roller pinion seems very printable...

Not that I want to encourage more dependency on printed parts but this seems like something higher performing and more easily printable and easy to manage than the herringbone:

[pdf.directindustry.com]

Could easily be a printable replacement for standardized nylon rack and pinion (which would be the boot strap option).

Someone else (see Diary of a Technocratic Anarchist) is already working on this idea!

Actually what I was posting is in response to that, the "roller" and pinion, rather than the rack and pinion (as that blog shows), is a slightly different design which I think would be particularly well suited to printing. The roller could have the "teeth" elements as bolts with spacers screwed between 2 printed hubs or possibly 100% printed. The herringbone is fine but very complex to model and requires pretty high printing tolerances, the roller and pinion is much simpler and could probably get by being pretty messy.

Because the profiles are 2D for both the hubs and rack and the provided documentation lists dimensions it should be easy to build stl's for (rather than the herculean effort for the herringbone modeling)

Interesting idea.
I think that I would use OpenSCAD to create a parametric model for this then you just specify the appropriate parameters and magic you have the proper STL file.

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

Neat! It would be good to tack up some stuff here including the stl, .openscad file, etc.
[objects.reprap.org]

Now how do we make the rollers? It seems like it'd be pretty hard to print a bunch of low friction bearings for each roller.

how about 606 bearings? they would make the wheel turn out quite large but it will five you a lowfriction roller surface. ask around at NYC resistor, they have a bunch of them. or anyone else who has them could try. I think 608 bearings will simply turn out way too big.

my 2ct

Why do you NEED bearings for the roller? I think you could get by with static rod/spacers mounted in the hub. VXD generally has any size small-medium sized bearing 10-packs for $5-$15.

goinreverse, the bearings are necessary for the anti-backlash part of this design. You could do away with them, and get a significant increase in either friction and wear rate, or backlash. I agree that this would be a neat idea to experiment with printing, sans bearings so perhaps we could try various fits and find a compromise between friction and backlash?

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

The rack's tooth profile is cycloidal, so unlike an involute gear on a rack, the pins/bearings' contact point slides over the tooth form rather than rolls on it. Without bearings, the rack teeth will wear down to something like a bike cog shape, losing the design's biggest selling point, the opposing two-point-contact, introducing prominent lash and only getting contact when the pins are at the bottom of the valleys. It would end up being something like pin-and-lantern gearing that was used in mills before the industrial revolution.

Edited 1 time(s). Last edit at 02/12/2010 09:21PM by murd.

I think a lot of that is speculative, It could potentially taking a very long time to wear down if it where plastic on plastic. It wouldn't be very hard for someone to print and compare the bearing vs. non-bearing construction and the respective tradeoffs.

Edited 1 time(s). Last edit at 02/13/2010 01:26AM by goinreverse.

Since I'm speculating, and you're not, why don't you think a length of bicycle chain with a cog rolling along it is essentially the same device as you're proposing, non-speculatively?

yep chain and sprocket is an ideal non-roller example.

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

I meant the amount and time to wear down is speculative and material dependent rather than the concept, sorry I was not more specific. I don't know what the co-efficient of friction of printed abs (or pla) on printed abs (or pla) is or if it could be modified by some sort of lubricant to further reduce wear and maintain 2 point contact for a sufficiently useful duration. It does seem to me to offer several benefits over a possible herring bone printed gear, that is the only point I was trying to make. Even it it does require bearings to be implement I don't think that is a major impediment.

One issue I see is that while this may be easier to print than a herringbone rack, it would require more non-printed parts. Even if it doesn't need bearings (which would require quite a bit of testing to determine, since you're talking about seeing how it wears down under longterm use), it does absolutely rely on precise and very stiff rods - ie, metal ones. A herringbone pinion has no need for metal elements - with proper design, even the shaft could printed.

I would also disagree that this design requires less printing precision than a herringbone gear. In order to maintain the high stiffness, high speed, and low backlash that makes it attractive, this design needs three of the rods to be engaged at any given time. That means the rods need to be precise as do their positions, and the rack also needs to be built to high tolerances.

If it turns out it does need a bearing for each rod, then I can't see any justification for it's use - you're talking about doubling if not tripling the number of bearings the whole design needs - the example pinion on page one would need 20 bearings alone! Needing to add 40 bearings (60 if you wanted to do the Z axis this way as well) is a major impediment in my book. By your pricing range, that's anywhere from $30 to $90 worth of (nonprinted) bearings.

By the way, chain and sprocket setups do involve rolling, they just don't have bearings at every joint. Which is why #25 chain isn't advertised as 0 backlash or having industry leading service intervals. Each link in the chain can rotate relative to the next around their shared post, which allows the post to rotate around the sprocket.

By the way, if you take away the bearings what you have is a very old technology called a cage gear or lantern pinion: [www.thefreedictionary.com]

One last note - according to that pdf, this design is patented.

It's not particularly hard to print a herringbone rack and pinion, once you have the scripts to generate the shapes.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison