbeam. I think I'm pretty set on eliminating the wheels in favor of linear guides though. If you are willing to share your plate designs, it would give me a great starting point, a little redesign to allow for Hiwin guides. I would keep them to myself of course.
As a followup, I've got a working setup now with CNC Router Parts R&P combined with C-Beam rails. This uses the plates I linked you to. Rapid speed is up to 5000 mm/min (I think the limiting factor is my xPro 12v driver setup). I have just started leveling my bed. The only down side is that my travel is relatively restricted. I am using 1 meter C-Beams. X-axis travel is just a bit over 650mm, and this is the most that can be had without longer beams. Y-axis travel is 590mm but could be increased up to 750mm by changing how I mount the Y-axis C-Beams. -D
Yesssss! One of the perks to being in the commercial construction business is meeting so many great people who work for awesome companies. My three "Sample" pieces of heavy T-Slot showed up today, very exciting! The beefy TS30-60 x 750mm piece will be the X gantry, and the 2 1100mm TS20-40 are for the Y axis. I plan on attaching a single 20mm Hiwin linear guide to each Y rail and a pair of 15mm Hiwin linear guides to the X. As of now the Z will also be a pair of 15mm linear guides attached to a 400mm piece of 20 x 80 from OBPS. I'm starting to get anxious about leveling the linear guides on the extrusion, any thoughts?
