Linear Motion Precision

Question: what is the minimum degree of control (in metric) that is needed in the X\Y plane to get decent resolution on FDM prints?

What would be optimal degree of control which is the maximum degree of control where beyond which improvements in control would not make a difference given current extrusion methods?

I assume someone has had to determine these numbers...at least I hope so...

Kevin

Those are subjective questions - how much resolution do you need? That will depend on what you want to do.

Here's what's worked for me - 0.1 mm resolution, using a 0.5 mm filament, gets me enough resolution to make more repraps, including simple gear trains like this:
[objects.reprap.org]

Speed is the other limiting factor though - I have been running at 16 mm/s, and I find that it's pretty slow. It takes a week of extruding time to extrude all the parts for a Mendel, which is a long time. By increasing my head speed to 50 mm/s, and decreasing the resolution, I hope to build parts a lot faster, but it's still very experimental.

Your mileage may vary, of course.

Wade

Since the filament we use typically fluctuates in diameter by about 10% there is not much point if having the axis resolution better than the filament diameter over 10. In practice I doubt you would see any improvement with resolutions less than 0.1mm.

Microstepping makes the movement smoother and reduces noise and vibration though, so often the resolution is higher for that reason.

---

[www.hydraraptor.blogspot.com]

Since the filament we use typically fluctuates in diameter by about 10% there is not much point if having the axis resolution better than the filament diameter over 10. In practice I doubt you would see any improvement with resolutions less than 0.1mm.

< 0.1 mm resolution is useful for the folk doing powder printing or surface deposition manufacturing.

thanks all...it seems the consensus is .1 mm or better. I will use this as a guideline for frame\linear motion\electronics and motors design\choices...

Kevin

There is a big difference between practical accuracy and theoretical accuracy of the underlying frame and traversal movements.

I think the real target needs to be .01" (0.04mm) with components selected on that basis.

Extruders are seeing dramatic improvements right now with the work that everyone is poking at (through feed, paxtruder, bowden, pressure management, makerbot MK5, etc) going down to .1mm nozzles I think. Linear motion will need to keep up.

Also, at some point, you run into "feature quibbling", because everyone's ideal RepRap has a different feature set.

The trick is to ignore that, and just keep researching, since you probably know what part of RepRap features and performance envelope you want to work on.

Oh, and don't forget to document on the wiki.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

SebastienBailard Wrote:
-------------------------------------------------------
> Since the filament we use typically fluctuates in
> diameter by about 10% there is not much point if
> having the axis resolution better than the
> filament diameter over 10. In practice I doubt you
> would see any improvement with resolutions less
> than 0.1mm.
>
> < 0.1 mm resolution is useful for the folk doing
> powder printing or surface deposition
> manufacturing.

... or <0.01mm for some of my applications

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

goinreverse,

have you been able to get that level of accuracy with t-slot linear motion? what steppers and electronics are you going with?

X and Y I have been right on the cusp and definitely under .1mm, I am side tracked playing with extruder tech right now so I have something printing wise that can use high accuracy motion. Two primary problems on that, nozzles and driver motor. I am waiting on a higher torque motor to see what the highest pressure that can be achieved with the feed through is. Turning by hand crank I can get the filament to snap rather than strip. It would be awesome if someone had the math on the pressure to push ABS at X temperature through Y nozzle (like .05", .01" and .005"). Alternatively I had a brainstorm on extruding UV resin which I need to experiment with. It is a liquid about the consistency of simple syrup (a bit thicker than water) and when exposed to UV light hardens. If that works I have something between in inkjet and FDM that I think would be a pretty neat solution with .01" accuracy.

I am using 0.9 nema 17 steppers with a rack and pinion drive with the pololu board (though I think the standard reprap 2.3 boards could do .01" accuracy without a problem). Not production ready just yet though. I am currently using nylon rack and pinion and I am waiting for the steel replacements, There is a really 'mushiness' in the nylon rack and before I spend of lot of time trying to make that rigid I wanted to see if the steel solves it.

Waiting on some more parts from 8020 to see if using rods and bearing sleeves will be as good but cheaper than the 8020 linear motion pads. Also need to do more experimentation with creative pad configurations. My garage looks like an 8020 demo van at this point.

On Z I am waiting on some parts to come in to try some solutions, the biggest issue is how to symmetrically drive both sides of the gantry. With off the shelf parts it is almost cheaper to just use 2 steppers, one for each side then to get the pulleys and belt to sync them. I would also like to see if I can do rack there to instead of screw just because it would be consistent. The rack is great because it fits directly into the t-slot and recesses slightly so that the gear has some edges to keep alignment.

I am 65% through spinning up a forked project (domain, website, settings up forums, etc) and as soon as I have my final extruder solution, a basic frame option (waiting on 8020 to create a kit SKU) and a solution to electronics I will post those there. The electronics is the biggest dead end right now. From what I can tell makerbot will move to Gen4 pretty soon (unverified but I heard recently as a conference that they are waiting on their first batch of completed gen4 boards from china). No one I could find commercially wants to commit to building more Gen3 until that is clear. Makerbots march batch is already sold out and april is close to selling out.

I am 65% through spinning up a forked project.

Why? Does RepRap seem insufficiently forked?

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

I am confused. Why are you you using 8020 linear motion pads if you are using rack and pinion? I would have assumed this was a replacement for 8020's linear motion solution.

I also assume you are not getting your linear motion stuff from 8020...did you find a good supplier?

anyway...it sounds like you have made some good progress. it will be interesting to hear what your final t-slot and rack and pinion setup cost sans electronics and steppers.

Regarding the sprayed resin, this is exactly what Objet uses and from what I understand their technology is the best out there for plastics part printing. this makes sense since you can get great precision with inkjet type technology. I am surprised actually that no one here has tried this yet. In fact it was high on my list for our new group here to consider. I know there are several brands of photopolymers, but don't know the best one or what would be needed to use them. can you use a standard inket type printhead for example?

also they are able to spray out support material that can be washed away later so there is not really a limit on printed structure.

Kevin

Kevin,

Here's the state of RepRap inkjet research.
[objects.reprap.org]

We've got some separate uv and photopolymer research buried in the forum:
[dev.forums.reprap.org]
much of it original research by fernando, (user name: spota)
[dev.forums.reprap.org]
I think we have zero photopolymer stuff in the wiki.
[objects.reprap.org]
(new wiki page with notes to forum threads)

Also, regarding Inkjets, remember that there are Thermal and Piezo, and Thermal might cook your plastic.

I'll ping spota, he's been away from his keyboard for a bit.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

@Sebastian As I already posted in previous threads I just think the reprap project is waning, it is turning away new contributors with inaccessibility and attitude at a profound rate. Further I fundamentally disagree on just about everything with the founders and current core team members. Certainly it has been forked before but I don't really think it has been credibly forked except by makerbot. Rather than continuing to just bang heads with a lot of folks here I would rather put together a like minded group and start fresh.

The 8020 linear motion is just the bearings and bearing carriages. I have been using their stuff but I think that with a little work I can do things more cheaply with rods/sleeve bearings, still experimenting. The 8020 linear motion doesn't actually move the planes, for that the rack and pinion is still the power transfer mechanism to actually move the planes.

You can't buy directly from 8020, they use distributors. Flomoco.com seems to have good pricing, instant quoting and you can use their 8020 sketchup library to auto-create BOMS and orders.

I did not realize the objet worked that way... interesting.

As I already posted in previous threads I just think the reprap project is waning, it is turning away new contributors with inaccessibility and attitude at a profound rate.

I'm working on the inaccessibility on a number of fronts - email me for details, but trust me, you don't want them. I can't do as much about the attitude, but I think it was possible there was a bit of conflict you presented a "perfect prescription for improvement" without trying to engage the community culture here, which resulted in a bit of impassioned discussion.

There's a large body of RepRap "lore" and community embodied in the forum and in people's heads - I think if you'd lay out a clear road map of each idea on a wiki page/new forum post, with a napkin sketch of your RepStrap/RepRap and then do an RFC, you may have more luck.

Any one large block idea gets tons of flack in this forum, due to the slightly intractable "Your perfect RepRap/RepStrap is not my perfect RepRap/RepStrap" phenomenon.

Again, if you chunk your ideas into: "my electronics proposal", "my RepStrap/RepRap proposal", etc, in the relevant forums, you'll find a lot more enthusiasm and sincere discussion then your ideas are one large block. Also, I'll run interference when twits raise the "but that's not my perfect RepRap feature set" nattering-point.

Rather than continuing to just bang heads with a lot of folks here I would rather put together a like minded group and start fresh.

If you think the best way to contribute to RepRap is to not actually contribute to the RepRap website/community, I don't know that I agree with you, but I've learned that there is not much point in discussing this particular matter.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

@Sebastian: "I'm working on the inaccessibility on a number of fronts - email me for details, but trust me, you don't want them." Of all the crucial things to publicly discuss in the forums that sure seems like the #1 priority to me.

"So - given that I am the leader - I regard one of my jobs as being to prevent leadership..." - Adrian Bower

Perfectly fine for him to take that position, 100% confirmation for me that this is not a project I am going to be able to work with. I can't think of a single mainstream open source project without a strong leader or group of leaders at its center. Think linux, redhat, apache, ubuntu, bind, mysql, postgresql, gnome, kde, firefox, wordpress, bsd, php, python, ruby, perl, arduino, all of them have strong consistent leadership.

I think widely available 3D printing represents a crucial and desperately needed economic tool for the public good. I think with the right group and the right project with the right 3D printer, it can make it into 10,000 hands in the next 24 months,

This post here: [dev.forums.reprap.org] was my attempt to provide constructive criticism. I fully admit it is not an ideal way to go about it but the points I identified are not nit pick items, they are the fundamental foundation of virtually all open source and certainly of the successful mainstream project I identified. I received a number of private contacts of contributors/potential contributors who agreed but didn't want to post in the forums because they thought it would not be constructive or just didn't want to be heckled, that to me made the need for a fork clear.

One of the most critical parts of any project is deciding what doesn't belong in it, that takes some form of leadership be it a group or an individual.

Forest can now make his obligatory post about how this bullshit doesn't belong in a thread about linear precision. QED YMMV

@Sebastian: "I'm working on the inaccessibility on a number of fronts - email me for details, but trust me, you don't want them." Of all the crucial things to publicly discuss in the forums that sure seems like the #1 priority to me.

Nope. It's taken me a little while to figure this out, but I finally realized passive-agressive ranting on my part is not nearly as effective as reasonable-type server functionality improvements which create and reshuffle community and rejigger/nuke social hierarchies.

This post here: [dev.forums.reprap.org] was my attempt to provide constructive criticism.

There are some good points in there, but they also are a massive utter derail of rack and pinion versus belt. Also, it's a huge intractable monolithic block of issues. So, you got shouted down. Raise the points one at a time, in the right place (General Forum, or maybe Library Forum), we can discuss them, one at a time.

One of the most critical parts of any project is deciding what doesn't belong in it, that takes some form of leadership be it a group or an individual.

Yup. Adrian does that quite well with Mendel-Stable. You can do that quite well with Goininreverse-Bot-Stable as a set of wiki pages. And, if you think each part of your new toolchain (hardware and software) deserves a dedicated forum, ask for one. (Although I'd start with a forum thread for each link in the toolchain). Bootstrap your rebellion on the server RepRap server itself. This is RepRap - you can fork it on the wiki/forum. Trust me.

Adrian's point on Leadership was in terms of "Heated Bed, should it be on the server?" And he was being very diplomatic in order to maximize contribution. Hell, he's working on color-extrusion, while simultaneously making sure that we've got one eye on Gada prize.

Hypothetical-Forrest has a point about this thread not being suitable. I think, between the two of us, that we killed it. I shall hope to see your contributions on the RepRap wiki and forum. I'm going to disengage now because I think neither of us accumulates any merit by discussing this further. "Argument Kills Uploads", and all that.

This whole situation happens every other month or so. Someone new comes in, decides they have a better idea for how to do things, and wants to convince everyone that they should focus on it. Often, they also talk about how they are working on their own fork of the design based on available technology y. You're really, really not the first, goinreverse. Forrest shouts at most of them (insofar as you can shout in a forum thread). Nonetheless, while reprap has and continues to develop rapidly towards working designs and improved capabilities, I have yet to see any more closed set of workers produce something comparable without taking advantage of reprap progress. The Makerbot and the Rapman are the definitive 3d kit products getting out into peoples hands by the hundreds and thousands (by the way, if the ramp-up notice about 1000 makerbots per month is true, you're going to be beaten to that 10,000 mark, goinreverse - unless that's who you meant by the right people with the right printer?), and they got that way by taking their preferred forks from the reprap project. The thing that people miss is that, at the moment, the reprap project is almost entirely about research and development - not end product releases. When you compare it to something like arduino or firefox, what you have to realize is that those are not groups that are doing original scientific research (correct me if I'm wrong). What you also have to realize is that the end results of any days work with those projects is code - it can be implemented and tested immediately, and updates can be released continuously. Every development made by a reprapper must be rebuilt by others for testing, and at the same time updating a machine is a significant effort. So what happens is that everyone interested in the project works on whatever they can, and the design increments slowly. When significant enough changes have accumulated, a new release is assembled from them.

tl;dr: My point is that the timeline for a distributed research project like reprap (and make no mistake this is a research project) is much more difficult to fix to a timeline than a software-oriented production project like most open source initiatives. In most cases, a project has a good idea of what needs to be done and how to get there. We have to invent the paths by which we progress. The only way we will ever know which ones are suitable to achieve our goal (remembering the goal is a self-replicating/printable machine, not just one that is cheap and easy to get a hold of) is if we try all of them.

I'm not trying to convince anyone, agreed that results speak for themselves... and makerbot is showing everyone else how it is done. I think their problem is just they can't (or barely can) keep up with demand at this point, As I understand it Adrian is a major investor in makerbot... not hard to see why. I hope they do reach 10,000 3D printers in peoples hands. Seems like all your points are the same as mine and the reason I am forking, I don't want to contribute to a project focused only on R&D, I think self-replicating is a lethal obstacle to adoption (at least until it really works as the replicate in one day concept per adrians papers characterization), and finally the prevalent attitude of the project seems to be dismissing and turning away (shouting down) potential contributors instead of organizing, focusing and harnessing those contributors to improve the end result. Seems like all the breath that has been used "shouting me down" probably could have been used instead to setup a bug tracker or make any kind of release schedule or choose maintainers for sub-systems or ...

At this point I am seeing all the next generation concepts, bowden extruder, heated bed, functional software, next gen electronics, color printing, etc taking place outside of the reprap project.

The only thing that really still seems disingenuine is that the reprap website happily shows the mcwire as the official 'seedling' to build a reprap. Only it can't actually build one, or at least no one has ever done it following the instructions there... maybe just replace it with a page that says "Just buy a makerbot"

... i won't state it so 'harsh', but it's partially true

The basic idea of self-replicating is good, but with our actual situation it's more an academical abstraction, than a real-life-issue ...

My position is very different: - if needing something really sophisticated RP-ed, i have the possibility to make it with my CNC-mill at home or print with a comercial 3D-printer at all the needed accuracies and perfect suited materials.

But this is a killer-argumentation

My sympathy with the OS-/DIY-scene at first and the need for much higher accuracies and other materials than common comercial systems can support, let me dive into developing some 'more perfect' fabbing-methods, compound materials and submicron accuracies.

My aim is more the realization of micro- and nanotech in a DIY-range, so skilled individuals can overcome actual price- and accessibility-hurdles for really 'serious' technologies ... but this is a long way there, so i'm happy with small steps and some occasional evolving leaps forward without shouting loudly for assistance or consensus.

There are some people and groups accompaniyng me or asking for help and more informations, but it's not really a stable community ... more occasional associations for specific problems, dissolving after reaching their predefeined goals.

So i'm happy tinkering around with ideas and methods, collecting and developing more and more parts for my 'unified micro-fabber' capable of making all sort of parts from many materials with different methods and accuracies down to 1 micron in the first stage - before stepping further to serious nanotech

And in the meantime i can help the reprap-development with some ideas or improvements

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

goinreverse, you've raised a number of very interesting points. Which I'm sure this community will be very happy to discuss at length.

However, this is a pretty toxic thread, and we're going to see zero damn progress on any one of these points. So, I'm locking this thread.

In the future, pick one point, either a technical innovation or a fascinating political landmine, and stick to that damn topic. (I've noticed a slight tendency on your part in this thread to rotate through a list of things on your annoyance list, and it's a bit frustrating ... )

I would suggest putting your energy into designing a better post-mendel that all those 10,000 Makerbots can go print. I maintain that the best way to design a better post-mendel is to politely and thoughtfully engage this community, but you may choose to do otherwise. I also think that after this little interchange, you may want to hold off on pointing out the corrupt political process, entrenched, corrupt, inefficient, etc reprap bourgeoisie, etc, and build up some political capital by getting research done and some decent innovations into the wiki, and only then throwing more hand grenades.

Stick around and we'll change things and make things better. And get some research done. Unless you continue to maintain that the best way to help RepRap is to not contribute docs, parts, projects to the wiki and forum.

Lastly, don't be vulgar and bring up forking in the middle of perfectly innocent technical threads, because then I'm going to reliably jump down your throat.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

goinreverse,
I think you are quite right to fork a new project. RepRap is about making a self replicating printer but you are not interested in that. Your aim is to make a cheap 3D printer with 80-20. Nothing wrong with that if you live where you can get cheap 80-20. Two different projects. I don't see any problem with forks that are open source. They don't take anything away from Reprap, they only add to public knowledge.

Quote

At this point I am seeing all the next generation concepts, bowden extruder, heated bed, functional software, next gen electronics, color printing, etc taking place outside of the reprap project.

That is not true, they are all developed by people working on RepRap. The Bowden extruder was pioneered at Bath and the experimental Mini Mendel uses it. These developments are the bleeding edge, so are not part of the current Mendel release, but nothing to say they won't be in the future if they move the project forward in its aims.

---

[www.hydraraptor.blogspot.com]

thanks for pointing me to spota's work. The only issue is that his curing time was 10 minutes! If we have to wait 10 minutes for each layer of printing then its a non-starter.

Objet and others print and cure in a very quick succession. Now perhaps they have the secret sauce photopolymer and that is the cornerstone of their biz, but maybe there is some quick curing polymer available on the general market that can be used.

Kevin

Lock didn't work, but now we're getting good technical stuff, so it's all good.

Also, we have 3 RepStraps coming online in the wiki:
[objects.reprap.org]
[objects.reprap.org]
[objects.reprap.org]

---

I got this email from spota:
Well, to tall the truth, since I have changed jobs and apartment, I haven't been doing much research on the UV-resins front.
I don't have a space where I could install my workshop at home anymore and all my material is lingering in a storage bay somewhere...

I know it's a shame, but quite frankly, unless I can rent some garage or similar to start doing trials again, I don't see how I will be able to help you guys out, except with my knowledge.
I still answer with great pleasure and depth of detail any request of further details that I get, though.
I have been talking with the guys over at diy_3d, with their inkjet model. It's quite interesting what they do and the resolution could be staggering, if you'd get the cartridges to work with methacrylate.
Over at RepRap you have come quite far, right? You are allready shelling out RepStraps at a major rate and people are doing nice pieces with extrusion. Do you know if anybody dealt with resins yet?

---

It may be good to create a photopolymerization working group, or maybe a polymer working group. We've got some of that filed in various forums, but I think if we create a space "Polymer Working Group" (and announce it on the blog, along with the occasional status report), we'll get 10x or 100x more contribution/R&D on this subject.

So, do you guys want a "Polymer Working Group"? If I get 3-4 yes votes, I'll start one.

> So, do you guys want a "Polymer Working Group"?
> If I get 3-4 yes votes, I'll start one.

YES!

Hi Sebastien,

count me in

I received some mixtures from Fernando/spota but found that my UV-LED's didn't have enough energy to cure the resins in normal fabbing time, so went to comercial UV-resins/glues ...

What's interesting too are heat-curing epoxies - at room-temp they have some days pot-life, but harden fast, when heated over 80 centigrades.

With diode-laser-heads this should be a much easier methode than using UV-mixtures and UV-lamps ...

Edited 1 time(s). Last edit at 02/24/2010 05:21PM by VDX.

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Viktor, what is your current UV-LED setup? Would something like this help for UV cure?

[ledshoppe.com]

ask google about the DIY open-air nitrogen lasers

---

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

Hi Corwin,

... look here at the atached image "Disp-UV.jpg" - more and brighter should help, but it's mostly an issue with the wavelength too ...

On the subject of keeping threads on topic.

Is a potential solution to implement moderation, where the thread initiator is the default moderator.

[dev.forums.reprap.org]

---

Necessity hopefully becomes the absentee parent of successfully invented children.