[Gavilan Steinman] just printed and assembled his own RepRap machine and filmed the process. This isn’t news but we found it very interesting to watch. He started with a RepStrap, a rapid-prototyping 3D printer that as built by hand instead of printed by a similar machine. This is the seminal step in the self-replicating process.
From there he prints an extruder head which improves the quality of the parts the RepStrap can produce. We then see time-lapse footage of the printing process for a Mendel unit, the second generation of RepRap machines. We’ve embedded the video after the break. It’s a great way to spend ten minutes on a Sunday afternoon.
33 thoughts on “From RepStrap To RepRap; A 3D Printer Is Born”
Cool. Dude knows how to make a video as well.
I would love to build one of these.
Now, take this with a “salt of grain(?)”:
Seeing the video I first notices his setup – two desks, 3xLCD displays, notebook and 2D & 3D “printers”. Looking at that the video represents no feat of challenge in my point of view. If the guy can afford all that equipment, why not go all the way and buy something professional?
Saying that I would welcome any kinds of ‘cheap’ hacks that people with say no or little income could afford to implement.
But hey, it’s just my thought..
anyone wanna start a group to pool funds/parts to follow this plan of boot-strap to reprap and use that to produce parts for the initial investors to make their own mendels?
Brilliant video indeed!
>”…If the guy can afford all that equipment, why not go all the way and buy something professional?”
Uh, because the journey is more important than the destination? Because you don’t gain any soldering skills by pulling out the mastercard?
And how do we know he didn’t hand craft those computers from parts the same way he made the worktables everything sets on?
If I’m envious of anything, it’s the clutter-free workspace.
The computer equipment really isn’t all that expensive, and “professional 3d printers” are still costing thousands. 1500 bucks in computer equipment and a few hundred in parts is still by far cheaper than 5000 dollars for a basic 3d printer.
On a second note, I hope he finishes the electronics off with a more solid and hopefully professionally done pcb.
did anyone notice the subliminal “open-source-is-the-future” message ??? CREEPY AS HELL..! lol
The Mendel is just sexy, but sourcing all the parts is a long term project on its own, which is a shame really.
I reside in the UK and dying to repstrap a mendel if anyone has any source please share.!
I always get all excited when I see RepRap news…and then I notice how much of the machine _isn’t_ printed. I mean seriously, he prints what, maybe 10 parts? Appears to be less than a tenth of the machine that is actually printed, and it appears to be parts that I could probably build faster than it prints out of scraps of wood and aluminum that I’ve got laying around. I mean, I love the idea of RepRap, but it still has a long way to go before it’s anything near replicating. Let me know when they can print everything but the electronics….
The problem behind printing everything except the electronics goes back to accuracy and what you have beforehand. You can’t print the aluminum or steel crossbars, you can’t print the screws/bolts/adjust heads. He printed the majority of what needed to be printed. Again, I hope he switches out the breadboard for actual pcb though.
I work with G. Steinman. I will send him here to answer your questions!
I think Urza’s point dealt directly with the “he printed what needed to be printed” bit; what exactly NEEDS to be printed? It didn’t look like anything he printed for the Mendel couldn’t have been fabricated with common materials and a little skill, possibly even faster than waiting for them to be printed.
Realistically, what is saved in terms of time when going this route? If he had put just a little more time and effort into the original build, there would be no need for it to build parts for the improved version. It seems like people get a little too wrapped up in the idea of self-replicating machines.
On the other hand, machines like this are excellent for one-off items or objects intricate enough to make hand fabrication impractical (well, at least, sufficiently annoying).
Hey, nice work, and he used my extruder design to boot. Good to see it getting replicated.
There’s always somebody that complains about the percentage of replicated parts, but as long as you’re within reach of a hardware store, there’s not much point in trying to print stuff that you can buy cheaper.
Once we start building these things on the moon, then we’ll have to worry about not using any outside parts. Then we’ll also have to worry about berserkers. :)
Also, just to reply to MS3FGX, there’s quite a difference between even a good Darwin and a Mendel. I have two Darwins sitting on my workbench, and a single Mendel blows them both away in terms of speed, reliability, and accuracy. Plus, the Mendel copies itself (at least the 100 printed parts) every 48 hours or so with little input from me; hard to do that with hand tools and common materials.
Well, it really isn’t a “self-replicating” machine, he’s building a different type. Darwin vs Mendel. Not only that, he printed a new extruder for the old unit, to help print out quality parts for the new one. In addition to that, he shows that the old one was crafted out of wood using hand tools. He saves a lot of time, increases accuracy, and probably has a better setup going than if he would have taken the time to mill it out of aluminum and other parts.
He did very well, and made the video an excellent showcase to what he did. I hope he adds more to the build layout so maybe others can replicate it. It looks very nice, and is an excellent example to what a reprap can do.
Cool setup. Maybe it is time I stop lusting after a 3D printer and build one.
Great job, love the slick video. Not to take anything away from all the hard work going into these 3D printers, they’re a great concept, but my main gripe (and why I don’t have one) is the lack of accuracy.
Take a look at this – http://www.thingiverse.com/image:10754 produced with a Makebot and this – http://www.pp3dp.com/index.php?option=com_content&view=article&id=130:cathedrals&catid=49:art–photo&Itemid=70 produced with essentially an extruder on a steel Cartesian frame and even the more professional one shows lines and burn marks.
Anyway to up the precision on these?
@deathventure: Must be some other universe I live in, >1500$ any kind of $ is still awful lot of money for me..
cool. i wonder about how much they cost to build…
The “Rep” in “RepRap” is supposed to stand for “replicating”. It does not even remotely do this.
It’s also a bit disingenuous on the one hand to DISMISS buying vs building equipment because “it’s part of the journey” and then JUSTIFY buying parts vs printing them because “it’s easier and cheaper”.
If RepRap had some realistic path from where they are (printing corner brackets that hold together the other 80% of the machine) to (nearly) full true replication, I’d find them a lot more interesting. After a year of subscribing to their blog, I never saw such a plan.
You also don’t know if he bought it all at once vs pieces over time too. I don’t know about you, but my overall pc cost (including monitors, new tower and such) probably around $1500, but that was over time. The tower itself was only 650 one time cost. The rest of the peripherals were added over time.
I’m really enjoying these interesting perspectives! Keep them coming!
I’m happy to share the details of my setup with you. The tables I made out of doors and saw horses for $30 per table. The desktop computer in the background I built. The tower was made from newegg parts for $300. The 17 inch monitors were $75 each on ebay.
The Darwin cost me ~$700 because I was learning and making rookie mistakes. The Mendel cost me $300. Wow, I spent $1,000 on 3d printing tech over the course of 2 years. That’s a lot of money. However, I now know how to write firmware and hostware to interface with microcontrollers, thermistors, endstops, and microstepping motors. I know how to solder through hole and surface mount components. I know how to model CAD parts to spec, and design schematics and circuit boards with Eagle. All this with only a software engineering background.
I feel I got my money’s worth.
I can’t believe no one picked on my t shirt. Any thoughts on what my shirt should say in the next video I do?
Thanks for all the perspectives!
That’s a pretty excellent breakdown. Done the old door tables thing before. Nothing like reusing old materials for new stuff.
700 for a basic first off printer like that isn’t bad either. I may have to give it a go.
Are you going to replace the breadboards with a pro pcb, or leaving it as is? I think it would be a good investment. Leaving some open IO’s for any possible addition or work arounds as well when you’re planning. Otherwise, I think it’s excellent.
Dude, what the hell? Everybody knows a computer nerd worth his degree only wears black t-shirts! White t-shirts should only be worn on laundry day, or under your Storm Trooper hoody sweatshirt.
Nice builds dude.
This really is the way video’s should be done! Only thing I’d suggest is a second camera on the table to show the parts.
ok as you mention your shirt…
well there was about 10 seconds @ 25.000x printing stuff and some more minutes assembly-time @ 1.000x
only from that the project took longer then 86 hours…
AND YOU DIDN’T CHANGE YOUR SHIRT ALL THE TIME !!!!
“It’s also a bit disingenuous on the one hand to DISMISS buying vs building equipment because “it’s part of the journey” and then JUSTIFY buying parts vs printing them because “it’s easier and cheaper”.”
Diminishing returns? There’s no real cheap or easy way of fabricating small quantities of custom 3d parts to reasonable tolerances. You can buy, sure, but it’ll cost you a fair few pennies.
On the other hand there are well established ways to build, say, stepper motors cheaply, in large quantities and to reasonable tolerances. You could try and make your own, but it will take you a while, probably cost you more and the end result may not even be as good as parts you could have purchased.
People seem to forget things when commenting on this specific project.
1. because he can
2. because it involves more people than just him
3. because it’s fcking awesome
4. because while the machine is working he can do something else.
A commercial 3D printer is way over the price range of the RepRap – and who cares if his equipment may or may not be more expensive than the project? I fail to see the point here – it’s open source meaning people other than himself will benefit from this. If he had “just bought” a 3D printer he wouldn’t have gained any knowledge and he wouldn’t have done YOU a favour either.
Are you seriously flaming the guy for doing YOU a favour? Come on :D
I for one think it’s awesome as i’ve always wanted a 3D printer for small parts. I don’t have the space for a workshop and a printer like this would help me a lot.
For those that comment it on the replication matter please bear in mind that no machine to this day is 100% self replicating – if it was it should also gather its own resources or what?
Barry, thanks for the second camera angle suggestion.
deathventure, yes I’ll go to a pro pcb eventually, but I still need to add a hot bed and possibly more heads, so I don’t want to lock myself into my current vision, even if I leave some IO open. I could need more.
All, thanks for the support.
i like to say this guy is totlaly alsome! if you go to his website and look at his work bolgs youll see what he did to get ware he is today he is even making his own boards true its crude and not plesent to look at but it works and that what counts in my book
The beauty about circuits in general is that they can be modularized if you want to look ahead too ;)
Board stacking would be a good option. Locking yourself into your current vision will only apply if you let your design be locked into 1 point. Expansion boards can be made and put in place if you design it well enough.
To do with your comment on accuracy. The images you show, while the same file, are not printed at the same scale. The Makerbot and UP printers have approximately the same resolution (as do the much more expensive Stratasys machines). However, the UP and Stratasys have larger print volumes, which allows parts to be printed larger, and therefore, to have a higher signal-to-noise ratio in the finished print.
You do bring up a valid point, however – Repraps and Makerbots allow control of all print parameters, which means that they also allow operators to get any of these parameters wrong :) Since most of these machines have only been in the wild for a few years, most makerbot and reprap operators have less than 100 hours of printing experience. This combination of high control and low experience means that lots of prints made by the open 3d printing community are lower quality than they could be.
When you buy into a proprietary print system, with proprietary plastic material, you get higher quality output, and you sacrifice some of this control.
As other people pointed out though, you also sacrifice the community, and the innovations made available by the community. I doubt very much that the UP printer people will make an improved extruder head design available for free, every 4-6 months (as the reprap community has done for the last 4 years).
In my opinion, this alone makes open printers worth the time it takes to learn to use them well.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)