PDA

View Full Version : Timing belt instead of rack!



ptjw7uk
17-12-2009, 02:54 PM
Hi All.
Been thinking(bad for me) and I thought why not use a timing belt for the X-axis on a router system. Two lengths of belt could be stretched along each side of the X-axis and interconnected timing pulleys could supply the drive.
So is this feasable and what backlash would be in the system, the pulleys could also negate the need for a bearing as they would run along the belt.
Not sure about tensioning the belt as yet just in the thinking stage.

peter

BillTodd
17-12-2009, 03:09 PM
Have a look at Mike Everman's belt drive here:

Best belt drive ever! (If I do say so myself) - CNCzone.com-The Largest Machinist Community on the net!

ptjw7uk
17-12-2009, 04:40 PM
Similar to my idea but he's using 2 belts which will up the cost plus the extra bearings used.
Also makes mention of reducing backlash etc.
Just had a thought against my idea lack of actual drive as only one tooth would be in contact.
Back to drawing board!

peter

Ross77
17-12-2009, 06:29 PM
Similar to my idea but he's using 2 belts which will up the cost plus the extra bearings used.
Also makes mention of reducing backlash etc.
Just had a thought against my idea lack of actual drive as only one tooth would be in contact.
Back to drawing board!


You can use one belt but it still needs to wrapped around the pulley, this causes probems with backlash due to the belt stretching and sagging. (and between belt and pulley)

In the other design two belts (or 1 cut in half :smile: ) are required to solve this and increase the tooth count on the pulley. The bottom belt is fixed and acts as a rack. the top belt is used for driving and the bearings force the top belt back down to mesh with the lower one forming a solid. Belt stretch is greatly reduced as it is only the small portion over the pulley and the sagging is removed. one thing to watch is that the belt can take back bending.

As you pointed out this all added up and I have mainly seen it used on on large machines where ball screws would need to be massive. you could compromise and just provide support for the belt to remove the sagging.

BillTodd
17-12-2009, 08:18 PM
Have you considered a chain rack? Something like this:

Ross77
17-12-2009, 09:33 PM
Have you considered a chain rack? Something like this:

Now thats an idea :beer: is that a normal bike chain and turned down spocket, or a special?

Have you used it? and do you Know accurate is it?

BillTodd
17-12-2009, 09:42 PM
Have you used it? and do you Know accurate is it?
I made it in pixelanium and, as designed, has negative backlash thus improving any machine that it is fitted to by 1000% :joker:

The chain is normal, but the sprocket isn't a normal profile because the chain doesn't wrap around it. The profile shown is only a rough approximation.

Bill

ptjw7uk
17-12-2009, 10:46 PM
Nice idea on the chain but my original thought was to kill 2 birds with one chain so to speak in that the pulley running on the belt would effectively act as a bearing. You would still neeed another lower down to give some rigidity but the 2 belts per side increases the complexity and I was after simplicity mind you I realise that simplicity and accuracy dont go hand in hand.
Mind you ball screws are just not cheap enough.
Peter

BillTodd
18-12-2009, 11:49 AM
I realise that simplicity and accuracy dont go hand in hand.
Mind you ball screws are just not cheap enough.

Have a look at roller nuts:

http://www.cnczone.com/forums/showthread.php?t=13593&page=2

This is a very long thread, but an interesting read with some good ideas for low cost alternative to ball-screws.


my original thought was to kill 2 birds with one chain so to speak in that the pulley running on the belt
If you could rig up some kind of positional feed-back (i.e. from a linear scale) then you could use a traction drive.