Beaded belts

There is a long thread in the blog about beaded belts as a replacement to the belts we presently use in reprap. Thought it deserved to be mentioned here too.

I started looking around for everyday easily-available items using a string or wire and small balls spaced out evenly along side.

Two things came to mind:

1) The kind of chain found hooked to the plug in a houshold bathtub.

2) Cheap plastic "jewelery" popular among young girls.

They both tend to be durable but dunno about the long-term tension, nor how to splice them together.

... the beaded strings of jalousies would be fine ...

For selfmade-types i think on two halfs of aluminium with 5 or more milled halfspheres (one half with holes for injection-molding) and a v-grove through - then inlay the string, close the halfs and inject hot plastic in the drilled holes, until the spheres are filled.

Let it cool down, open the halfs and shift the string, so that one bead is in a outer hole ...

This procedure can be reproduced, until you have enough beads.

Then you close the loop, lay both ends with the end-beads in the fittings and mold with the last sphere the both ends of the string to a closed loop ...

Ciao, Viktor

Edited 1 time(s). Last edit at 09/24/2007 09:13AM by Viktor Dirks.

Sounds cool!

Wonder if one could inject the plastic with an ordinary glue-gun, and if that kind of plastic would be useful in a beaded belt?

It would probably make sense to make a long aluminum form since one would otherwise have to wait a long time for the balls to cool!

One could drill conical inlets to each ball to make a descent fit from the glueguns nozzle, also a small scratch (hacksaw) to fit the wire in would be nice.

Then one would have to trim the excess off.

I wonder how the precision of beaded belts compare to a gear system.

The large gain, though, would maybe be that the drive wheels themselves could easily be made with rapid prototyping, since there is not the same need for precision as with conventional toothed gear.

Hi mimarob,

... you can make a slip-free system without gears or beads.

I build my first penplotters with 0,5 mm thick steel-strings over rolls, i made from plastic-coated ball-bearings, in which i lathed a v-grove.

Two steppers drived bigger rolls, over which i had 2 to 4 rounds of the string and at one end i had screw-spanners (?), which apply some amount of stiffnes in the loop ...

There was no slip and the penplotter was very precise.

Some high-speed pick'n-place-SMD-robots and SMD-paste-dispensers use this kind of driving too, so it seems to be not so bad

Viktor

hmm yes but this does not make an infinite loop or?

In the reprap design there is a need to run several revolvations for the Z-bed.

... yes, it's limited of free length, but why not fixing the z-bed directly to 4 vertical steel-strings, which comes from two closed loops, that runs over a single motorized roll, instead of 4 screws which are synchronized with a (bit more elastic) belt?

In the plotter i drived with 2 stepper-motors (which sat on the same plate) two wire-loops, which moved the rail (in X) and the toolhead (in Y) over an Area of 60 x 40 centimetres.

The 200-step-motors were driven in halfstep-mode, and the plotter had an overall resolution of 0,1 mm.

The same design i used at developing a cutting Laser-plotter, but then with a thicker and stiffer steel-string, better mechanical parts and finer motordrivers (IMT901 with 1/8-step).

Viktor

Because the Z-axis use thread drive and only needs to move slowly you could replace the expensive 400 stop motor with four cheap 48 step "tin can" motors and wire them in parallel dispensing with the belt, pulleys, tensioners, etc.

Edited 1 time(s). Last edit at 09/26/2007 08:13AM by nophead.

---

[www.hydraraptor.blogspot.com]

Yep, already thought about that. The problem is: what happens if you miss a step? Then your platform isn't level anymore, and because you're driving them in parallel, you can't resynch. You would have to have hard stops at the bottom of the travel, and then just bang the Z stage back into synch. The hard stop would have to be something that stops exactly at a certain place in the rotation. Maybe a set of four nuts, two of which are used as locknuts for the other two, which have a step ground into them which is less than the travel produced by a 360 degree rotation (aka the pitch of the threads on the rod).

There is no reason to miss a step with a properly designed stepper drive after initialisation. You would only get a problem if you powered down on the phase which is opposite to the phase you power up on. In that case it would be 50/50 which way each motor went. You could either always power off on a fixed phase or perhaps have a very crude shaft encoder to tell you which phase to start the motor. A disk with 12 holes and an opto would do it I think.

---

[www.hydraraptor.blogspot.com]

I recall about a year ago that they were having a little bit of step skipping going on in Reprap. Can't remember the details, though. I'd suspect that Simon and/or Adrian could fill you in. I haven't heard about it happening in a LONG time, however.

Just thought of an easier way to solve the startup sync problem. If the motors are always stepped with one coil on in normal use then power up with two on or vice versa. That way it will always be deterministic which way the motors move to "lock in" phase.

Edited 1 time(s). Last edit at 09/26/2007 02:28PM by nophead.

---

[www.hydraraptor.blogspot.com]