First RepRapped RepRap?

So far, I don't think there has been a full set of RepRap parts made by a RepRap machine. I've seen a few parts here and there but nothing that constitutes a full replication.

It occurred to me that this would be an important milestone in the project. Sure, there are going to be quite a few Darwins around pretty soon, mostly based on moulded parts, but these are all, technically, first-generation machines. There's something special about the first real second-generation RepRap.

By the looks of it, we should be seeing one soon. There ought to be a ceremony or something. :-)

I agree. Some sort of "birthing" ceremony or celebration. I like the idea. Maybe a cordinated toast around the world? Funny idea.

Demented

I think at minimum there should be a message thread with a compiled list of what remains to be done... and edited message marking the date of each accomplishment.

Then we will have a good idea of how close we are...

We should have a shower for the new born. Don't want to miss it because they grow up so fast and have kids of their own before you know it.

haha, birth control is in order now!

Demented

I would expect the first RepRapped RepRap to be made by either:
- Adrian's team using the original Darwin
- someone who builds a Darwin from moulded parts, then replaces them one by one

Then there's also the first RepStrapped RepRap. That's pretty much anybody's game at this point.

Hmm, does a Darwin made from moulded parts count as a RepRap or a RepStrap?

i would say that a moulded darwin is still a RepStrap... after all, RepRap stands for replicating, and a RepStrap is a machine designed to be built with non-printed parts to be used to print the parts to build a real machine. technically, i dont think there is a single *real* reprap out there (if you dont count stratasys prints)... basically the first person to print out their own parts and build a machine out of it that can print more parts will be a full, real, RepRap machine. the crucial test is that the machine made out of printed parts is able to print parts as well.

And then we raise the bar and demand that the first reprap print an improved version.

Sean,

On that issue. Why are we having people build a RepStrap to build a RepRap that cannot do anything unique? IE the RepRap can make all the same parts that a RepStrap can and nothing more. Shouldn't the second machine you make/print be more capable than the one you have? I'm not sure what the "official" stance on this is but I've been mulling it over for some time now.

Demented

Hey Demented,

I'm not entirely sure what you're grappling with, but the main reason for the RepStrap design is to lower the barrier to entry. Until Adrian and Ed prove me wrong, Darwin is an experimental design that has yet to produce its own parts. Until it accomplishes that, it will simply be a very expensive testbed that may or may not work. Even with the molded parts now available, it is a rather expensive design. Darwin only truly makes sense once we have cheap, reliable 3D printing resources on hand.

Now, thats where the RepStrap design comes in. Its cheap, easy to build, and I'm very confident it will be a reliable platform for printing. Tom McWire proved that it can be used for light milling, so it should be adaptable for printing as well. In the next couple of weeks, I hope to have some solid proof to back that up as well.

The unique thing about Darwin and its future variants is that they can reproduce. The raw cost of the plastic for a darwin is literally $10. When (not if) we achieve this then it will drastically change the nature of the game and make a RepStrap type machine obsolete. The problem is that we dont have access to that stuff yet, so the only viable route for people like me is to create a RepStrap machine that i could use to print a darwin or its variants.

Once this technology becomes widespread, we'll see an explosion in the growth of the technology. People will print out experimental versions, new toolheads, and all sorts of cool stuff. But we need to get a critical mass of machines out in the wild to get to that point. I personally dont think molded parts are going to cut it, but its a great step forward and I applaud Ian for his efforts.

Until then, I think that the goal of a solid RepStrap kit is the way to go. Otherwise we simply have a community where people like Adrian do all the work and the rest of us watch and wait in anticipation. Wouldn't you like to be able to help and contribute too?

Here here I am trying to contribute also and Demented it has been growing if anything I can print shot glasses...

Bruce

Hmmm. Yes, I see what you mean but my question still stands.

Essentially once you build a RepStrap you can do anything that a RepRap can. So, why build a RepStrap and then build a RepRap. You already have the capability on hand at that point. I'm not arguing with the need for quick, cheap, reliable 3d fabbing. I'm arguing with the idea that all of us who build a RepStrap must then build ourselves RepRap v1.0 Darwins.

Bruce, I agree things are moving rappidly so my question is mostly theoretical in nature.

Demented

Except give a practically free repsrap parts kit to your friend.

Once YOU make a repstrap, there's no reason for you to use it to make a reprap, for yourself, until the next design iteration. But, you can make a reprap to give to your friend, or you can decide the repstap is too slow, you want to make that art-project lightswitch cover AND the doorstop you promised your mother-in-law in less time... So you make a second machine to double capacity.

One advantage of a reprap over a repstrap would be its ability to print its own replacement parts. If you made your own repstrap, this point is, admittedly, weak, as you obviously have the skills and tools to repair it.

However, as you'd have all the expensive parts once you built the repstrap, (motors, controlling electronics, etc,) going ahead and printing out the reprap, and swapping all the parts over, becomes easy enough, and cheap enough, some of us would probably do it just to say we'd done so. Would it be an improvement? Debatable. That square is, as I understand it, not nearly as structurally solid as the McWire, but I suspect it has a larger working volume.

I see your point. It is good to have a replacement on hand. As to the printing area, that depends on how big you make your McWire. I've got about a foot and a half volume in my RepStrap though it isn't strictly a McWire design. Took some parts from some other sources.

For me, the next iteration is where it's at: Mendel. I want that advanced capability.

Demented

... i have first to bring my repstrap to full functionality - it's a big CNC-system designed for professional milling (costs new something about 4 - 6000Euros or so and was used as glue-dispenser), so i'll order a better mill-head first.

With the freedom to make all my parts at home i can do anything i want, as the system is capable to handle wood, plastic, aluminium, brass and can do PCB-engraving as well ...

The next step would be to order or make my own extruder or syringe-system, then i can go with additive 3D-fabbing.

Then i want to adapt the diode-laser and do some more sophisticated experiments ...

And then i'm in the timeline to start 'professional' working with my tripods (and then REAL BIG ONES too!) and build some reprapable setups for my own and to give to some friends ...

---

So the CNC-repstrap would have his niche as CAM for the harder and milled parts as long the plastic- and paste-fabbers are capable of otrun the CNC-milling in capability and usability ...

I think the point, when i'll switch from repstrap to reprap completelly should be somewhere in the next two years, but it's no need to throw away the CNC then

Viktor

The Way I saw it was that a repstrap is something that's been bashed together out of odds and sods to make the parts for a reprap like Darwin. Repstraps would, unique to the maker, although they'd all share similar properties (3 axis control, a plastic extruder/miller etc).

Once you've made your reprap you could simply bin the reptsrap as it would not be as accurate, nor as useful. I mean, the bed size of Darwin seems to be about 12 inches cubed (30cm cubed) while the actually resin/plastic parts for Darwin don't seem to be larger than 4 inches cubed (10cm cubed).

However should you just go straight for it and drop the money ($500 or about

... as long, as you can't fab the bed, the struts, all feedings, belts and especially the electronic and motors, ist's mainly a repstrap - Darwin and Mendel are assembled of metal-parts with many components as screws, connectors, cables and other stuff, you buy ready from shelve elsewere ...

I think the first 'real' RepRap would be something pneumatic or hydraulic with heat- or pressure-morphed elements from resins and some reprapped pistons - but the part of the electronic, powersupply and pump/motor is still to be developed reprappable ...

With my tripod-system i heavily reduced the amount of needed parts and components and it's more reprapable with actual repraps as the cartesic robots, but the software is to adapt, so it's a way too.

The next year should be very interesting, as many people with skills and motivation are involved and the count should rise fast, as the information spreads, so most hurdles are adressed to overcome!

Viktor

It seems that it will be a while till someone repraps the stepper motor. The housing may be able to be replaced with a plastic material and maybe even the shaft. But the coils most definetly willl not be able to. There is also the bushings or bearings (depending on the motor). I will have to look up the melting point of copper.

Joe McKibben

The melting point of copper is 1984.32

Hi Joe,

... i think, ready manufactured copper-wire should should be alowed to use for a reprap

Here [forums.reprap.org] i discussed a linear stepper motor, that could be self-made from iron-bolts and copper-wire for the coils.

Some other sophisticated parts (screws, nuts, ...) are also easier to make in a different process, so it's a fuzzy definition, what's a core-reprap and what's a repstrap ...

When pointing towards pneumatic or hydraulic systems, then it's easier to find reprapable parts, so it's a bit a race, what's faster - a full-pneumatic 'real' reprap, or an extruder, capable of all materials, needed for building the core-systems ...

Viktor

A RepRap doesn't have to be capable of making copper wire, as long as it is capable of making a machine that makes copper wire. Same for winding the copper wire. RepRap just has to make a winder. Making a crucible for melting copper from some kind of clay should be possible. Making a wire drawing or wire rolling tool should also be doable.

But making copper wire only becomes interesting when getting it becomes the bottleneck for making stuff. Which is still a long way off. Currently the bottlenecks are:

-Hardware - getting the special parts like extruder heads and axis stages (although both the moldings and the McWire design have relieved a lot of pressure).
-Software - printing large things without crashing (Using bits from Fab@Home or Forrest Higgs might fix this).
-FDM - making stuff without warping (using ABS or PLA should fix most problems here).

--Blerik

I think when you look at it like that it's easy to get disheartened, probably to recognize that that's the projects eventual aim, not something we're aiming for by 2008.

I'd count something like Darwin as a reprap, a device that can print a fair chunk of it's own parts, meaning you have to only buy the bare minimum. Just the ICs, Motors, wire, metal rod.

It's be nice to be able to print copper wire, or the actual IC die but thats pushing the limits far further than even commercial printers can think about achieving in the next 5 years.

I see. I was just pondering the technology that would be necesary for the RepRap to accomplish some of these goals.

"A RepRap doesn't have to be capable of making copper wire, as long as it is capable of making a machine that makes copper wire." That is an interesting concept, I didn't think of it that way.

Joe McKibben

It's like what qualifies as a computer? The first computers were less capable than the all-solar scientific calculator in my pocket, but I'd not call that calculator a computer. The calculator can't be reprogrammed, but until EDVAC, or rather the duplicate of the concept built by its inventors student, the first computers had to be manually rewired to "reprogram" them anyway.

As for wire, and wrapping.
If the reprap went to four axis, or six, you could theoretically "print" the coils using conductive ink around the axis of the motor. Also, rigging a wire-feed mechanism might prove to be a better way to "draw" traces than would paint, although I'd prefer paint, or sinterable metal particles myself.

Think of it this way. The first one gets built, and is declared the first true "reprap". The second model comes around, and the first is still a "reprap". The third, then fourth comes out, and the first is only a "reprap" out of tradition, but new versions with comparable capabilities are not considered to be "repraps" as the definition has gotten stricter. Ultimately, it might be that to qualify as a "reprap" it might have to take "raw" materials, (whether that means pure, or unrefined might itself be a point of contention,) and produce, from this, another reprap...then you can have it produce a minimug filled with Earl Grey to celebrate. :>

I think the definition of what is a RepRap vs a RepStrap is going to be a fluid one for the next few years as the machines capabilities expand. The ultimate Reprap would be something you could turn loose on the planet's surface and watch as it made a working copy of itself out of dirt and air without any help. A Darwin machine can reproduce all of its' Darwin specific parts with the exception of the actual nozzle tip and the PCB's (Somebody correct me if I am wrong here). Aside from brackets and connectors (Reprapable) everything else is stock materials, which are widely available and not specialized for Reprap. The skill set needed to assemble a Darwin is not small but aside from soldering up the PCBs not exotic.
In the next year or 2 I expect to see the printing of PCB's and the creation of most structural members within a Reprap framework. I also think the issue of automatically soldering or otherwise connecting electronic components to form working circuits will be addressed. After that things get sticky.
There has been talk on other threads about the possibility of printable linear motors for moving the cartesian platform, but electric motors and electronic components are hideously demanding to produce and may need to be considered raw materials for quite some time. While there has been some very interesting work done with the idea of printing IC's and other circuit components to order, it is questionable whether there would be much to be gained, aside from the gee-whiz factor, from doing so. Consider that the Fab@Home group has been able to print a working battery, but only by mixing up a number of exotic and unstable chemical compounds. Is there any gain in creating these compounds, moving them from place to place and disposing of them and their wastes, as opposed to doing the same with a standard factory made battery? Usually not. In the distant future when fabbing has become the predominant form of manufacture, yes.
Could a Reprap machine create the parts to build a machine to make nichrome (or copper) wire? Probably soon, but when was the last time you had a big chunk of nichrome (or copper) to feed into such a machine?
Reprap can gain a huge economic advantage over traditional manufacture by producing customized (as opposed to standardized) goods from locally available resources (as opposed to those transported from far away). But things like wire, transistors and machine screws gain their power from being standardized. If the materials for them are not locally available there is no economic advantage in making them locally from imported material. Again in a micromanufacturing economy the balance shifts, but I am not expecting to live that long.

The thing is, some of the "raw materials" such as bolts that we've been taking for granted could definitely be replaced by a clever design made out of reprapped parts (self-locking plastic pins, for example). We don't need to be able to fabricate all of the parts currently in the design, we only need to be able to fabricate parts that are functionally similar. Viktor's linear steppers are another example of this principle. Although it may be a long time before we can fabricate permanent-magnet stepper motors and threaded rod with the required precision, if these parts can be replaced by a home-brewed linear stepper motor with workable tolerances they will not be such a large obstacle to self-replication.

... i'm a real fan of 'organic designs' - imagine a halfsphere from resin, which consists of several separated inner sectors, that can be pressurized and evacuated through valves.

When you set a tool on top of this, you can very precise move and point the tip in different directions ...

With other designs you can perform more travel-range, higher forces and accuracies and build objects, that can morph in many different modes.

I give some ideas to guys at Fraunhofer, how they can build microgrippers, bending their fingers with microhydraulic, piezo- or SMA-actors or other methods capable of micron accuracy and scalable down to appropriate sizes ...

When going the other way, then i can imagine macroscopic objects, morphing and moving like the tentacles of an octopus (cuttle-fish?).

What should be developed first for this kind of driving are reprappable pumps and valve-arrays - i aready sketched some ideas and think it's no real problem to make some from silicone or polymer-resins (look here: [forums.reprap.org] )

Other ideas for 'home-brew-motors' could be found in low-temp-steam-powered systems, SMA- and bimetal-actuators, oszillating Ferrofluid- and MRF-systems and maybe in some one-way-processes like baking-powder and vinegar ...

Viktor

I don't see why screws and bolts can't be made in RepRap made molds then filled with another nozzle that extrudes molten metal into the mold. However, I would prefer clever design using locking plastic parts, than molten metal nozzles.
At some point we will have a larger machine that will 100% reproduce itself and it will drive itself off the platform to a new location and unfold itself to be the same size as the original. It is just a matter of time and if you look at the evolution of the computer, it may only take twenty years to do that. Every month from here on out we will be able to mark another advancement made in this project because more and more people with different backgrounds will get involved and push this project further and faster then any of us could dream. Of course, I

Screws and bolts are not usually cast, because castings generally aren't as strong as people would want screws or bolts to be; you don't want the threads to crack under stress.

Has any design work been done in the area of self-modifying or expanding repraps? For some jobs, the machine would have to be able to build specialized attachments or tool heads, and then use them to finish the job; the above-mentioned wire-making machine is an example. I think that this would ultimately be more practical than attempting to build a machine with every possible task allready built in.