What Are the Current Problems?

Hi,

I've read through the website and I am wondering what the current problems are that stop a Darwin machine from creating it's own parts. Can someone please point me to the relevent webpages or post a summary? I would like to understand the current state of the project better. Thanks!

regards, Andy

One of the machines was able to make the parts to a small working gear assembly, which looked like flat parts, and there was also a very small reprap machine part made, so I'm guessing that the limitation was in making big, thick, solid parts. Could someone verify this?

If this is the limitation, maybe someone could develope a variant reprap design that could be built from parts laminated together from wafers.

I think that that depends on what you mean by "thick" and "solid".

I don't think 'big, thick solid' parts are a problem... it's more a matter of making a 'big, thick solid' part that isn't solid plastic... (ie: cost/weight/runtime/etc)

inner support structure with gaps/etc and support material/etc would be more the 'problem' needed to solve the 'big, thick solid' part 'problem'...

I think the core team should respond to this but looking from the outside in the current problems I perceive are :-

1) There are delays in the comms that affect accuracy.
2) The melting point of CAPA is too low for to be used to make the motor brackets.
3) HDPE warps a lot because it shrinks when cooling, I don't know if CAPA is the same but my experience with PolyMorph would suggest it is. Perhaps Vic's PLA is the answer here.
4) I think some parts need support material but I don't think the support extruder is fully rolled out yet.

---

[www.hydraraptor.blogspot.com]

i'm only speaking for myself here, and not the development team, so dont take this as gospel word or anything, lol.

here are what i see as problems with the project currently:

problem: darwin needs a few more design iterations to get to the point where its easy to assemble, reliable, etc. until 3D printing is ubiquitous, we'll continue to be stuck in this nightmarish chicken/egg situation.

solution: create an intermediary research platform that is easily / cheaply manufactured through conventional methods. see: McWire Cartesian Bot

problem: development on the PIC is *not* fun, or easy, and we have problems with the underpowered PIC 16F648A, as well as the open source toolchain for it.

solution: arduino.

also, i agree with the things nophead said above. we shouldn't kid ourselves. this project is tough, and we have a decent amount of work to do before its easy for your average hobbyist (let alone your average person!)

that being said, i still firmly believe that we're on the right track, and that this project is achievable. we've done some really great work (polymer extruder, darwin, support extruder, reprap host software, etc.) and i'm looking forward to seeing how things progress.

I looked up PLA and it does have a smaller coefficient of expansion than most plastics do. The only thing that works against it is that it melts at such a high temperature.

As to problem(s), here's my take.

1) Just as there is a whole body of knowledge about how to design injection moulds to account for shrinkage we are going to have to learn how to design and extrude things in ways that will leave us with acceptable degrees of dimensional changes. This implies that simply designing a solid, slicing it and tracing a set of paths for the extruder on the slice isn't going to do the job by a LONG shot.

2) Right now we have very little experience with running these systems for many, many hours. I've got several hundred hours in actual operation time, but until we have many thousands, we aren't going to have a clear idea for how the parts of a reprap machine REALLY need to be designed. I've been continually astounded at what requires maintenance on Tommelise, what wears our quickly and what hardly wears out at all. Any ideas about what would give trouble and what wouldn't that I had when I started have almost inevitably been wrong.

Truth is, the good folks here keep reinventinting the wheel. Can't get the wheel rolling if you don't get it out on the road. Don't let them riprap ripoff you. Keep your money untill they can sell(RRR)something that works.

kevin Wrote:
-------------------------------------------------------
>
> Don't let them riprap ripoff you. Keep your money untill they can
> sell(RRR)something that works.

This is certainly good advice if you look at RepRap Darwin and the variations that are being developed as merely another consumer product that you'd as lief bought from someone like HP, for example. If that's where you are at you certainly shouldn't be buying parts at this point. I think that Dr. Bowyer has publicly made that clear on any number of occasions in the past few months.


If, however, you want to be part of the development process you have to be prepared to get your hands dirty. Putting up with mistakes and blind alleys is part of that experience.

Edited 4 time(s). Last edit at 09/29/2007 09:28PM by Forrest Higgs.

> I looked up PLA and it does have a smaller coefficient of expansion than most plastics do.
> The only thing that works against it is that it melts at such a high temperature.

So, integrating ( Exp(temp) dTemp ) from melting temp to room temp might be the same, or more than other plastics. Easy enought to determine experimentally.

Regarding that kevin fellow's petty little comments, Forrest, you were rather more polite than he deserved. We should send him off to go pester someone else.

You're right Sebastien, it was petty of me and I must admit it was done deliberately with bitter disgust that a project that is in the developmental stage would bring new blood into the fold selling items that cost good money when it is known that there are many problems with software and firmware and PCB designs.

Look back at all those who wanted to get involved, who bought items from the RRRF and who have been left by the side because you guys are too busy being electroheads. It's easy to move on when you can't solve a problem and call it a learning curve or blind alley, It's harder to stay the course because then you must admit that you're just not that smart and should never have gone done that alley to begin with.

Be up front with people, put it in big letters on the main page that most of you will move on to other designs and systems because you can't solve the current problems and that if you buy the products in the RRRF there is a big chance you will not need it two months from now.

I'm sorry if you can't see the intent behind my post, or that it is taken as a personal attack on the Reprap project, it's not.

Kevin Wrote:
-------------------------------------------------------
>
> I'm sorry if you can't see the intent behind my
> post, or that it is taken as a personal attack on
> the Reprap project, it's not.
>

Again, I don't see what your beef is. The foundation was set up in order to save builders money by buying bulk and redistributing in small lots components that would otherwise be quite uneconomic to buy by individuals. Recently, Zach has been trying to create a set of control boards for the Darwin reprap machine that takes advantage of his board design skills and again bulk buys. If you've ever had to build up one of those boards from scratch on stripboard even having to apply a few patches to one of Zach's boards to cure errors is a damned sight easier than starting from scratch. Zach deserves many thanks for his boards for making reprap development work more accessible to people with limited electronics skills.

If the Reprap project had been misrepresented as a proven technology DIY exercise, I'd see that your rather tedious jeremiads might have a point. Anyone, however, possessed of even a room temperature IQ can get a ready grip on the fact that Darwin or even Reprap as a concept is by no means a "done deal" simply by following the blog entries and forum patter.

I must say if your posts aren't to be taken as an undeserved smear on the project then your intent is pretty opaque to me at least. What you are attempting to do here rather reminds me of those wonderful sorts of people who try to sue McDonand's for making their coffee too hot or sue the overhead fan companies when they loft their newborns into one, viz, manufacture a case of grievance where none has any proper right to exist.

Let me ask you this. Have you personally bought a bunch of components from the Foundation and feel that you have been somehow robbed? If you have, I'm sure that there are several here who would be more than glad to take up a collection and buy them off you so that the rest of us could get on with trying to make the goal of Reprap a reality.

Edited 3 time(s). Last edit at 09/30/2007 03:22PM by Forrest Higgs.

My take on Kevins comments is that he has a point but that in the grand scheme of things its a very small one.

Currently the whole project (especially the electronics in the last few weeks) are in a state of flux. My opinion on the matter is that this is quite clear from following the posts on the forums even if only for a week or so. It might be helpful to put disclaimers on various key pages and probably on the RRRF site making this clear to those who just give the project a once over before diving in and forking out money.

Most of the stuff sold currently does actually work so its not so much a matter of being left with non-working parts. Its more that it may be harder to make every thing work together if you buy it peice-meal over a period of months (as I'm doing and I would recommend other newbies do as well).

It is also something to keep in mind as the project advances and try to make sensible decisions on what sort of backward compatibility should be supported. I think quite a good job has been done up till now and long may it continue.

In the medium term i.e. once the Darwin 1.0 is out in the wild and replicating like mad, this becomes less of a problem as we can just archive software files of the various versions of the printed parts. The non-RP parts will be more standard and take less time to construct - therefore less of an issue in both the rate of change and the time you need to build it.

In the longer term I expect there will be many forks of the Darwin 1.0 'kernel' and those who wish to can set up their own website selling the physical items needed to build their version.

I see our current state as one of trying to get the best possible design and making it easier for the future at the expense of a few blind alleys for the present.

Hey Kevin,

I'm sorry you feel that way, but I understand where you're coming from. As the others have said, we're definitely still in the development state. Also, even when we are out of that development state, things will change. Thats the nature of this project, open source, and life in general.

I'm not entirely sure what changes you are taking offense to, so let me break some of the changes down for you:

Electronics - yes, its true that i'm designing a new set of boards that depending on how they turn out could be a candidate for v2.0. however, the current boards we have now *do* work. i'm personally unhappy with how easy it is to use them, their somewhat limited capability / non modularity. i've set out to design a set of boards that solve the problems i have personally encountered. if we never fix any problems, then how can we make any progress?

Mechanics - i'm also working on a lasercut style design that will be easy to get manufactured. if you've followed the forums, you'll know exactly the types of chicken and egg problems we're running into. hopefully this will solve that, and if not... well then its just another mistake i've made along the way.

Extruder - the extruder design has pretty much stayed the same for a while. it works, and it does a good job.

One thing to note is that each of the systems above is designed to be interchangeable. the extruders will all work regardless of what type of machine its mounted on, the electronics are all capable of driving any cartesian machine, and the software is capable (or can be made to) the electronics to drive the machine. since we know that each part of the system *will* change due to the nature of development and progress, we've made a good effort to keep the systems modular and interchangeable.

The nature of open source means that everything we do is exposed to the public. That means that we share our experiences with you: both the positive and the negative. We're not some closed door company that only tells the public about our successes. If we fail, its out there for everyone to see. Some of us are also amateurs. But we have passion and a drive to succeed, and I think we will.

Finally, if anyone has bought something from the RRRF and is not satisfied, please contact me. If you were under the impression that the raw PCB you were buying was the final, perfect version, then I'm sorry. I'm willing to work out some sort of refund or something. The RRRF is a non-profit foundation setup to help researchers. We're a long way off from selling finished products, but all of the stuff on the store is useful to people. Hopefully you haven't decided to give up on our project. For all its flaws, RepRap is a beautiful idea and a dream that is shared by many here.

Well said Zach. I have taken up enough of everyone's energy on this issue.

"Silence is a form of complacency"