Edit Jan 24, 2018:
I've finally completed a test assembly, and corrected a few issues that I had with things like endstops and some parts that needed adjustment in order to have some success with assembly. I also added provision for a cable drag between the heated build platform and the chassis so that wires don't get pinched. This takes the form of a tab on the Y nut holder and a small piece on the feet to which the drag chain can be attached. This is not incorporated into my test build yet.
At time of writing, I have the frame and axes assembled, and have the Marlin 1.1.x firmware configured to move the motors at 400 steps/mm (4000steps/mm for the Z axis)
By happy coincidence, I was able to use the existing frame to hold the gantry. This was not intentional, it just happened to work out this way. I had intended to (And still intend to) change to a box frame. The problem is that the feet were not entirely appropriate as they were to hold the frame at the correct offset to get the correct printable area. I added some spacers to the feet file in order to handle this, though I'll use a different method for testing, and I may change my mind for the full production model, provided that I get good results from testing.
Original summary follows:
Redesign of my 3D printer when I got sick of problems with belts and tensioners.
Redesign to make use of T8 leadscrews instead of timing belts. Still uses Nema 17 motors. You probably don't need to use the feet if you have a RepRap I3, that can be left alone. The motor holder file should be all that's needed to convert the Y axis.
You will probably need a complete change of the gantry in order to use the X and Z files, but I'm annoyed enough at my printer that I don't care.
I've included endstop mounts for switches with mounting screws 0.75" (19.05mm) apart, with the switch perpendicular to the mount. The only provision is for a 3mm screw/nut on one of the Z axis motors, centered 15mm below the top surface, alongside the smooth rod. This should be easy enough to rig something up, and I'll put together something with a reasonable range of adjustment when I get this put together. Of course it will depend on how you mount your print head for what's reasonable for a range of adjustment.
NOTE: I have not yet finished building this, so this should be regarded as Beta, if not Alpha. I believe that this is ready to try at least, and I intend to use it as-is.
Based off of the Prusa I3
Long story short, I bought an I3 printer kit, and it was terrible. Poorly printed parts, nearly impossible to keep the frame from twisting every time I walked past the desk, and belts that seemed like they were a different length every week. I spent more time leveling the bed and calibrating the steps/mm than I did actually printing with it.
I re-did the feet in order to try keeping the frame square. I also added in parts to keep the gantry frame square to the Y axis, so that X/Y would remain at 90 degrees. Z remains a slight problem, which is the reason that I will be going to a box frame in the near future.
I had trouble with the print head not remaining stationary relative to the carriage, which is no end of a nightmare. I'm moving to a mostly metal solution, which is still in the process of being fabricated. I may or may not go to a Bowden setup, still making that decision. I have some options as far as mounting decisions still.
I changed over the motor mounts to drive leadscrews instead of belt and pulley systems. This is the experimental portion of the project. Leadscrews are effective for CNC machining, but offer slower movement than the belt and pulley systems do. In theory they should also offer greater accuracy. My hope is that I can get similar speeds to what I could get an acceptable print from before, but with greater precision and dimensional accuracy. We'll see.