Experiments with Bearings
A variety of bearing ideas were tested. These are some of the attempts.  The unit on the right was designed to tension the rails and to allow space underneath for a rail support. The system uses a cam on one bearing that is twisted to tighten the bearing against the rail. It is made by tapping a threaded rod off-center into the head of another bolt. It is a simple idea and it works well, but it takes a while to build and requires welding. |
 A routed 2 x 4 served as a rail support as shown on the right. The 2 x 4 is routed to give clearance around the bottom of the bearings. ........................
Another method that worked fairly well was a pipe that supported channels, which held the bearings. Image below right. The slop can be removed with the mounting studs. The studs were soldered into the copper pipe. This was made from shop jetsam, so it is not pretty, but it is tight and works fairly well. |
 The bearings are placed 120° apart around the pipe; in this image only two bearing channels are visible. The problem with this system is the gantry has to be disassembled to slide the bearings over the rails. This is a nuisance when work is done on the Z carriage. The system also does not permit supported rails, which is a deal breaker for aggressive cutting. |
| ............................... |
 There was an attempt to keep things as simple as possible with the version shown on the right. It gains its tension by flexing the aluminum supports. It works fair but the metal is a bit too soft to hold tension well. Were one leg more solid with the other left as 1/8 inch aluminum, the system would hold tension. The flexible leg would always keep the bearings tensioned without the risk of overloading. ................................
|
 These next units were made of PVC male adapters and conduit. The threaded section of the PVC was cut so the caps, which were cut down to rings, could tighten the male adapters onto the conduit. The adapters were glued to PVC pipe and held into place with conduit straps. The 3/4 and 1/2 inch EMT conduits just happened to fit inside the plumbing fittings. The plastic has to be plumbing fittings rather than electrical fittings because plastic electrical fittings do not have tapered threads. The threaded tapers allow the cap-rings to close the fittings around the conduit. |
 The tension and
play is adjusted by how far the ring is tightened onto the male adapter.This system worked remarkably well, but the parts' price added up rather quickly. Surprisingly, skate bearings are cheaper. Also the unit would freeze when left dormant, and the tension of the PVC threaded rings would have to be re-adjusted. Again it was quite solid and gave a satisfactory quality of work after adjustments were set. However, the rails could not be supported, so the table had to be fairly small. ............................
|
 So it was back to bearings; the
next option was the unit
shown below made from a
hunk of 2 x 4.Rip the 2 x 4 at 30° (it is 60° from the horizontal) as shown in the small picture below. Flip one piece end to end so it mirrors the other. Right images.  |
 Drill holes A, B and C. A is for a through-bolt or threaded rod which holds the top bearing, and ties the blocks together. Hole B is drilled small enough for the bearing's bolt to be tightly screwed into the wood; the wood acts as a giant nut. Hole C is drilled oversize so that the bolt can wobble in it thus allowing for tensioning adjustment against the pipe-track. The head is countersunk to clear the top of the wood block so the block can be attached to its axis. |
 All bearings are held fast by nuts tightened firmly on each side. In the top image a small bolt head sticks out of the side of the block. This bolt is threaded into the wood and acts as a tensioner which pushes against the bearing's axle (bolt C) so the bearing will ride tightly against its rail. This bolt makes it much easier to adjust the bearing before tightening bolt C. |
 Update:
Here is a set made of 3/4 inch Corian, though they have been manually
run on a
pipe they have not been tested on a machine and I suspect they would
easily fracture under normal use. The purpose of this exercise was to see what design modifications were required when a thinner material was used. The only difference is threading the right block for the A axle since there is no room for the axle bolts to cross. This idea has potential. |
 Finally, a picture of the aluminum version of the trucks; they are easy to adjust and work well. They have been installed on a 4 x 8 ft. servo drive CNC machine that has become the work horse of this shop. |
| |