rack and pinion versus belt?

I look forward to Sebastian's response about repraps potential growth in the US, I would love to be persuaded to change my mind on that. I've been working in open source projects for 20 years and my personal opinion is that currently reprap has little to none of the crucial philosophical and practical project management infrastructure I see in a linux kernel, ubuntu, apache, debian, etc. Others disagree and I am happy at this point to agree to disagree. As explanation here is why I have come to that conclusion:

*Is there a release schedule, no.
*Is there a target features release system, no.
*Is there community infrastructure to have a benevolent dictator accept or reject patches, no.
*Is there a community infrastructure to vote to accept or reject "patches", no
*Is there any articulation or documentation of the professional and financial conflicts of interest among principal participants, no
*A lot of things are under licenses which really can't be applied to them, such as applying the GPL to things which aren't source code or compiled source binaries. There are pretty compelling legal reasons why the GFDL and the CC license exist, they were created to address coverage of articles that are not computer programs.
*There are huge patent, trademark and copyright issues with most of the solutions being explored here.
*Are their any clearly articulated goals for the current release, not that I could find. --There IS the general "RepRap is a free desktop 3d printer capable of printing plastic objects. Since a large part of the RepRap is made of this same plastic the RepRap can be considered a self-replicating machine - one that anyone can build given time and materials." Which is fine but extremely ambiguous, even two 'official' contributors (nophead and sebastian) seem to have mutually exclusive interpretations.--
*Is there any tracking or response infrastructure for resolving bugs/glitches, no (I don't count a wiki page).
*Competing projects are drawing more attention and brain power (makerbot, cupcake, contraptor, 3dp, etc).
*I and the people I try to bring on to the project are immediately dead-ended on the bootstrapping issues, #1 most frequent advice is to send people to a competing project (cupcake). To me this is like ubuntu saying, ok to install our system please first install redhat...
*Is there a defined process (proposal guidelines, patches guidelines, anything,,,) for contributing other than "add it to the wiki", no

I could probably keep going on that list... I want to be clear that I am not just trying to nay-say the project and that I really wish I could find a way to contribute but I feel like I would have to brutally beat my head against every single one of those issue. My perspective seems to be going 180 from most other active contributors.

If reprap was a 0.01 alpha project I would be fine with those but officially this is supposed to the on 2.0 .

If Adrian is the 'Linus' of reprap I would expect him to post and say "you're an ass and take your energy elsewhere" or "I agree with 1,4,5 and disagree with 2,6,7, here how I think we can resolve those or propose a way to resolve those". I am fine with it either way, what I seem to get instead is "I don't agree but I can't propose another solution or I don't/do agree and/but the problem will solve itself eventually".

You are not alone, goinreverse. Many of your statements resonate with me too, and I agree that the RepRap project has quite a gap to fill to achieve the level of open source projects such as Open Office, several Linux distributions, etc.

The feeling I have is that everyone seems a bit afraid to take the leadership role because it might stifle innovation, and to some extent, I believe this is probably accurate. On the other hand, I also believe that growth in the project in terms of goals, focus, and expanding reach to a large audience suffers to an extent because of this deficiency.

Some of this may be because the current official release is just plain hard or expensive to do without an existing printer - the classic chicken/egg problem. There are a few really unappealing options to get around this: Sebastian offered buying a $1000 hobby milling machine, add a RepRap extruder, then invest another $600 for parts to build the printer, then sell the milling machine. This has the fundamental problem of turning away anyone with a $1000 budget. People are not going to spend close to $2000 for a $600 device. I suspect psychologically, they do not view the milling machine as a temporary cost, and it can be a lot of uneasiness, both for to order the device, then to modify it, then to sell it. Makerbot is a fine solution, and I think the most effective one to date; a commercial enterprise selling a complete kit. Unfortunately, they have some production and supply issues, and coverage to non-US sites is expensive and flaky at best, from my impression of posts on this board. The McWire solution in principle seems great, but there is still the lack of a definitive guide or kit there as well, as well as the extremely slow printing speed from the current advertised design.. and again, you are basically left to your own resources to collect and figure out how to acquire and build the device. My own project, RepolaRap, I think has a good potential (I am considering offering kits in a few months depending on the success or failure of it), but because I have traded off mechanical complexity for software complexity, the going is slow there too because I cannot leverage a large majority of the firmware written by others.

I suspect that once someone has the definitive 'kit', with production able to satisfy the "Ships within 2-3 weeks" for any and all orders, with well documented resources for DIY's that can order and purchase everything they need to create an official device *without* having an available printer within 6-8 weeks, and I think the project population will explode. Exponential growth, while very nice at the high end, has the extremely undesirable quality at the lower beginning end, effectively being basically flat up to a point compared to straight linear. I estimate the current doubling rate of second and beyond generation machines at about x2 every 4 months; this is great for about 6-7 years from now, but in the 2-3 year time frame, not so great; something has to happen to decrease the doubling time to sub-month, 2-3 weeks, where from the time you want a machine, you have a set of spare parts you printed yourself 3 weeks later; aggressive, but this would put a RepRap into anyones hands who wanted them within 3 or 4 years at most.

Sure seems like there is a lot more bitching and moaning than work going on around here. Seems an odd line of bullshit to find on a rack and pinion thread. Perhaps we should open a bitching and moaning forum just for goinginreverse and those of that kidney. Hell, you all could run a 24x7 pity party outlining how you feel that the Reprap project is just wrong in every possible way. I'll tell you what, though, I've been reading all the spleen venting and GADS its so tedious and small minded.

Edited 1 time(s). Last edit at 02/13/2010 12:07AM by Forrest Higgs.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

heh you tell 'em forrest

This is an open source project- things are done by those who want to do them, so there's no point bitching that something you want done isn't being done unless you're willing to do it yourself.

---

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

goinreverse Wrote:
-------------------------------------------------------
> I would love to be persuaded to change my mind on that.
>
Has it ever occurred to you that changing your mind just possibly isn't all that high a priority for people in Reprap?

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Seems an odd line of bullshit to find on a rack and pinion thread.

Well, we're really trying to sort out the RepRap [[Combinatorics Problem]]
[objects.reprap.org]

Hypothetical examples of stuff which is welcome and needed at
RepRap.org, and also issues at play:

The German-speaking Laser Expert is building a self-replicating Laser
Cutter and SLS-RepRap. The documentation is perfect. And in german.

The Russian-speaking New User is annoyed by the Russian langauage
forum because of russian-language pornographic spam that the
english-speaking forum moderator misses.

The Arabic-speaking Guy Who Wants to Translate RepRap into Arabic is
still waiting for me to make a arabic-RUG and initate a
arabic-Wikipage.

The Japanese-speaking Expert built a self-assembling set of
"[[RBS/Blocks]]". The documentation is perfect. And in japanese.

Joe User can't buy Mendel Parts.

New user-developers can't do this unless each step has been
documented. Experienced user-developers would rather research and
develop than document. Entrepreneur user-developers want to sell one
set of Mendel parts, or 50000 sets of electronics, filament, Mendel
vitamins and need RepRap to do documentation and support.

Jack Developer just forked his LaserCut RepStrap. And the RepRap
Development, Documention, and Object Library, and everyone with a
working RepRap now uses the Object Library to document their
post-mendel variants, cause they like to keep all their stuff in one
place. Also because people with working RepRaps don't go back to the
RepRap.org website unless there are interesting Library Object Files
there, unless they are making another RepRap.

A fellow in India makes molds for Mendels with his Mendel. Each mold
set casts 4 mendel plastic parts sets a day.

Please note there is also a "RepRap Service Work Problem", I would
prefer to spend no more than one hour a day doing forum and wiki work
to preserve personal time for play, figurative sculpture, building
Robots, designing new RepRaps and RepStraps, and working on getting my
machines existing, running, and self-replicating. To say nothing of
dreaming up and making new things for the RepRap [[Library]].

People who urge me to spend more than 1 hour a day doing RepRap
Service Work may at any time be urged in turn to either volunteer to
help with Service Work or to "go self-replicate".

Also, I need to scan and tidy up my concept drawing for [[LeCorb]], a
Tablesawable FlatPack Repstrap made from pegboard and BlodgeIt's
plastic cabinetmaking widgets. In other words, LeCorb is basically an
Eiffel with lots of RBS/Sheet and 24 tiny segments of RBS/Beam.
LeCorb will be very cheap and easy to make.

After that, I'll scan and tidy up my concept drawing for an Eiffel
PourStrap. Eiffel PourStraps are good SDM machines because they can
also support a spindle. And they should be fairly cheap, materials
wise, compared to generic benchtop mills and benchtop cnc routers.
Along with being (1) cripplingly expensive to ship, and (2) you can
make the molds with things that cut sheet. (3) molds are cheap to ship.

Oh, and I think we'll have formal bugtracking software in place at some point.

And I'm pretty sure Adrian would rather do research than work on more
than 2-3 of the above problems, until we've got the RepRap [[Cloning Vats]]
running, anyway. But I know he's in charge of figuring out all this
stuff. I don't know if he knows that. ... But this is all Adrian's
job to figure out. Really.

---

I'm glad there is so much energy in this forum thread from people
eager to volunteer to help and also people eager to go out there and
self-replicate.

Edited 2 time(s). Last edit at 02/13/2010 06:28AM by SebastienBailard.

I don't know if I would put it as strongly as Forrest, but I have to agree - while a good bit of what you're talking about, especially the problems that Sebastian raises, are important and deserving of discussion... they don't belong in a thread on designing rack and pinion drive systems. Figuring out what should and should not be covered in the wiki, what full releases should be covered, and how often they should be adjusted for changes to parts/electronics, etc is something that we should be talking about. In another thread.

Figuring out what should and should not be covered in the wiki,

I don't understand what "should not be covered" in the wiki, aside from handguns. I'd take wiki-deletionist arguments more seriously if they were presented by people uploading parts and text into the wiki. "should not be covered" really means: needs to be documented in a more careful manner if new user-developers are actually getting confused.

And frankly, if Person A is telling Person B that "your stuff doesn't belong in the wiki", then I'm going to get angry with Person A, because it's my job to help Person B get their RepStrap, RepRap, or new extruder, laser, etc up there. Unless we think that McWire and Mendel are perfect, and we are done.

If Person A feels that strongly about it, they should set up their own damn wiki and fiercely protect from icky foreign content and machines. Or go and keep everything in a lab notebook. Or, more likely, blog the hell out of it and upload every damn thing they make to thingiverse, while not uploading any damn thing they make to RepRap, as may or may not be happening right now.
what full releases should be covered,

Presenting and maintaining a "RepRap Stack" of modules that work together is a crucial RepRap developer responsibility. As is documentation. Mind you, I might be working on an [[Eiffel]], [[BitBanger]], Extruder-and-Spindle aka[[SDM]], and [[EMC]] Stack 50% of the time.

This is much easier with the Linux kernel, gnu tool chain, directory structure, X window software, KDE Desktop, and Firefox browser that I'm using right now, aka my "Browser Stack".

But helping maintain the wiki that we use to sort this all out is the responsibility of all of us.

Deleting everything but the current working Sat Feb 13 22:13:21 EST 2010 RepRap Stack is silly.

and how often they should be adjusted for changes to parts/electronics, etc

People who don't want to work on new parts and electronics shouldn't tell the people who do, aka "RepRap Developers" to stop doing research and (hopefully) documentation. But they do have the right to say "this isn't documented, except on your personal blog".

It would be a lot less work for us old hands if we just completely and utterly abandoned each old version in order to just blog our latest research. It is, perhaps, not the most ethical way to go about RepRapping, though though.

is something that we should be talking about. In another thread.

Yup. Agreed.

Forrest, what's new with you and Rack and Pinion?

Also, have you tried testing it to destruction with an automated test rig, say, moving a 5kg weight back and forth 10,000 times?

SebastienBailard Wrote:
-------------------------------------------------------
>
>
> Forrest, what's new with you and Rack and Pinion?
>
I got Herringbone Rack and Pinion working and got a mounting beam that the rack can integrate into designed and tested to an extent.
>
> Also, have you tried testing it to destruction
> with an automated test rig, say, moving a 5kg
> weight back and forth 10,000 times?
>
I have something like that in mind. I found that Pololu have Allegro A4983 breakout boards which get you microstepping down to 1/16x. I took delivery on those a few days ago, but have been ill this week so haven't had a chance to get them running with my I2C MCU. It doesn't look like any big deal to hook up and test. The only rather scary part is that the A4983 chip is supposed to handle 2 amps. The problem with that is that it is about as big as a sequin, viz, a 4-5 mm square, surface mounted chip with connectors you need a microscope to see. It's going to be tricky putting a heat sink on that puppy. Once I get the driver board working with the NEMA-17 setting up a test rig to see how accurate and repeatable the positioning is and how it wears will be happening.

Fortunately, that breakout board lets me directly drive the rack and pinion instead of going through reduction gears. That makes it both simpler and probably more reliable.

Right this minute, I'm recommissioning Slice and Dice, the pixel-based STL slicing approach that Adrian used to hate and then decided that he loved a month or so ago for the Rapman printer that I own. I ran into problems making gcode out of the light, openwork structures that I have been working with lately when I use Skeinforge. Doing that sort of thing really tests Skeinforge's performance envelope, so I'm reviving my own code. That lets me dive into the code rather than having to ask other, busy people when things go wrong for me.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Forrest, I posted recently about how I heatsinked my A4983s. method works really well! Note that the Vbb (Vmotor) capacitors and the current pot are taller than the chip, and some of the other smt parts are very close to the same height as the chip so a shim or riser is highly recommended

---

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

rhmorrison Wrote:
I particularly like the use of
> plain paper, adhesive and a cutter that creates
> wood-like objects at a very low cost - the main
> cost being the special adhesive required. Others
> use a roll of plastic, adhesive and a cutter which
> makes even stronger parts but at a significantly
> higher cost (plastic vs plain 80/160 g paper).

I agree. I admire the mcor technologies machine, but it's still pricing itself out of the market, at between 20K and 25K euro. I've been daydreaming about a Mendel bed without Z axis, just pen up and down like a plotter, with say five y axes, each carrying, instead of knives, dentists' drill air motors that have 1.5mm burrs that rout slots in a sheet of paper instead of cutting with a knife. With multiple Y axes, the x travel only has to be 60mm with the y axes varying their rate and lifting and dropping. Dentists' drills have a free running speed of up to 500,000 rpm and I don't think that sharp burrs slotting a sheet of paper will slow them much. They're compact, so lifting and dropping, they only need about 100mm of headroom, so the Mendel beds could be stacked to run multiple sheets simultaneously. Adrian's experiments have shown that at least four repraps can be run from USB on a laptop, so coordinating the operations would be slightly more complex than that, but less so than, say, a five axis milling centre. Each sheet of paper is a .1mm layer, but if you're building a small model, say a 50mm square footprint, each sheet could give twenty or more layers, 2mm thickness, with the same sort of speed as a reasonably quick printer. If the burrs cut two holes for register in each slice, they could be manually stripped and stacked on posts quite quickly, then drenched in cyano and sprayed with accelerator to set them instantly. I think that surface finish would be more consistent than extruded plastic. Prototyping value only, but truly rapid. Multiple beds could be stacked vertically to increase speed by multiples. Cheap and fast. Just a thought.

Triffid_Hunter Wrote:
-------------------------------------------------------
> Forrest, I posted recently about how I heatsinked
> my A4983s. method works really well! Note that the
> Vbb (Vmotor) capacitors and the current pot are
> taller than the chip, and some of the other smt
> parts are very close to the same height as the
> chip so a shim or riser is highly recommended

Triffid: I'm assuming that you are talking about this...

[picasaweb.google.com]

where you park a fan over the breakout boards?

If so, that's good. I salvaged a fan a few months ago that is just the right size from a PS2 that my son wrecked years ago.

Edited 1 time(s). Last edit at 02/14/2010 11:34AM by Forrest Higgs.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

yep that's the one. Note that I cut channels into the heatsink face for the Vbb capacitors to sit in without being crushed or shorting to the 'sink

---

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

all i can say sitting back and watching this dialog is that if certain people here are representative of the core RepRap team than many people will not want much to do with it.

There should be a certain sense of decorum between people on such projects. My experience is that friction of ideas fosters the creative process, but only amongst mature individuals. if it degrades to insults or calling something "bitching", then this means that the core team is really not in favor of progress as much as protecting their "baby" from constructive criticism.

This is too bad because I think that goinreverse wants very much for this project to be successful, and his points are all valid. I have never laid eyes on a RepRap in person, much less made one, but I know a thing or two about running projects and creating an environment of innovation...and frankly this project is a mess. This is not to blame anyone, because it is no ones fault. It is due to the fact that doing 3D printing is so new and there are so many tangents that can and will be explored by all the smart people here, that it is hard to manage.

there are probably 4-5 core skill sets involved in 3D printing including electronics, materials sciences. software, and mechanical design, and it is clear that each of these areas has extreme innovation going on separately. So managing this without a strong coordinated effort is VERY hard, if not impossible.

if I had the means I would organize a team of 5-7 key people with expertise in key areas who have worked on this project to work together for couple of years to create the machine we all dream about. Unfortunately I don't have the time\means to do it today...but I think this is the only way this can all be effectively managed. it is a full time job for a dedicated and coordinated team.

Perhaps it takes commercial interest to look at all the things being developed by various parties and to then integrate that into a great design, ala MakerBot, and that is fine with me. My hope is that a 3D printing machine is created that we can all have that can produce many everyday devices such that the cost of the device is less than the products it will produce. For me I could care less whether it self replicates...cool if it can, but not really essential IMO. If it is cheap enough then self replication really is irrelevant.

Kevin

All this may be true, Kevin. Please have a go at this page:
[objects.reprap.org]
or
[objects.reprap.org]

Also note that we're running into the fact that RepRap is a constellation of different technologies which we mix and match when we build a particular instantiation. Like bolt a ceramic extruder onto a bridgeport mill.
[objects.reprap.org]

---

Now that we're done with that derail, is your machine going to use belt, or rack?

I've been thinking about extended bed machines which use belt or rack to move 4 feet of bed back and forth, but on a mendel with a footprint of 1-2 feet. We might want outriggers to support that. Or lay rails on our basement or workbench floor.

I already had a go at that page. I added the verbiage around the t-slot RepRap version...

I already had a go at that page. I added the verbiage around the t-slot RepRap version...

I just had a go at it myself.

It occurs to me that t-slot seems purpose made for us to slip RepRap-made Herringbone Rack into it ...

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

Kevin,

You can take offense and get upset and preach and complain and threaten till the cows come home. It will make no difference whatsoever to the Reprap project. With Reprap, your worth and reputation are dependent on what you do, not what you say and especially not what you demand. If you look at the people who are highly respected you'll find that they are the guys like, inter alia, Vik and Nophead and Zach. They put their passion and energy into making Reprap happen and have been for years. That's how you gain stature in an open source project, pure sweat equity. You don't just walk in off the street and start making pronouncements and expect to get paid much attention to. You solve a problem that advances the state of the art and people start noticing you fast.

I'm probably the "certain people" you so coyly refer to. I'm a big boy. You can call me by name. I don't mind. I call 'em like I see 'em.

"I know a thing or two about running projects and creating an environment of innovation..."

So do I. I've been a senior scientist, team leader, project manager and the like at universities, gov't labs and corporations for onwards of thirty years now. I've worked research and development teams in strongly hierarchical management structures, flat management models and unfunded quasi anarchies like Reprap. I've been damned good at it.

Problem is that in all of my non-Reprap experience, I could motivate team members with money, respect and advancement. You suggest that...

"..if I had the means I would organize a team of 5-7 key people with expertise in key areas who have worked on this project to work together for couple of years to create the machine we all dream about. Unfortunately I don't have the time\means to do it today...BUT I THINK THIS IS THE ONLY WAY THIS CAN ALL BE EFFECTIVELY MANAGED. IT IS A FULL TIME JOB FOR A DEDICATED AND COORDINATED TEAM {emphasis mine}."

That's just great. Basically, what you've said that is that the only way to make Reprap succeed is with a conventional R&D team. You even have the chutzpah to say that...

"I have never laid eyes on a RepRap in person, much less made one"

Conventional R&D teams got 3D printers down to a cost of about $50,000 in 2005 when we began with Reprap. It took conventional R&D teams about fifteen years to get the state of the art to that point.

It cost about $30/cubic inch to print things with one of those machines. In the five years since then our anarchistic, uncoordinated, and in your opinion doomed to failure effort has reduced the cost of 3D printers to well under $1000 and the cost of printing to about 15 cents/cubic inch. In the same time, conventional R&D has managed to reduce the cost of printers to about $15,000. I can't but wonder whether they would have moved nearly so fast if they hadn't had the wolves of Reprap snapping at their heels while passing them.

If that's failure, I don't know about you, but I could do with a lot more of it.

People are here because they want to be here. They aren't paid. They're doing what they're doing for the pure love of it. The core team is here because early on they were captivated by Adrian Bowyer's vision of a future with virally diffusing technology where the scaling of manufacturing endeavours drops by several magnitudes. They want ordinary people to be able to design and make things. They don't believe that the future belongs to a priesthood of "DEDICATED AND COORDINATED TEAM(s)".

Even as hopelessly retro a publication as Wired has finally caught on. Their lead story this month is "The New Industrial Revolution". LOL! It took them five years to catch on, but they finally did.

It's the 21st century. The aptly self-named goinginreverse is hopelessly committed to the 19th-20th century notion that mass production is the only way to do things. You're similarly committed to 20th century R&D model for technical innovation. Both notions are living on borrowed time.

These days it's OODA loops, Swarming and Crowd Sourcing. Like it or not, that's what is happening in Reprap. We're solidly 21st century.

Forrest

Edited 7 time(s). Last edit at 02/15/2010 09:03AM by Forrest Higgs.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

I agree with you Forrest.

Warning - it is after 3 AM where I live. I will disavow anything
I've typed here until I've had a chance to sleep on it. As opposed to
usual, when I just blame the cat for things I wish I hadn't
written. Anyway, this is an utter ramble.

---

Documentation needs to improve.

But we do need bugtracking, and I assert that we need to build social
networking into the wiki so that people get feedback (praise,
favoriting) when they upload to it. Right now the wiki, she is an
emotional black hole.*

We've got an extruded aluminum RepStrap candidate:
[objects.reprap.org]

But sam's not going to want to upload without social networking-type
"Nice job", and "800 people favorited this post", etc.

I agree that "1 team" is silly. But regarding "COORDINATED TEAM". We
need COORDINATED TEAMs. 1024 of them. Like the Heated Bed team, the
SpoolHead team, the skunkworked SDM team, and the new extruded
aluminum team.

Being able to offer email will help. So will having the wiki do all
the user comms for some teams via shoutboxes.

---

Distribution needs to improve. e.g. a "tender" system for Mendels,
etc.

---

I think that debian does do dedicated subteams very well. (It's been a
while since I lurked over there.)

Although I think we mostly want swarms.

The linux and linux kernel analogy some people invoke, "adrian ?=
linus" is an utterly wrong analogy. Because we have stacks, or machine
configurations which are RepRaps without using any element from
mendel. But they may still be utterly cool.

(I think the extruder is the kernel. Unless ... it's the slicer?)

I think Adrian is the Debian Project, maybe?

---

quasi-self-replication can be a more realistic and desirable feature
for pre-nano RepRaps. Like a RepRappable lathe, and a dedicated
extruded aluminum pick and place, and so on, all of them open source
and in the wiki, capable of donating features and parts to post-mendel
candidates, and also obeying the "Everything in One Place" scenario.
[objects.reprap.org]

---

*On generic wikipedia, that is done via the "discussion" tab. But that
separated wikipedia into two populations. And that won't work for
RepRap.

And most importantly, Rack and Pinion needs to improve.

Forrest, what's the longest summed Rack you've made? And is it easy to bolt to pegboard?

SebastienBailard Wrote:
-------------------------------------------------------
> And most importantly, Rack and Pinion needs to
> improve.
>
> Forrest, what's the longest summed Rack you've
> made?

I printed one that I think was 220 mm long. That is the diagonal length of my Rapman print table less a little bit for width and raft.

> And is it easy to bolt to pegboard?

The script to build the rack is a hairy little piece of java. I extended the rack with flanges on either side of the rack in the script.

As to tacking it to anything, you need to design a connector that mates with the rack. I designed a column system that slots pieces of rack...

[technocraticanarchist.blogspot.com]

... for my Delta 'bot pick and place project.

Personally, I wouldn't screw it to pegboard. Dimensional stability of pegboard isn't up to much. If it were me, I'd glue it to float glass. That stuff is cheap and for practical purposes optically flat.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

My take on goinreverses comments, as I was dinged on "sympathising" with goinreverse's sentiments:

>> *Is there a release schedule, no.
>> *Is there a target features release system, no.

A release schedule looking forward 1 or 2 years would be nice, but not necessary. What this provides to new or semi-new people is an idea for whether waiting for higher availability and better printing capability would be worth it, vs. simply jumping into the fray now, based on that individual's level of frustration tolerance or ability. When will the next version of official reprap (Darwin, Mendel, ...) machine be in 'beta'? I realize a lot of this cannot be predicted, and commiting to meeting a deadline challenging when rewards consist entirely of non-financial payment.

>> *Is there community infrastructure to have a benevolent dictator accept or reject patches, no.
>> *Is there a community infrastructure to vote to accept or reject "patches", no

The advantage to this becomes one of consistency. If you have one person or tight group acting as gateway to allow new features or fork merges back into the official release, you get that person's or group's vision for what RepRap represents. Having the swarm mentality, you get a bit more churn in the mix, but vastly more innovation. Advantages either way; disagreements about this as important seem to stem simply to which priority takes precedence: ease and simplicity vs. innovation and power.

>> *Is there any articulation or documentation of the professional and financial conflicts of interest among principal participants, no
>> *A lot of things are under licenses which really can't be applied to them, such as applying the GPL to things which aren't source code or compiled source binaries. There are pretty compelling legal reasons why the GFDL and the CC license exist, they were created to address coverage of articles that are not computer programs.
>> *There are huge patent, trademark and copyright issues with most of the solutions being explored here.

I feel this also is a large sore point. Given the lack of information about what "hardware" GPL means, especially considering that the FSF and Richard Stallman have stated that "GPL applied to hardware" is nonsense at this point in time. I have yet to see an answer that discusses the rights granted to the non-software components within the RepRap heirarchy. I'm less interested in the financial conflicts of interest; as it stands, making the hardware rights made known should be adequate for either commercial or non-commercial use of *ALL* the components (software and hardware) of "RepRap".

>> *Are their any clearly articulated goals for the current release, not that I could find. --There IS the general "RepRap is a free desktop 3d printer capable of printing plastic objects. Since a large part of the RepRap is made of this same plastic the RepRap can be considered a self-replicating machine - one that anyone can build given time and materials." Which is fine but extremely ambiguous, even two 'official' contributors (nophead and sebastian) seem to have mutually exclusive interpretations.--

This is one point on which I disagree. The goals may be somewhat open ended, but the reprap website home page fairly clearly states the goal of producing a low cost self-reproducing desktop manufacturing machine. To quote: "The primary goal of the RepRap project is to create and to give away a makes-useful-stuff machine that, among other things, allows its owner cheaply and easily to make another such machine for someone else."

>> *Is there any tracking or response infrastructure for resolving bugs/glitches, no (I don't count a wiki page).

This would be nice; Sebastian, how hard would it be to add one of the several bug tracking systems into the reprap site with home page "http://bugs.reprap.org" or something like that? Is there already something like this set up?

>> *Competing projects are drawing more attention and brain power (makerbot, cupcake, contraptor, 3dp, etc).

I feel this is a mixed bag. In one sense, it's good because it gets greater visibility to the idea that cheap 3D is fun and inexpensive. However, I've also seen a bit of press on RepRap too; I think it's all about having a 'front-man' who's sole contribution is in marketing and promotion; this can be tricky to find when the rewards are non-financial.

>> *I and the people I try to bring on to the project are immediately dead-ended on the bootstrapping issues, #1 most frequent advice is to send people to a competing project (cupcake). To me this is like ubuntu saying, ok to install our system please first install redhat...

This is key. I wanted an official "Mendel" version of a RepRap machine myself. The current model is near impossible for a semi-novice, semi-tech capible person to build without access to a pre-existing 3D printer for nominal cost above one with RP available. If I had available a 3D printer, I suspect I would have had a Mendel up and running for about US$800. As it is, I have a few wooden parts that I've been able to make, and a $600 pile of parts. I do feel this may be a short term problem, but it is one very keenly felt at the present; it is not a valid option, in my opinion, to double or triple the funds needed up front, even if some of that cost can be recouped later (E.G, the "buy a millstrap", or "buy a MakerBot", or "buy a RapMan".) I want a Mendel. If I wanted a millstrap or a makerbot or a rapman, I'd probably be a much happier person right now.

In my vision, a RepStrap version of the official line should be almost as easy to make without a 3D printer as the second generation version; this has been my goal with RepolaRap; So far, I still think I can attain this goal; with $100 in tools, $100 materials, and $200 in electronics, I've made a working and fast X+Y axis structure. I estimate the Z axis structure and extruder will be another $400 at most, so overall cost competes quite well with Mendel. The biggest problems facing me have been software/firmware related; I've probably spent 3x to 5x more effort getting the software going than the hardware. This is a very good thing, because leveraging that effort to copy a bit of software is vastly more simple and easy than copying a bit of hardware, if you assume you have a computer and internet, but no 3D printer. I realize this still doesn't get me a Mendel, but as I'm using that $600 part bin in the corner, the extra material costs actually come in at less than $100, so, if I or someone else who's made a RepolaRap wants, the can use it to print out all the Mendal parts, then scavange all the motors and electronics off it to make the Mendel. Final investment over and above the cost of Mendel: $100. Final time investment assuming you have the software, a weekend or three. I believe others will find this appealing; I certainly would have. (I am aware of McWire; from what I've heard and seen, it is very slow.)

>> *Is there a defined process (proposal guidelines, patches guidelines, anything,,,) for contributing other than "add it to the wiki", no

Sebastian, is there a wiki page for how to add images, content, source code, board designs, etc. to the wiki, and recommendations to maintain some consistency for how content is managed and layed out? If not, it seems it should be pretty straightforward to create one based on the steps you usually use when you create a page for others, and add a "How to create content for this wiki" as one of the first items in a FAQ and menu item or something like that? And by this, I don't mean a link to a template, but rather, use case flow oriented: Step 1, click here and type the name. Step 2, add a summary section. Step 3, click the "add image" button to add images, (etc.)

Forest,

Have you released your rack and pinion script yet? I'd like to see if I can print it and show off the parts to some people I'll see this weekend.

---

---

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

Personally, I wouldn't screw it to pegboard. Dimensional stability of pegboard isn't up to much. If it were me, I'd glue it to float glass. That stuff is cheap and for practical purposes optically flat.

How about optically flat polymer concrete with 1" regularly spaced holes?
[objects.reprap.org]

But prototyping on masonite (or steel) pegboard. Or erector set. Or anything else with regularly spaced throughholes.

---

BeagleFury, I'll get back to your queries when I've got a moment. Or you can bring them up in the Library Admin Forum.

So getting back to the whole printed rack and pinion idea. The problem with rack and pinion is that we have to move the relatively massy y axis stepper motor with the X axis. More mass means more inertial forces on the rack(if we're aiming for a print time of less than 10 days we're going to have to go fast!) Anyone have an estimate of how much mass the x axis could have?(the mass of darwin's x-axis is fine) I'd really like to do some finite element analysis on the rack and pinion to see if the deformation under inertial loading is significant enough to make printing inaccurate....

Also I'm worried that the printed rack and pinion may not be accurate enough to make itself. Though if worst comes to worst it might be possible to 'grind' racks and pinions.

I'm not sure that it necessarily follows from having a rack and pinion setup that your x axis needs to sling your y axis around. I would start from the mendel setup for this for that very reason, in fact. One axis of all-thread for z still seems to make sense to me because threaded rod is so good at station keeping - no need to keep the stepper energized 100% in order to stay on the layer. That's definitely up for debate, though. Even with rack and pinion on all axes, I think it makes the most sense to do a mendel style arrangement with z and y combined over an independent x axis. That way the heaviest thing getting moved around is the extruder itself. If you mount the steppers at the center of each axis so they remain stationary in their axis, there shouldn't be any real problems.

The problem with the z and y combined axis and independent x axis is that when you're printing really big things you have to move all the material that was printed. The mass of a full build volume is probably more than the mass of the x-axis and y-axis combined. Now when you want to print a bunch of parts in one go this becomes a problem....

It's also hard to do a conveyor belt part ejector on a moving bed.

I see no reason for using rack and pinion on the z-axis, unless we come up with some weird locking mechanism.

Gene Hacker Wrote:
-------------------------------------------------------
> So getting back to the whole printed rack and
> pinion idea. The problem with rack and pinion is
> that we have to move the relatively massy y axis
> stepper motor with the X axis.

So? You already do that with both Darwin and Mendel. It is the same with Darwin and with Mendel the z-axis raises and lowers the x-axis..

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

mccoyn Wrote:
-------------------------------------------------------
> Forest,
>
> Have you released your rack and pinion script yet?
> I'd like to see if I can print it and show off
> the parts to some people I'll see this weekend.


You are, of course, welcome to them. The rack script is clean. The pinion script has a glitch which causes a flaw in the gear which doesn't affect the mesh surface. I haven't had time to chase down how that flaw is happening, but discovered that I could repair it with the free Netfabb download.

Alternatively, I could simply generate a rack and pinion, clean up the pinion and ship those to you, if that is more appealing. I'm attaching the scripts to this forum message. If you want the other, send me a private message. You will probably want bigger ones than I have been printing with my 0.3 mm extruder head.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Forrest,

First off, please forgive me for not following your work on printable rack and pinion drive systems more closely. What software is needed to run your rack and gear scripts?

Have you made any measurement of the backlash, due to the rack/pinion themselves? (That is, the back/forth motion possible when the pinion is not permitted to rotate. I realize that this will also be a function of the slop in the linear slideways.) Backlash becomes very important for milling, something I think will become important for speeding up the building of large parts. (Extrude faster, using a larger extrusion orifice, and then mill off the blobs that ooze outside the desired profile, once per layer.)

My repStrap currently uses leadscrews, but your work really has me convinced that printable rack/pinion should replace bought threaded driverods for the machines I attempt to print. I also find it notably cool that the Herringbone gear pattern can be extruded, but would be quite difficult to make using a conventional mill -- or even a hobbing machine (at least the ones I've actually seen outside of books.)

---

Larry Pfeffer,

My blog about building repstrap Cerberus:
[repstrap-cerberus.blogspot.com]