Friday, January 06, 2006
Pulley accuracy test rig
Results on the nylon cord will follow next week.
It looks like this...
I found this in an article about building a
string tensioned telescope:
I looked and looked for Kevlar string, because of it's low stretch characteristics. All of the Kevlar string I found was braided. Even though the Kevlar strands don't stretch, the braid tightens up, and the overall string does stretch. Someone on the ATM list suggested that I try an archery shop. I found a string with the brand name of Spectra. It has separate strands that are not braided, with a cloth outer jacket, and it has very little stretch. This string works well as is, but I was able to take the core (the fiberglass) out of the jacket, double it, and put it back in the jacket, making it have half as much stretch.
Ed, I have a very old pen plotter with a design which reminds of your rig. It uses thin metal wires to do the job. The wires looks like bicycle-brake wires, but thinner. I think that ordinary bicycle wires could be used with good result if the radius on the 'wheels' where made big enough to accommodate for the bending radius of the wire.
Ed, you might want to take a look inside a flat-bed scanner and see how they arrange their cables to make the bar light and scanner head go back and forth.
This problem of stopping cables from stretching is the chief reason I try to avoid such mechanisms as much as possible - when I do use them, I personally prefer steel cable.
Another, more long-term problem is that relying on specialist cables, such as high-end bowstrings, decreases the ease with which the machine can be constructed and distributed - I doubt there are many specialist archery shops in some third world countries where the reprap technology might eventually spread. Linkages like the peaucellier or sarrus have the advantage that they could be entirely made by another fabricator using base materials, though I suppose one could use a fabricator to manufacture a specialist string-making machine and then manufacture the bowstring using that....
Oh, and keep looking at the Stewart platform, and also "Crank" variants which require fewer multi-directional joints. I reckoned they were vey nice but waaay to complicated for the initial attempts. Also quite expensive - 6 stepper motors...