I wanted to build a test rig for my cloverleaf cable drive. A simple weightlifting platform.
I only had nine 2020 extrusion, 300mm long to play with. Not enough to build a cube and a bed frame. So I started to minimize extrusion count and came up with this:


The single linear rail will lift the 'U'-shaped bed frame. The bed frame will be stabilized by 'parallel bar' cables. I believe, the extra friction will keep the bed from dropping or at least will slow it down.


I have to print some brackets before I can test it. But if it's halfway successive, I want to build a CoreXY printer with two more cloverleaf drives.

Well this is very interesting! Can't wait to follow your progress.

Interesting indeed. What material is on the shaft for friction grip? It looks like the pulleys are different heights, any issues with the cable touching and rotating? And what cable are you using? Sorry, a unique design spawns many questions!

The material in the pic is silicone tube, but it's too soft. I try rubber tube now, but need a long run to confirm. A sandblasted aluminum idler pulley would be next on the list.
The pulleys are offset by M3 washers. The cable hasn't given me any headache yet, it's a 0.5mm Aramid fishing line.

I have a mocked up bed frame running with 1.3mm PP-cable for the parallel bars. That cable it too stretchy and I'll replace it with aramid.
But the parallel bar is a full success already. There's a significant difference with cables attached or just lifted by the linear rail.
Maybe I have to add diagonal struts on top of the 'T-bar'. It doesn't deflect now, but the PP cable isn't very tight.

Just made a video of the Cloverleaf drive.
It's running a long test now to test for slipping or wear of the rubber tube.

Here's a closeup of the stepper.

Edited 1 time(s). Last edit at 09/18/2018 11:52AM by o_lampe.

It's either the Ramps_test firmware or slipping, but the bed frame looses position constanly.
I thought of wrapping the cable twice around the stepper shaft before wrapping around the next idler pulley. Have to add washers to increase offset.
Also have to write a code, that moves xx# steps up and down instead of the timed approach in Ramps_test.

If the cable is slipping on the rubber shaft, you could increase the rubber diameter for more cable wrap length (2x?). The shaft as shown is pretty small in diameter, smaller than a 12T GT2 pulley it looks like.

At the risk of making it over complicated, you could add a second set of idlers at the corners and replicate the setup for double the cable wrap. That could possibly be less conducive to cable binding than wrapping the cable twice?

Making the center pulley bigger is probably the answer, since double-wrapping didn't work. Adding a second set of idlers would make the whole cable much longer, which counters the lower stretch ratio vs. belt.
I also need a better bottom idler, it limits the max. tension now.

What, if I replace one of the parallel bar pulleys with an encoder wheel?

Slip wouldn't be a problem with a closed loop stepper driver. ( here we go again )
Something like the mechaduinos?
Surely the resolution wouldn't be good enough. Maybe, if I fit it on the idler pulley? That would also work for the XY-gantry.