The current T5 timing belt thats used on the Mendel has a few shortcomings. Its expensive, can’t be printed, needs to be split with a dangerous belt spiting jig and its hard to obtained locally.
So what are the alternatives? Here is a quick run down of what’s currently being explored.
The idea is simple enough, print a ball chain pulley (thingiverse link) and use the same type of ball chain you find on most blinds and shutters. Its advantages are that its dirt cheap and available just about everywhere. Its downsides are mainly over its reliability as its more likely to stretch, break or slip than the normal belt used.
Using 35mm cinema/photography film is a fantastic alternative to the standard belt and is current being explored by BalanceSeeker. You could possibly obtain it from your local cinema for free and a single role would be enough for 10’s if not hundreds of Mendel’s. Even off cuts may be long enough. It can also be joined with simple tape and should not stretch during use. If that wasn’t enough, you could also add barcode like stripes to the film which would allow the belts home position to be known at all time when read with a rotary encoder type setup.
Rack and Pinion
The herringbone rack and pinion option being explored by Forrest Higgs its a fantastic option for the mendel in that its 100% printable. Its only shortcoming as far as I can see is that it would require the x-axis stepper motor on the Mendel to be attached to the extruder carriage, adding a lot of weight. Alternatively, the stepper could be stationary with the rack attached to the carriage. However this would increase the foot print of the Mendel in the same way that a typewriter extends outside its foot print as you type.
This is a bit of an odd idea I had this afternoon. Why use belts at all when you can use magnets and frick’n lasers?
The general idea is this, you have a perforated surface below the print bed that acts like a air hockey table to produce a cushion of air beneath the print bed. This allows it to move in the x and y direction with very little resistance.
The bed is then moved by4 electromagnets placed, two for each axis. By supplying current in different directions the electromagnets may either attract or respell solid state magnets placed on the bed. The bed’s position is found from two laser range finders, with one for each axis. A representation of what I’m describing is below.
In the image the bottom section is the air table, the arrows represent the directions that each electromagnet (the spirals) can move the bed and the round hockey puck shaped objects are the solid state magnets.
For obvious reasons this would be a nightmare to control and most likely would suffer from very poor movement resolution and vibration. However, if the electromagnets in my CRT monitor can be controlled precisely enough to guide an electron travelling at close to the speed of light to a point less than a milimeter in size on my screen, then why not a print bed? Dozens of orders of magnitude more mass is an obvious contender, but not something that couldn’t be overcome by someone with enough technical skill.
Its simplicity of only one moving part is very attractive but this is something I wont be attempting any time soon that’s for sure.