Retrofitting the contraption with ACME screw and Delrin nut

Online research I've done shows that all-thread rod is not the best candidate for a lead screw. Triangular thread is not designed for power transmission or positioning, so the accuracy of the thread is not important, and all that could explain the irregular intervals from my accuracy tests. The trapezoidal ACME thread, on the other hand, can be and is used for these purposes. One popular solution used in DIY CNC is an ACME screw with a lead nut made of Delrin. This arrangement apparently has little friction and little backlash. This is what I decided to try next.

Delrin nut

Since I could not find an ACME screw of 3/16" diameter, I had to figure out a way to fit 1/4" (smallest ACME screw available) into existing sliding element and 3/16" ID bearings, without subjecting the concept to a major overhaul. I figured that 1/4" 16TPI ACME rod coupled to 3/16" ends should work - I just needed to find adequate couplings. Another thing I could not find was 1/4" ACME tap. Some googling yielded homemade ACME taps, so I decided to make one too, albeit with much simpler tools:

With a power drill and a file I tapered a 6" piece of the rod, then fixed it between the wood planks in a vise and made two flutes (sort of) with a hacksaw. I think Dremel with a cutter attachment should be the right tool here, but the thing was good enough to tap a 1" long piece of 1/2" Delrin rod, though not without extensive physical effort - repeatedly driving it in and out of Delrin, moving forward little by little. It took some 30 minutes until it was tapped.

To be able to use the belt drives in future contraptions, the threaded rod axis should lay on the imaginary 4th edge of the slide angle (as if it was 1" square tube). This imaginary edge (and the rod axis) is 1/8" away from either wall of the sliding element tube, which is why 2 sides of the nut are filed down almost to the screw thread. It's important that these sides are parallel to the rod. The other sides are also filed so that the nut takes more or less square shape in the cross section. Using round needle file, two pairs of grooves (perpendicular to the rod axis) are made on the thicker sides of the nut, to accommodate four 6-32 mounting screws. The grooves are slightly slanted so that the nut is compressed more and more as the screws go deeper. In theory this should allow to regulate the friction and backlash in the nut.

Shaft couplings

Next, I needed to make shaft couplings for 1/4" diameter to 3/16" diameter. This turned out to be harder than I thought. I quickly discovered that in a shaft coupling, it is very important that both shafts are coaxial, otherwise the whole thing wobbles and whips. Initially I planned to drill a 3/16" hole in a 1" length of round Delrin stock, then drill 1/2" depth with a 1/4" bit, using first hole as a guide. I tried that several times (with a power drill) and in every case the holes were not parallel. I thought that because Delrin seems to grab the drill bit, the second hole cannot be drilled perfectly parallel to the first hole (with a power drill, anyway). The next idea was to press the 3/16" coupling nut halfway into 1/4" hole in Delrin. That didn't work either. Then I tried the first approach (3/16", then 1/4" halfway) with aluminium rod - same result, no alignment. Eventually I gave up on trying to get coaxial holes with a power drill.

The backup plan was nested nylon spacers - actually it was one of the original ideas for couplings, but because the hardware store didn't have the right sizes, I decided I'd first try making couplings from Delrin. I ordered the spacers from mcmaster.com for a very reasonable price of about $10 for a 100 pack each, and they worked pretty well. The small spacer is threaded onto the 10-24 screw or piece of rod, then inserted into the large spacer. Two 1/8" radial holes for 6-32 set screws are then drilled in the large spacer. Both rods need to have a slot for the set screw, which is easily made with a file. The wobble seems to be minimal with this arrangement - if anything, probably due to the rod not being perfectly straight. I hope that whatever remaining wobble won't wear down Delrin too quickly.

Part		Dimensions		Mcmaster.com
Nylon spacer	1/2"OD 1/4"ID 1"L  	#94639A506
Nylon spacer	1/4"OD #8ID 1/2"L 	#94639A295 

To be continued...