Blog for the RepRap project at www.reprap.org - a project to create an open-source self-copying 3D printer. To get all the early posts on this blog with all the images as a single PDF visit this page.
Pages
▼
Friday, January 15, 2010
A printable, high speed alternative to belts?
Herringbone racks and pinions might well be a practical, printable alternative to belts for high speed, NEMA stepper driven Reprap machines.
i saw those advertised in store once. from what i remember some of the things they said about them, are that they make less noise, they do not slide off the pulley, and are supposed to have a longer life. i just don't see them being easier to make the belts, and i have not seen a reprap make much for a belt yet, though i am keeping my hopes up for that.
Looks great, you'll run into the whole chicken and egg problem though of not being able to print something bigger than an axis travel unless you print the rack on an angle.
The biggest problem with using a rack and pinion for precision movement will be backlash. Mendel gets away with screws for the Z-axis by using the weight of the printhead. I don't know if backlash can be managed through software for FDM like it can for subtractive machining.
I don't see the use of a rack and pinion being any more susceptible to backlash than the belt and pulley would have. Hwin and THK sell very expesive Rack and pinon drives for CNC machines so it must be a sound solution.
I also don't see a problem with the rack length as you would design it to have a joins so it clips or slides together. Particularly its an ideal shape to make something like a dove tail join to make longer lengths of Rack.
That is why I like the design it looks like an ideal solution to remove the need for belt drives.
Having worked with the Rapman for a few months now, I can assure you that the belts it uses do stretch and the speeds at which it operates most certainly causes an elastic backlash, small but real.
Could you get around the problem of the maximum fab-able length of rack by using two racks placed side-by-side - but staggered:
------------ ------------
You'd need the pinion gear to be twice as wide so that it can grip either or both racks and you might need to make some kind of registration mechanism on the sides of the racks to ensure that they line up well - but that's certainly do-able. Then you could make racks of arbitary length from shorter pieces:
The rack, as pictured, doesn't need any modifications to be able to be run to any length needed. The "break" nests right into itself, so you can run as many lengths as needed end-to-end.
If backlash is a problem, that could be managed with a second co-driven gear, offset the distance of the backlash from the drive gear. But with the herringbone shape, the backlash distances are going to be quite minimal.
This comment has been removed by the author.
ReplyDeletei saw those advertised in store once. from what i remember some of the things they said about them, are that they make less noise, they do not slide off the pulley, and are supposed to have a longer life. i just don't see them being easier to make the belts, and i have not seen a reprap make much for a belt yet, though i am keeping my hopes up for that.
ReplyDeleteLooks great, you'll run into the whole chicken and egg problem though of not being able to print something bigger than an axis travel unless you print the rack on an angle.
ReplyDeleteThe biggest problem with using a rack and pinion for precision movement will be backlash. Mendel gets away with screws for the Z-axis by using the weight of the printhead. I don't know if backlash can be managed through software for FDM like it can for subtractive machining.
What a lovely design Forest I like it ;-).
ReplyDeleteI don't see the use of a rack and pinion being any more susceptible to backlash than the belt and pulley would have. Hwin and THK sell very expesive Rack and pinon drives for CNC machines so it must be a sound solution.
I also don't see a problem with the rack length as you would design it to have a joins so it clips or slides together. Particularly its an ideal shape to make something like a dove tail join to make longer lengths of Rack.
That is why I like the design it looks like an ideal solution to remove the need for belt drives.
Having worked with the Rapman for a few months now, I can assure you that the belts it uses do stretch and the speeds at which it operates most certainly causes an elastic backlash, small but real.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteNice bit of work Forrest!
ReplyDelete--------------------------------
It may be good to have this in our documentation.
http://dev.forums.reprap.org/read.php?58,32811,32866#msg-32866
Documentation task #0012 - check Forrest Higg's work into RepRap.org docu-wiki under RBS/gears/herringbone.
http://objects.reprap.org/wiki/RBS/Gears/Herringbone
Ticket opened: Fri Jan 15 22:39:10 EST 2010
http://blog.reprap.org/2010/01/high-speed-alternative-to-belts.html
--------------------------------
Remember guys, teaser links can go the wiki along with to your personal blogs. :)
Could you get around the problem of the maximum fab-able length of rack by using two racks placed side-by-side - but staggered:
ReplyDelete------------
------------
You'd need the pinion gear to be twice as wide so that it can grip either or both racks and you might need to make some kind of registration mechanism on the sides of the racks to ensure that they line up well - but that's certainly do-able. Then you could make racks of arbitary length from shorter pieces:
---------- ---------- ---------
---- ---------- ---------- ---
It is much simpler than that. You can just print the rack diagonally on your table so it can be longer than the individual axis travel.
ReplyDeletei was just thinking that if you made them interlocking some how that it would not matter how long you needed to make it.
ReplyDeletei would love to use a system like that on my rep rap but i am not that far just yet.
The rack, as pictured, doesn't need any modifications to be able to be run to any length needed. The "break" nests right into itself, so you can run as many lengths as needed end-to-end.
ReplyDeleteIf backlash is a problem, that could be managed with a second co-driven gear, offset the distance of the backlash from the drive gear. But with the herringbone shape, the backlash distances are going to be quite minimal.