In that sense, acme rods are designed for linear movement. They're threads are much more vertical.

However, when i say they are bad because they are high friction, I think I poorly spoke. Because they are better for use in a CNC machine (3d printers are technically cnc machines, I think.) However, all of the rod based systems are "high in friction" because their basic principle is that of a worm-gear.   And especially where the X & Y are moving quickly, the static friction and intertia of the rod are very high.  If you want to make a bigger solidoodle the problems multiply, because thin rods bend easy and the wobble matters more as the length increases. So not only does the inertia increase because your rod is longer, but you start to need to consider getting a bigger rod.

They are designed to keep a nut in place or to seal. They also have much lower tolerances usually, and as such you can have sticking points. Add on to that the finer pitches, you have very, very poor linear motion components, especially when we want 300 mm/s travel speeds.

Overall though, screws aren't what we are looking for. They are good for moving larger loads than what we deal with. Belts are king for a reason.

I see inertia of the rod, and wobble, to be the biggest problems. Rods are more suitable to a big CNC machine, where the forces are high anyway because of the fact you're cutting into a metal stock.

If anything, with 3D printers, we need to reduce the inertia of the system with a bowdon tube (some 3D printers use a tube that goes to the nozzle, and feeds the filament from farther away. That way the extruder motor isn't on the nozzle, and the whole nozzle assembly is lighter. This means you can move it faster.)

solidoodlesupport wrote:

Some have mentioned trying threaded rods for all axis controls, just like the Z.

I would have thought that this would slow down movements fairly significantly without a significant gearing change? I could be really wrong though.

I'm not sure myself. It could be something to look in to.

Some have mentioned trying threaded rods for all axis controls, just like the Z.

I would have thought that this would slow down movements fairly significantly without a significant gearing change? I could be really wrong though.

Wire rope is great stuff.  strong as hell, flexible as can be, and it barely stretches under minimal workloads.

Roll-on/roll-off is a great alternative to belts, screws, and chain.

I have used wire rope for a decade, and it is one of my favorite rigging materials.

Now to experiment with some smaller stuff...

Some have mentioned trying threaded rods for all axis controls, just like the Z.

Roll-on/roll-off is a great alternative to belts, screws, and chain.

I have used wire rope for a decade, and it is one of my favorite rigging materials.

Now to experiment with some smaller stuff...

It uses wire instead of belts, chains or screws.

Screw system could actually be worse. They have more inertia for each movement. So they're likely to have a lower top speed. You might be better off weeding out elements of backlash in the belt system. Then again, there might be a reason why reprap seems to be mostly focused on screws.

In my experience, chains actually can have more stretch than belts. While they do not need nearly as much tension, they do have play in every link.

My experience is with bicycles, but those are if anything higher quality chains than I would expect of plastic chains.

We would rather move to a screw axis system than chains, but belts are here to stay.

My experience is with bicycles, but those are if anything higher quality chains than I would expect of plastic chains.

http://www.servocity.com/html/plastic_chain.html

Would a chain like this be an improvement over belts?  It probably wouldn't work in the current SD2 but if the case were a bit bigger it might.