Side bottom view of motor mount plate, also showing 1/2" spacer below the lead screw. It’s also a better shot for seeing the extra belt, which is relatively snug around the trapezoidal pillow block mounting screws.
I did these parts in aluminum. The spacer could be plastic, but I would expect the motor mount eventually to creep under load if it were made from plastic.
missing/deleted image from Google+
I noticed when I was test-fitting that the X 2020 gantry (484 mm long) seems to just barely fit between the MGN12 trucks on the linear rail on top of the Y gantry 2020 top of the frame. But I haven’t mounted them, they were just sitting there loose, so I did the math…
Total X width is 530mm, so 510mm between centers of the Y gantry 2020. The MGN12 trucks are 27mm wide. There is half of each of the two trucks on either side of the centers, so 510mm between centers minus 27mm (two halves of the trucks) is 483 mm to fit a 484 mm extrusion.
Maybe I can mill the alignment brackets off by 0.5mm and make it fit? Alternatively, I can mill 0.5mm off the top 10mm of both top sides of the X gantry 2020 to make it just perfectly fit. I’ll have to think about that! 
I milled 0.5 mm off the top 10mm of both sides of the x gantry 2020. I milled alignment jigs for mounting the linear rail, and mounted it with all the M3x8 screws and t-nuts that didn’t interfere with existing screw heads in the 2020.
I mounted the X linear rail on the side of the x gantry extrusion. I haven’t yet decided what print head to hang off it first, but the nice thing is that I can do anything that matches the screw pattern of the MGN12 trucks, which should make it easy to change up.
I will have to mill spacers to add trapezoidal pillow blocks to the tops of the z screws because they wobble. Sadly 1/2" stock is about 1.5mm too thick, so I’ll have to face them.
Then I’ll have to mill T shapes to attach the Y trucks to the X gantry 2020. That should be easy enough. But if they aren’t perfectly square, my prints will be skewed.
I milled the spacers for the z lead screw tops, for mounting the pillow blocks. I made them just like the bottom blocks, then realized I couldn’t get the t-nuts into the slot in the 2020 without being able to drop them at least part way down over the lead screws. Rather than worry about how deep a pocket to mill, I just drilled all the way through the middle. A pocket would have made installation a tiny bit easier, maybe.
This is not a precision part, and it is not under significant stress. The hole and half holes are intentionally oversized for easy fit. If I hadn’t already been set up to mill these pieces, printing them would have been appropriate. The half holes could easily be normal through holes in a slightly longer part.
Due to the flexibility of the lead screw, these are not actually an over-constraint. They solve the problem of free end whipping when the bed is at the lower travel extent, and reduce vibration/chatter throughout travel range.
Because I removed 1.5mm from the face, M5x25 screws were about 1.5mm too long. So I put three washers under the screw heads. Good thing this isn’t a precision part!missing/deleted image from Google+
No pictures tonight, but… I milled some prototype X gantry holders. I used .25" aluminum, 26mm x 100mm, which was just enough to try out the movement and not enough to mount pulleys on for driving the movement, but it let me experiment with a design to keep everything square.
I milled a pocket in the bottom 1mm deep. It was 2mm from one long edge, and 20mm wide, leaving 4mm on the other edge. The pocket went all the way to the edge at one side and to 26mm from the other edge. The pocket fits the 2020 extrusion to hold in straight. In the last 27mm, I drilled 4 3mm holes, 20mm on centers, to fit the M3 screws needed for the MGN12H carriages, and I pocketed 1.25mm deep around each of those 4 mounting screw locations so that M3x8 screws would be the right depth to mount to the carriage trucks. I drilled 3mm holes centered 10mm in from each edge of the pocket and 8-10mm from the ends of the pockets. I used M3x8 screws and M3 t-nuts to mount to the X 2020. I made two of these, in mirror image, for left and right mounts. The 2mm edge of the pocket is toward the front, and is small to avoid interference with the front-mounted linear carriage on the X 2020. There is more room on the back, thus 4mm. The 2mm and 4mm edges hold the 2020 extrusion square.
Mistakes I made included using a chuck instead of a collet to hold the center drill for starting holes in the first of the two holders. This made the holes insufficiently precise and one of the four just doesn’t match up, even after drilling out with a 3.2mm oversize bit. Using a 3/16 collet and a #2 center drill to start holes for the second holder was more precise. (I know better than to have tried the chuck…)
It runs square. It’s a little bit stiff, and it’s stiffer toward the back, which indicates that the linear rail isn’t quite straight, even installed with the jigs I made. Or maybe that the frame isn’t quite as square as I think. That will take some investigation.
The real holders will have to be larger and have mounting built in for pulleys. I’ll probably use two pulley stacks with vertically-aligned belts aligned with the front of the extruder carriage truck. Alternatively, I could just use the existing holders and screw some square tube on top to hold the pulley stacks. I’ll see.
A whole album of pictures of the next (final if I’m lucky) revision of the gantry mounting plates, plus discussion.
I milled some 1.5" square tube (more scrapyard special!) with 1/8" walls, lining up 5mm holes for M5x45 bolts through into the M5-tapped holes in the gantry plates. I did two in mirror image, one for each side. They sit on top of the screw heads for the gantry plates, which isn’t obviously optimal; maybe I should instead use some spacers or bushings. But since they contact only three screw heads and three points determine a plane, maybe this is actually the best arrangement. Don’t fix what ain’t broke?
I noticed, finally, that I had lost track of where I should drill the tapped holes in the gantry plates, and ended up with the belts lined up with the bottom of the front-mounted linear rail carriage instead of with the ball bearing slot in the linear rail as I had originally intended to do. Doesn’t really matter, probably. It will just change my carriage design a little bit, and since I haven’t started that design it’s a distinction without a difference.
I haven’t yet added bushings/spacers to the M5 bolts to hold the idlers in place. The bottom right 20T idler in this picture is just there in place of a bushing I will add later. I just needed to test out the height for now.
The way I’ve stacked this up moves the belt path much higher than it was on the printer as shipped. I’m thinking of mounting the front corner idlers inside 2" square tube, which would require buying 55mm M5 bolts. An alternative would be just using 1/2" aluminum pieces which would not require support at the top, perhaps. I also need to think about machining new motor mounts.
To drill the bolt holes more precisely, while I did center-drill the holes with a collet-mounted center bit to start the holes, I also used a 5mm collet to hold the 5mm drill bit instead of a chuck. The holes lined up very nicely!
missing/deleted image from Google+
I occasionally make progress on this project. A month or more ago, I added bushings where they are missing in the picture above. I reused some of the bushings that shipped with the printer, then added a few more in custom lengths. A few weeks ago I milled out some thicker aluminum blocks as spacers for mounting the motors and some inch-thick blocks off the front corners for mounting corner pulleys. The thick blocks mean, I hope, that I don’t need to support the pulley bolts top and bottom; I’ll see how well that works. Because of the square tube holding the pulleys on the Y carriage, I had to put the front outside pulleys well outside the original printer envelope to keep the belt out of the way event when the carriage is at full back Y travel. The alternative would have been yet another idler.
Yesterday and today I milled a single-extruder bowden extruder carriage with belt clamps. I used a 45⁰ chamfer bit to score the belt-holding faces with .2mm lines 1mm apart all across the face. After loosely installing the belt, I tightened the screw part way, grabbed the belts with pliers, and pulled tight. They stayed tight while I screwed down the clamp. What I have now is still the stiff steel-core cable that shipped with the printer, but I plan to replace with glass-core because it’s pretty stiff going around all those pulleys… But both belts pluck about the same note so they are approximately equivalently tight, and the XCR3D XCR-BP6 mounts tight; it’s not going anywhere.
The belts are a bit too far apart in height because I used an M5 screw for the clamp. In retrospect, an M4 might have been a better idea, and I have more of those handy anyway. The larger screw will create a bit of angular deviation at the extremes, but probably not enough to measure.
I’m also planning to make a carriage for direct feed with the aluminum titan clone. There are three known problems with those: the hobbed gear is installed in the drive gear upside down so it doesn’t align with the filament correctly (fixable by buying the original), it doesn’t have a filament constraint (fixable by making and installing one) and it doesn’t use a bearing around the motor shaft used as a pivot for the arm (I hope fixable by milling it out and installing MR58 bearings).
Sorry for the ugly picture.
missing/deleted image from Google+
I think the 2mm pitch on this timing belt is a little off in some places… 
I bought fiberglass core belt to replace the steel core that shipped with the printer. I’ll see how that works.missing/deleted image from Google+