The future

I am fairly new to reprap. I have read all the recent blogs, read a lot of posts on this board, looked at makerBot videos, visited thingiverse, etc. It seems after taking all this in that RepRap and its derivatives still haev some way to go to being truly useful in any practical way.

Setting aside the issue of just building one, which is non-trivial, there are key problems. From a pure economic standpoint, the machine is no practical. By this I mean that in its current form, I cannot produce enough useful goods such that the economic benefit of the machine exceeds its cost. The only exception seems that I could produce parts for other Mendels, etc. and sell them to recoup my investment. But I am more interested in the standard consumer who is wanting to manufacture everyday goods for their own use instead of becoming a retailer of goods.

So what needs to change for the cost\benefit to change? I list below my current perceptions of the RepRap's weaknesses:

1) The resolution is poor. Most pictures of parts show ripples and uneven surfaces. Corners are not well formed. Commercial printers have much better resolution...and are even capable of producing parts that can be used in mechanical devices right out of the machine instead of just being used as non-working prototypes. RepRap needs a major improvement in this area. I have not seen any resolution to this being proposed yet.

2) The build area is too small. Presumably reprap can be scaled to any arbitrary x\y\z configuration, but to date this has been an issue. Printing time takes too long for large builds, and it appears that larger parts even on todays machines suffer from warping. A practial printer should be able to print chair size objects...even if the chair needs assembled. So we should be talking in terms of square feet instead of square inches of volume potential.

3) The choice of materials is limited. A truly practical machine will be able to print metallic features as well as various plastic features, and to do so interchangeably. This will enable at least simple electronics which greatly expands the possibilities of the printer. I know some research work is being done by reprappers in this area.

4) Printing items that have features > 45 degress is not possible. If i want to print a plate for example, I cannot do this because plates have a gradual slope at the edges. Commercial printers enable some supporting material to be laid to enable this. The supports are then removed via washing or a chemical process after production. Not being able to print these features is a major limiting factor.

5) color is not supported. while this is the least of the issues, we cannot dimiss the importance of color on printed goods. Solid white or solid black is not the norm for many everyday objects.

6) there is real lack of useful designs for everyday items. this is probably due to the current limitations of the technology since it does not make much sense to design something that cannot in practice be printed on a reprap. But I do think that a concerted effort should be made to start to build up thingiverse with practical everyday items. Plates, bowls, drinking cups, table lamps, chairs, etc. i think that laser scanning is the best way to go here as it is much faster than someone designing from scratch. We already have the goods to scan, so no need to reinvent the wheel.

I am well aware that RepRap is intended at this point as more or less a research tool with the hope being that if enough people working on it that somehow miraculous solutions will be arrived at. I have some doubts that this will work. It seems that each of the above issues should be taken up by teams of people with proper expertise in materials science, mechanical and electrical engineering as well as software engineering. I think that these issues are *hard* to resolve, and this means that a directed team effort is required to solve them.

I am very excited about what has been started here. i am a big advocate of a resource based society, but I beleive the only way to get there is through replication for the masses. this is the disruptive technology that can be the great equalizer for the everyone and can start us towards a society of plentifullness for everyone....

I look forward to your comments...

Kevin

...here is a sub $3,000 printer that offers som very nice capabilities versus makerBot\RepRap...

Perhaps the commercial enterprises will address the concerns I list above at a low price point. Of course I would prefer the open source solution and these commercial items cannot replicate themselves either...

[www.solido3d.com]

Kevin

Quote
maitri982
It seems that each of the above issues should be taken up by teams of people with proper expertise in materials science, mechanical and electrical engineering as well as software engineering

I wouldn't doubt the skills of the people here. Some of the community members are indeed expert/professional engineers, although many others may just be hobbyists. But everyone is running on the same, uber-low budget, so the challenges will always be high. I'm quite impressed already with what's been achieved so far!

I don't want to diminish what they have done at all...I probably could not have done it.

But right now its a 3-axis movement device with an extruder attached. I would call this the "low hanging fruit" of the ultimate goal. i think much of what has been done has to be rethought...especially the extrusion process.

If my situation (time and money) were different, it would be fun to assemble a team to try and address these problems head on...alas I am not in that position as we speak.

Kevin

maitri982 Wrote:
-------------------------------------------------------
> I am fairly new to reprap. I have read all the
> recent blogs, read a lot of posts on this board,
> looked at makerBot videos, visited thingiverse,
> etc. It seems after taking all this in that
> RepRap and its derivatives still haev some way to
> go to being truly useful in any practical way.
>
> Setting aside the issue of just building one,
> which is non-trivial, there are key problems.
> From a pure economic standpoint, the machine is no
> practical. By this I mean that in its current
> form, I cannot produce enough useful goods such
> that the economic benefit of the machine exceeds
> its cost.

It's currently a research project. They didn't find a good use for lasers for 20 years after they were described, and didn't make any decent lasers for another 20 years after that. We're hoping for a slightly shorter timeline, but the same principles are in effect.

>
> So what needs to change for the cost\benefit to
> change? I list below my current perceptions of
> the RepRap's weaknesses:
>
> 1) The resolution is poor. Most pictures of parts
> show ripples and uneven surfaces. Corners are not
> well formed. Commercial printers have much better
> resolution...and are even capable of producing
> parts that can be used in mechanical devices right
> out of the machine instead of just being used as
> non-working prototypes. RepRap needs a major
> improvement in this area. I have not seen any
> resolution to this being proposed yet.

You haven't seen some of the ABS parts coming off nophead's heated bed? They're stunningly perfect compared to the average reprap.

> 2) The build area is too small. Presumably reprap
> can be scaled to any arbitrary x\y\z
> configuration, but to date this has been an issue.
> Printing time takes too long for large builds,
> and it appears that larger parts even on todays
> machines suffer from warping. A practial printer
> should be able to print chair size objects...even
> if the chair needs assembled. So we should be
> talking in terms of square feet instead of square
> inches of volume potential.

yep reprap can be scaled to as large as you want, however smaller scales are far easier to get started with. Even the wright brothers started with small scale models, and moved up when things looked like they were working nicely.

> 3) The choice of materials is limited. A truly
> practical machine will be able to print metallic
> features as well as various plastic features, and
> to do so interchangeably. This will enable at
> least simple electronics which greatly expands the
> possibilities of the printer. I know some
> research work is being done by reprappers in this
> area.

This is just a matter of time. The desire to print circuits into objects is the main driving force, however once you can print metallics in you can do as much cosmetics as you want

> 4) Printing items that have features > 45 degress
> is not possible. If i want to print a plate for
> example, I cannot do this because plates have a
> gradual slope at the edges. Commercial printers
> enable some supporting material to be laid to
> enable this. The supports are then removed via
> washing or a chemical process after production.
> Not being able to print these features is a major
> limiting factor.

There's plenty of development around the paste extruder at the moment for this exact reason

> 5) color is not supported. while this is the
> least of the issues, we cannot dimiss the
> importance of color on printed goods. Solid white
> or solid black is not the norm for many everyday
> objects.

Apparently scribbling on the filament with permanent marker before it goes into the extruder colours the extrudate nicely, however it takes a fairly long time to change colour because you have to wait for the chamber to be flushed of the old. There's also stuff being done with changing extruder heads.

> 6) there is real lack of useful designs for
> everyday items. this is probably due to the
> current limitations of the technology since it
> does not make much sense to design something that
> cannot in practice be printed on a reprap. But I
> do think that a concerted effort should be made to
> start to build up thingiverse with practical
> everyday items. Plates, bowls, drinking cups,
> table lamps, chairs, etc. i think that laser
> scanning is the best way to go here as it is much
> faster than someone designing from scratch. We
> already have the goods to scan, so no need to
> reinvent the wheel.

This is what thingiverse is for. It's rapidly growing

> I am well aware that RepRap is intended at this
> point as more or less a research tool with the
> hope being that if enough people working on it
> that somehow miraculous solutions will be arrived
> at. I have some doubts that this will work.

> It
> seems that each of the above issues should be
> taken up by teams of people with proper expertise
> in materials science, mechanical and electrical
> engineering as well as software engineering.

That's exactly what the reprap project is! Many of our contributors have significant professional experience in these areas, and even those with none are coming up with new ideas, trying them out and telling us what happened, good or bad.

Personally I've been employed making robots for a large engineering firm, where my most demonstrably useful skills were cross-discipline debugging and microcontroller firmware wrangling.

> I
> think that these issues are *hard* to resolve, and
> this means that a directed team effort is required
> to solve them.

Again, that's exactly what the reprap project is! If you want all the best hackers in the world to work on a project, there's a simple 3-step process:- first you make it interesting, second you make sure the bar for entry is reasonably low, and third you open source it. See The Cathedral And The Bazaar by Eric S Raymond for an in-depth discussion of this.

> I am very excited about what has been started
> here. i am a big advocate of a resource based
> society, but I beleive the only way to get there
> is through replication for the masses. this is
> the disruptive technology that can be the great
> equalizer for the everyone and can start us
> towards a society of plentifullness for
> everyone....

Every good idea and well executed experiment (successful or otherwise) is a step closer to this, so instead of telling us how our research project isn't commercially viable (which we don't care about because we're hackers, not businessfolk), help us iron out the bugs!

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

Thanks for the reply Triffid...

i did visit nopheads site a few days ago, but did not see any major advancement in finish...I'll check it out again...thanks for the reference.

I am not concerned about "commercial potential"...exactly the opposite! I want an open source solution for the masses that has real practical value. I merely laid out what my novice assessment of the current state of the research.

these points are known by everyone who works with the machine...so i didn't have any special acumen in comprising the list. I wish I was competent enough to help, and I may contribute at some point, but I think I will wait for what comes after mendel...or just build my own (much larger) rig. But until the extrusion process is refined a bit to improve the product finish I could have the best rig on the block and it won't matter cause I won't be satisfied with the end product.

Better extrusion is clearly possible as Stratasys does a good job of it...so i wonder what the secret sauce is for them? and wonder why we can't replicate it here? Does it just come down to laying a thinner bead?

Anyway...i hope this project succeeds at its lofty goals...

I'll reply to myself and say that Nophead has done some good work...here are some good pics:

[sites.google.com]

if we improve on this slightly, solve the 45% angle issue and do it all larger...then you have something that can print usable everyday item methinks...

and here are Stratasys tolerances...which is a good mark to aim for:

The results of the accuracy study confirmed that the FDM 400mc
produces parts to a tolerance that is the greater of ±0.005 inch (0.13
mm) or ±0.0015 inch/inch (0.04 mm/mm). The specification is based
on a 95 percent certainty level (two sigma).

These machines are by definition "rapid prototyping", not "rapid production". In the days before rapid prototyping, skilled modelers built masters entirely by hand, tolerances dependent on the individuals' skill. A master rapid prototyped to .5mm tolerance establishes proportion and takes much less time and skill to improve that tolerance figure, consequently surface finish as well. Since proportions are maintained, it's very easy to design parts that have alignment and attachment registers that push together, making the machine build envelope size irrelevant. Consider Lego. Cost/benefit ratio jumps rapidly when a prototype is used as a master for molding or casting by other means. Imagination is often much more limiting than equipment.

Quote

I look forward to your comments...

I agree with most of your stuff, and will even raise you one: it's not reliable enough that you can just turn it on and walk away.

My view of the future goes something like this: Reprap will get better and better, but also more expensive. Heated platforms aren't free, you know. And developments to come (more heads, etc) will also cost money.

Professional machines will get cheaper and cheaper. There will also be 3D Kinkos, where you can get something printed w/o buying the whole machine. At some point, the cost difference between a professional machine and a DIY machine will be minimal.

Then nanotechology/molecular manufacturing will come along and blow it all out of the water. Plus people will spend a lot of time in virtual reality and won't need as much real stuff

And (almost) everyone who's developing Reprap will have a lot of fun doing it. But most who buy it this year expecting "a China on the desktop" will be disappointed. Maybe next year...

My assessment is exactly in line with yours. We are already seeing the 3D printers come down in price a lot. i think in 2-3 years we can all have one on our desktop if we like...it will be hard for open source to keep up...although I wish them the best.

I look forward to more advancements wherever they come from...particularly in electronics printing and bigger scale. If we get these two improvements then much can be done by the average joe to produce every day items which is what I'm most keen on. If these are just prototyping tools their societal impact is limited.

Professional machines will probably get cheaper, but I'd be surprised if their stock materials came down in price. Today, inket printers are super cheap, but the price of ink is thousands of dollars per litre.

I hope that cheap professional machines will help to spread the Reprap, as more people with printers (of any sort) means that printable parts will be more easily available, so more people will be able to get involved.

As well, keep in mind that the Reprap is not able to merely print itself, but also its own upgrades, including additional print heads that could potentially work with additional materials.

There's an expression which I heard recently, I think it was from a video linked from the Makerbot blog, discussing the future of open-source hardware. "No matter who you are, the smartest people don't work for you". I'm sure that part of the reason that professional 3D printers produce high-quality results is because they use precision-machined parts, special printing materials, and well-developed software. Not because their engineers are better than the ones working on open-source projects.

(I wouldn't be surprised if Stratasys' software engineers spend some of their free time developing Linux...)

Anyway, I'm excited to see how it will go in the future. No doubt, it will only get better than it already is, and it already is quite awesome.

I think most of - if not all - the questions raised are valid, also the premise that commercial 3d printers will come down in price.

@maitri982: I think you need to figure if you are in for the "ride" or for the "goal", if you want the "goal", high-rez, high-reliability, high-cost-quality-ratio etc. it is probably too early days. If you are in for the "ride", i.e. help make these things come true, and help things just for the pure fun of it, then the time is now. By taking part in the "ride", I believe you can help influence the "goal".

Call me a misanthrope, but I think the nature of a future low-end 3d printer will depend a lot on this - and similar - projects. Potentially 3d printers and rapid fabrication - rather than prototyping - will have the same effect of physical objects as MP3 has had on the music industry, with exactly the same problems, protection of IP rights, difficulties in finding business models, etc.

I'm almost certain that a commercial - low end - 3d printer will try to implement some kind of tie-in to that manufacturers other deliverables, both raw materials and designs. Think ink-jet color cartridges and a DRM version of Thingiverse.

I personally am just trying to get a handle on where the technology is versus where I think it needs to be to have a meaningful impact on society at large. As I mentioned before, the cost benefit has to be there for this to have some impact...meaning i need to be able to print useful items with greater value than the printer and its materials.

Right now it may be "cool" to print a simple coat hanger, but this is a long way from having a useful cost benefit on the scale I would like to see. My desires here are driven by my personal concern that we are already heading for a tipping point where automation will increase the efficiency so much that human workers for many jobs will become unnecessary. this will lead to major societal issues as the unemployment roles will swell. So automation is very good, and also very bad.

So where does 3D printing come into play? I beleive we have to completely change how society works. I think the capitalistic money\greed driven system was fine, but never really worked. The division between poor and rich is extreme, wars proliferate due to profiteering motives, prisons are packed and politicians are owned by corporations. This is appearing to be no longer sustainable.

So this means moving towards a Star Trek type society where money no longer is needed and everyones basic needs are met. The technology is almost there now where such a society is possible. many factories can run with almost no workers. We have plenty of resources such as energy and food for everyone on the planet, IF you remove the profit motive.

BUT...i think one of the major enablers of this new society will be at-home manufacturign of everyday goods. So I am hoping that in the next 5-20 years that some major advancements can be made in this area because I believe it can help make the necessary transition less painful. If many essentials, including homes, can be printed, then this makes it easier to provide for the masses in every corner of the world.

Sorry for the disseration, and I know this is radical thinking to many who can't conceive of a world with no money, but I think this is where things must go. i am just worried about the transition, which will be painful.

Kevin

maitri982 Wrote:
-------------------------------------------------------

> 3) The choice of materials is limited. A truly
> practical machine will be able to print metallic
> features as well as various plastic features, and
> to do so interchangeably. This will enable at
> least simple electronics which greatly expands the
> possibilities of the printer. I know some
> research work is being done by reprappers in this
> area.

How many commercial machines can o this? Zcorp prints in plaster, Stratasys in a range of polymers not that different from RepRap, their 15.000$ entry machine prints only in ABS, most of the systems print in 1-5 materials. Objet is the major difference but hey, they start at 150k?


> 4) Printing items that have features > 45 degress
> is not possible. If i want to print a plate for
> example, I cannot do this because plates have a
> gradual slope at the edges. Commercial printers
> enable some supporting material to be laid to
> enable this. The supports are then removed via
> washing or a chemical process after production.
> Not being able to print these features is a major
> limiting factor.

This is in my opinion the most valid point and one I would rely like to see some progress on, the other points are mostly "hey it should be better/faster/smaller/...", which is valid but not strange when you look at the young age of this system (2-3 years) and the rockbottom price point. Or "should do this and that" that isn't even available on 75% of the commercial machines. I hop you are familiar with the reality of rapid prototyping. Most print service companies have machines from ALL the major brands so to offer you the wide variation of materials. Btw I am printing in ceramics with some easy mods to my reprap.


>
> 5) color is not supported. while this is the
> least of the issues, we cannot dimiss the
> importance of color on printed goods. Solid white
> or solid black is not the norm for many everyday
> objects.

Again, most commercial brands print in cream or limited colors. Zcorp has full color but their models are not suitable for functional proto's.

>
> 6) there is real lack of useful designs for
> everyday items. this is probably due to the
> current limitations of the technology since it
> does not make much sense to design something that
> cannot in practice be printed on a reprap. But I
> do think that a concerted effort should be made to
> start to build up thingiverse with practical
> everyday items. Plates, bowls, drinking cups,
> table lamps, chairs, etc. i think that laser
> scanning is the best way to go here as it is much
> faster than someone designing from scratch. We
> already have the goods to scan, so no need to
> reinvent the wheel.

no, designing from scratch is much better. 3d scanning is a very labour intensive task, its not your everyday xerox... Reverse engineering from 3d scans is something they tend to avoid in the industry too. Besides that, you cant scan mechanical items like a scissor and expect them to work. If you feel like you need practical models on Thingi, go and make them. There are lots of very functional things on there but it is what people need themselves so don't blame them for not making the things you want.

>
> I am well aware that RepRap is intended at this
> point as more or less a research tool with the
> hope being that if enough people working on it
> that somehow miraculous solutions will be arrived
> at. I have some doubts that this will work. It
> seems that each of the above issues should be
> taken up by teams of people with proper expertise
> in materials science, mechanical and electrical
> engineering as well as software engineering. I
> think that these issues are *hard* to resolve, and
> this means that a directed team effort is required
> to solve them.

They are here, they are just not the most vocal group. Rather working in garages and labs all over the world.

>
> I am very excited about what has been started
> here. i am a big advocate of a resource based
> society, but I beleive the only way to get there
> is through replication for the masses. this is
> the disruptive technology that can be the great
> equalizer for the everyone and can start us
> towards a society of plentifullness for
> everyone....

Roll up your sleeves

>
> I look forward to your comments...
>
> Kevin

thanks for the comments unfold...and thanks for providing more insight on 3D scanning, although I think that it is likely that scanning is legit for many items, but perhaps not for most.

I am willing to roll up my sleeves, but I have learned it is smart to understand at least at some level what the heck is out there and what I want to do and whether I am competent enough to actually move the needle forward.

I am not sure RepRap is the best way to make an impact yet...it appears some of the best work is being done by naphead, but he does it on a modified CNC machine, not a full RepRap. He seems to be on track to do some very interesting things and it might be interesting to follow his lead.

It seems like the goal of self replication is holding RepRap back a bit. I believe that the production of goods by an individual might be more akin (in the near term) to a workshop where there are several tools and not just one. In that case the shop is able to replicate the replicator and itself, but the replicator cannot replicate itself as a pure extruder. So a replicator that also becomes a CNC machine and a torch and a drill perhaps seems very logical. I mean the basic idea of a cartesian coordinate movable structure with multi-tools seems most practical until things like nanoassemblers come along.

i am still trying to understand all the issues. This is one of many areas of interest to me around a resource based society, another being automated hydroponic farming. I run a software company right now and have a 1 year old son...so i need to manage my time as best i can.

I would like to see your ceramics...do you have them on the web?

Kevin

Kevin,

I am really glad to see you speak up on some issues that I think are very valid. It seems to me that there are some serious problems with the reprap as you outlined but a lot of the community back and forth ends up with simply rationalizing those shortcomings... I am with you in your "let's get is done" attitude.

The self-replication goal to me does seem like a serious obstacle. However I am not against it, in fact conceptually I am for it, but practically right now it is the major bottleneck holding back widespread adoption.

I have a mcwire running now (about 4 months of effort from initial research to build) only to find out that it is going to take something near 400 hours of painstaking print time to actually build a mendel (my math could be wrong on that time, someone PLEASE correct me if I am wrong). Certainly less if I mold and pour the duplicate parts at which point I could produce a lot more for others...

To recap my responses to your other points briefly:

1) "The resolution is poor": To me a properly tuned device has adequate resolution for many things like mold making, prototyping and tooling masters. I have some stratasys printed parts and while marginally better there is not a huge difference on my examples.

2) "The build area is too small": To me a 1 cubic foot total build volume is ideal (1x1x1) . Larger than that I don't see cases that would make economical sense to print things in plastic (abs filament is easily $15-20 per pound) and there are issues with heat and warping at least with FDM. Sintered powder can address this and also the angle issue.

3) "The choice of materials is limited": To me a base plastic is fine, direct manufacture just isn't that important for my needs. Standard manufacturing processes are pretty good like injection molding, casting, etc . Having a reprap makes those existing processes very accessible to anyone because you are cutting the master and tooling costs by 90% or more.

4) "Printing items that have features > 45 degress is not possible": In designing you can work around a lot of this issue by including support structures and so forth that get cut or melted off in a finishing process. Also a long term move to a sintering technology could resolve this.

5) "color is not supported": I don't really see this is an issue. Many finishing processes permit the application of color after the item is completed or again using reprap to produce a master for a traditional manufacturing process. You can flexibility in the base color at least.

6) "there is real lack of useful designs for everyday items": Thingiverse is a good start. Things like cups, plates, etc. I am not so sure about those, they are probably not food safe depending on your filament and printing very high volume items like that will never be cheaper on a reprap even in terms of electricity consumption. Also spending a few days learning a tool like sketchup means that you can create a wide variety of items in a few minutes or hours, there is also the 3d warehouse linked with that.

As a software guy you will most likely be horrified at the state of the reprap host software, replicatorG (much better but still some major issues) and skeinforge (Is that Tk UI from 1990 or what?)

In concluding my response I would call for feedback on the following:

1) Is there are 3.0 working group? If not lets start one.
2) To what extent is the self-replication goal something that should cause a project fork. I.e. I support it only if a single machine can be built with standardized OR printed parts (not necessarily identical representations of one another but practically interchangeable). At this point I am completely against the bootstrap device concept.
3) I think an additional mechanism should be implemented that sets up a distribution authority for printed parts and that requires anyone who receives a set of printed parts to contribute 2 sets of printed parts within X months back to the authority. Specific details like if there is a cost or deposit or such can be worked out.

Thanks, I am ready to roll up my sleeves a little higher.

thanks goinreverse...all elucidating for me.

I figured there were issues with scaling that no one would talk about since no one seemed to even be conceiving of a much larger RepRap. In a way, this is disappointing to me as many practical items are larger than the current capabilities.

I agree on the self replication. it has not really worked out as planned (one person makles two machines, then those each make two more, etc.) because new people who want to build a machine are having a hard time getting parts, or they have to pay some fairly decent fee to get them.

If RepRap can be fitted with a CNC toolhead, why can't it just produce its own parts to replicate itself? Self replication does not mean it all has to be extruded. Can a CNC produce most of the RepRap parts that are now extruded?

Kevin

PS - Haven't even looked at the software yet, but i had read comments that echoed yours. I didn't make much sense to look at it till I felt like i was going to make a machine. I guess the teams answer to the software is "if you don't like it, fix it!"...and since this is open source I suppose that is not a totally wrong response.

A lot of the parts don't lend themselves well towards only CNC machining because of curvature/taper, overhang, orientation of holes, etc. I am not a CNC expert but I had a friend look at it and his assessment is that it would take several iterations to rework the plans. You could definitely get some portion though. Material is another issue, some tolerances are pretty tight so you would need pretty thick plastic or metal. Wood kinda-sorta can get you there but I could only find one person who had almost done it.

I am investigating molding and pouring parts but it seems to come back to cost and effort involved. I am beginning to strongly lean towards just building a mendel 'simile' out of standard parts from mcmaster (see my extruded aluminum thread). After that I think I would have a hard time seeing value in the self-replicating goal and it feels to me like it would be a fork. I am really trying to work within the current project, it just feels like my perspective is goes totally upstream against a lot of core community contributors. There positions aren't invalid, they have brought a lot more to this project to date than I have, we just don't see eye to eye.

If you want to play with the reprap host software you should probably start now as even getting it to run is a project. I agree with your "if you don't like it fix it..." take. I have the source and am starting to poke at it, to me replicatorG is just a much better base although a total rewrite seems inevitable. From a directions standpoint I would also say to start with the electronics/motors and building an extruder. First test out the electronics as there can be some issues with those (I got a bad stepper board from makerbot, they replaced it right away but I didn't find out until the last minute. Very anti-climatic). Building an extruder from scratch is hell, if the issue below gets resolved I would strongly suggest just getting the printed mendel single piece extruder carriage from protovantage for $60.

There is one major issue right now holding everything up for me which I am awaiting resolution of. That is driving a stepper motor for an extruder from the makerbot firmware. If I can resolve that (I am working on patch) then I will document my whole simplified mcwire build, finish building a mendel with it and see what feedback I get. Once I have the mendel I can make a decision about v3, forking, etc. I have some practical things I need the machine for and I want to see how well it performs.

I also have another extruder design in my pocket but it doesn't use printed parts so I'll try to go self-replicating route first.

Little hint, go over to makerbot and see if any have been sold in your are (they have a map). Try to find out who bought it and offer them some spare parts (they will need them) in trade for access to their printer. You can print all the parts over a weekend if you have someone print while you sleep.

---

repraplogphase.blogspot.com

did you try printing nopheads extruder or did you follow the RerRap version?

It looks nopheads extruder is somewhat more straightforward...

I've been following the RepRap community for a while. All of the issues that maitri982 brought up have been considered, but none of them have received enough actual experimentation time to be resolved. Many people who are eager to work on them don't have fully functioning machines to work on. So, I tend to believe these issues will be worked out by 100 people trying lots of different ideas in the next few years. The real problem is that these people don't have RepRaps!

So, I've dedicated myself to helping other people get RepRap parts and build functioning RepRaps, knowing that they will do the same for others and some of these people will improve the design. It seems like this is the most efficient path.

I've noticed a bit of momentum building on this point. A few people have replicated in the last few months, something I never saw in early 2009. I just delivered my first complete set of parts in the last week. I think, by this time next year, we will see a lot of people working on these problems and that will be more effective than Nophead's efforts alone (which have been great, but he is not as imaginative as 100 people).

Most people think of exponential growth as something that is incredibly fast, but it can be incredibly slow at the beginning.

So, if you want these problems solved faster, the best course or action is to buy a MakerBot (it seems to be the easiest design to get working), get it working and get RepRap parts to other people.

---

---

Darwin clone, Gen 2 electronics, Arduino Duemilanove w/ AtMega328, 5D Firmware, Pinchwheel extruder
[www.codeerrors.com]

I really can't help but feel like this project is a failure if the best solution that can be proposed is to buy someone elses commercial project to start bootstrapping this one... that just seems wrong-headed. If I am going to buy a makerbot I will just wait until their larger format device (with heated bed, magnetic encoder, etc) comes out which is probably only a couple months from now and forget about reprap altogether.

mccoyn; what you are doing sounds great and like what this project needs. What if you require everyone you give parts to to produce and return to you 2 sets? Furthermore if you are willing to send me a one-piece mendel extruder carriage I will commit to making a mold and to whatever extent that is successful dupe off as many as I possibly can.

maitri982: It takes an extruder to print an extruder... From scratch I mean making an extruder without an extruder. The hell part is mainly that the documentation does not exist, and there are mechanical, electronic, and firmware problems that all need to be overcome to be successful.

I have built a McWire and frankly am not satisfied with the performance. For large parts it print fairly decent even good. The smaller parts come out a pile of goo. I have tried many thing to make it work, but in the end it is McWire slow speed that does it in. Print times take too long. That being said I have changed my sights from building a Mendel to a design with the following attributes.

*The mechanical parts should be available at the local home center (Except bearings and motors) and inexpensive .
* NO printed parts should be needed anywhere in the build.
* Speed should be comparable to the MakerBot. (I'm trying to do this without belts and pulleys)
* Easy to make

Hopefully this design will be something that would be good enough that it could be your first and last 3d printer if you choose not to go Mendel.

I have a sketchup model for the X and Y and am 25% through the build. I started this Sunday. For now I am keeping the design closed so i can keep the nay sayers out of my head and keep focused on my goals. So I think the mechanical structure cost is coming close to $80.

---

B^2 : [replibot.blogspot.com]

~~ We Are The Factory ~~

Ironic that you have problems with small parts on McWire. I avoid the pile of goo problem on small parts by slowing down to McWire speeds, i.e. 4mm/s. Otherwise heat is being delivered too fast in the plastic.

Perhaps your nozzle radiates so much heat that it melts things under it when it hangs around?

I use a fan blowing on very small parts to fix this problem. You do need an insulated nozzle to prevent it being cooled though.

You can go much faster if you use acceleration.

---

[www.hydraraptor.blogspot.com]

I have always wondered, because the small parts are such a challenge with mcwires, why you guys just don't print multiple parts combined into 1 stl.

So instead of 1 part small part, you have 5 small parts, causing your head to move away for a while in order to let your parts not melt.

Just curious.

---

repraplogphase.blogspot.com

I think it's not JUST the McWire that causes the problem. I also am using the MK4 extruder from Makerbot that is DC gear motor driven. The extrusion control is not fine enough for the speed the McWire goes. There really no stretch to the filament (layer thickness = 0.65) at 200/255 for the PWM motor drive. Any slower and the motor stalls. I have played with oozbane and cooling but without too much success. I have done multiple part prints but sometimes that is hit or miss depending on the and if you need only one it's at least twice as long in an already slow print session.

Using fans has caused the extruder to stall and sometimes carve out. I thought about rigging up a cooling straw to concentrate the airflow but such experiments are really just a bandaid over the larger issue. The need for speed.

---

B^2 : [replibot.blogspot.com]

~~ We Are The Factory ~~

Have you considered refiting your McWire with belts drive? That would solve your speed issue. And it would also be easier than redesigning the mendel to use all off-the-shelf parts.

This is a McWire modification that I'm seriously looking at.

---

--
I'm building it with Baling Wire

Well you do have a speed issue in that your extruder can't match the bot, but I don't think going faster will help you make small pieces. I find exactly the opposite. To make small pieces I have to go slowly, so a stepper based extruder, or a DC motor with a shaft encoder is essential in my opinion.

---

[www.hydraraptor.blogspot.com]

Has anybody used the 5D firmware on the McWire? I think it would go many times faster with acceleration. There are much easier ways of doing acceleration than the 5D gcode method though. I do it in a few lines in my firmware, which are published in this forum somewhere.

---

[www.hydraraptor.blogspot.com]