Why Repstrap?

Why are people trying to repstrap? Seems to me that once there's a design for a reprap, one person makes two, two make four, four make eight, etc.

So it seems to me that work towards the goal of the reprap should go into solving the design problems -- and they require the use of something very close to Darwin.

Not meaning to rain on anybody's parade.

The Darwin basic mechanical design, while not 100% ideal, is very sound. Once running, the Darwin will proliferate quite quickly. Still, I can think of two good reasons and one other reason to be doing RepStraps:

1) Seed sites. It's much easier to get a RepRap passed around within the borders of your local principality rather than international. If RepRap part kits are made in several sites, this helps to limit the disuse decay that will eventually happen (where one does not beget two). Its like crystal growing. One seed crystal can grow to become a large crystal, but several will grow much faster and have a better chance of producing the structure you're looking for.

2) Development testing. By having several repstraps worldwide working on the final bugs and testing, especially of the extruder heads, issues and variations can be caught that would not show up if the only tests were being done in Bath. Sitting here talking shop is all well and good, but someone has to test the ideas that arise.

3) You want to. Some people are impatient. Some are looking farther ahead (as I am, and trying to lower my costs at the same time!) to more advanced technology. Some are looking at a different variation that may not self replicate but be cheaper (I'm looking at you, Forrest!).

1) Seed sites .... well, Adrian is going to have to pick seeds regardless, by virtue of (very likely) having the first working Darwin. I'm sure that he will pick sites that he believes will make ten, not two, child repraps.

2) Unless repstraps are actually Darwins, I don't see how the lessons from one will transfer over to the other.

3) I'm impatient, too!! But I want to build something that can reproduce itself, so ... if I can, then it means that I could have gotten parts from somebody else.

That's why I suggest that enthusiast's time is best spent proving Darwin, rather than going off and building their own design. Once Darwin is reprap'able, the design problems will be found and fixed quickly, and people can use their Darwin to make Spencer, and Lemuel, and all the subsequent versions. Even Lysenko, if they want. :-)

Forrest Higgs is making a variant design that he dubs Tommelise. He's using more traditional motors, and a feedback system, to handle position. This is a concept that is being considered for Mendel, so any lessons he learns with his system can be applied to a future RepRap. Think of it as a development fork that continues to add to the body of knowledge of RepRaps in general.

A lot of the design work is in tuning software and the electronics. As these are also the parts that are most readily available, putting them on Something allows the user to work with them, even though the "official" design is in short supply. Many of the bugs, at least those detected, are agnostic about the actual design of the cartesian coordinate robot, allowing them to be corrected long before the official box shape is available.

I don't like the Darwin design... so I've created my own... and the repstrap is out of legos! (all but the extruder... which I'm still working on)

Did you go for a moving head (like darwin) or moving base (like mcwire) design? I've tired both and they both have their own merits. I made an paste extruder completely lego and a syringe that works pretty well. Here are some pics of the old design, i'm completely re-building it right now using a moving base design as the previous would have run into problems when mounting a plastic extruder onto it.

[www.flickr.com]

Have you got any pics of your's? It would be interesting to see how you've gone about it.

No extruder yet... The design is a moving xyz platform, with a stationary head(s)... I've designed it to be 99% reprappable (electronics, steppers, piece of thread/wire, copper strips that will be reprappable soon!)

You can check the builders blog for my postings... they were back in January and Feburary... I've already been planning a reprap version of it (with appropriate overhangs, fasteners, replaceable rack gearing, and expandability for a reprap to increase its own work area to reproduce itself same/larger size...

I, for one, have found that my work on a RepStrap, built from marine plywood and basic big-box hardware parts allows me to get started on the project, test the boards and extruder, and actively participate right away. The cost for the whole endeavor is still under $300, and almost all of that will get resued on the RepRap once I build it. I am only out a half-sheet of plywood, a couple drawer slides and some random hardware in exchange for being able to start now. This is really important to me, since I don't have access to inexpensive RP parts - hence why I am building a RepRap...

I am determined that once I get it to the point that I can replicate parts, I will start building Darwin parts for myself and others - I just see it as a way I can get there faster, and hopefully contribute some ideas (or at least some testing) to the project faster.

Besides, it is really really fun!

I think that the last post hit it on the head. A reptrapper might be buildable to someone that doesn't have the money for or access to existing rapid prototyping.

And there's a satisfaction in having built the thing in the field with your own hands. Reptrapping comes from the word bootstrapping, which is a reference to the expression "pulling yourself up by your own bootstraps" (a physical impossibility). Building the machine that can build the machine is a pretty cool trick.

Designers of compilers of new programming languages like to write an early compiler in another language, then write a final version in the new langauge, so that the compiler can compile itself, which demonstrates the utility of the new language, and is a neat trick besides. Same kind of thing.

With a self-replicating machine, there is the chicken + egg problem associated with devising the first machine.
If there is a clear design path of evolution it should be possible to fabricate a machine which can generate ever more sophisticated versions of itself.
A simple and rudimentary (read inaccurate) machine might have the capacity to produce the components for a more precise version. This is how we have gotten from knapped flint to particle accelerators. If there was a documented plan to get from the stone age to the subatomic age, we would have arrived here somewhat faster.

On a more practical level, if the component design of the Darwin RepRap could accommodate varying methods of fabrication and be tolerant to imprecise fabrication (possibly through a greater degree of adjustability) then the machine could replace its own parts through an iterative approach to precision.

To this end I have designed and am in the process of building a version made from wooden lasercut pieces which fit together and retain the same configuration as the RP originals. As lasercutters are still not as widespread as they ought to be, this leads me to wonder how the parts could successfully start off as cut from sheet with a drill press and a fretsaw.

tobyborland Wrote:
-------------------------------------------------------
> With a self-replicating machine, there is the
> chicken + egg problem associated with devising the
> first machine.
> If there is a clear design path of evolution it
> should be possible to fabricate a machine which
> can generate ever more sophisticated versions of
> itself.
> A simple and rudimentary (read inaccurate) machine
> might have the capacity to produce the components
> for a more precise version. This is how we have
> gotten from knapped flint to particle
> accelerators. If there was a documented plan to
> get from the stone age to the subatomic age, we
> would have arrived here somewhat faster.
>
> On a more practical level, if the component design
> of the Darwin RepRap could accommodate varying
> methods of fabrication and be tolerant to
> imprecise fabrication (possibly through a greater
> degree of adjustability) then the machine could
> replace its own parts through an iterative
> approach to precision.
>
> To this end I have designed and am in the process
> of building a version made from wooden lasercut
> pieces which fit together and retain the same
> configuration as the RP originals. As lasercutters
> are still not as widespread as they ought to be,
> this leads me to wonder how the parts could
> successfully start off as cut from sheet with a
> drill press and a fretsaw.


Well I have been working on making my Repstrap from MDF, Oak and for the corner blocks using square PVC bar stock and so far I have been able to get everything very strong and solid. It has taken a lot of trial and error and making 2 to 4 parts of each before getting it right. When you look a Darwin there are not a lot of parts that are not multiple of. My goal has been to make the parts using only hand tools. even though I have found it took a drill press, a chop saw and a Scroll saw. I am willing to help anybody and I have most of the parts in drawings or can make them for you. the laser cut parts is an even better answer to this problem in my eyes the only thing would be the access to a laser..

Bruce Wattendorf