Page 1 of 2


PostPosted: Tue Apr 21, 2015 1:38 pm
by scipione205

I would like to start a new collaborative project for the next printer. I choosed to call the project 'Nicola' after a man who worked hard for the Socializzation of the Enterprises, and was killed at the end of the WW2.
The goals of the self-replicating machine and personal fabrication are pretty close to the goals of the socializzation.

However, the machine must meet this requirements:

- simple to source parts (printed and vitamins)*;
- apart from printed parts, eeryone will be able to build the machin with very basic and simple tools;
- high rate of self-replication;
- cheap price (less than 200$/€ will be perfect)**;
- capable to print at least PLA and ABS;
- very good precision and repeatibility;
- good speed.

I have to explain something:
* please consider that some items are HARD TO FIND or even IMPOSSIBLE TO FIND in certain places; for examples, in the US you can buy sch40 PVC pipes and fittings for few bucks everywhere, but in Europe not (and in italy where I live, it's impossible); some pieces such as precision shafts can be quite hard and expensive to source.
** the key of the real third revolution is the acces to this technology. There are a lot of people which could do very much with the acces to those printers, but they can't afford the price. For example, I live in SOuth Italy and I am in a sort of vicious circle so that I don't have any money (any... don't speak about 400 or more euros) because I can't find a job and when I work they even don't pay me (I worked for SA Group for six months and they didn't paid me), or if someone pay me they pay very little (sometimes I work as DP or as Camera Operator and DP - yes, operator and DP at the same time... - and I receive 50€ for 8-10 hours of work, less than a third of the actual value...).
I feel very bad for this condition and I think that there are a lot of other people like me (and in those days a lot of them died in the sea while trying to reach Italy from Africa, and it happens very often).

I have some ideas... maybe the 'cross screw drive' (but I don't know how to easily and cheaply source the screw axis) or something like the LISA. I know that the design must be rid of linear shafts and linear bearings (and also the linear bearings obtained by arranging skate bearings, because of backlash).

Any idea?


PostPosted: Tue Apr 21, 2015 2:59 pm
by Nicholas Seward

Code: Select all
| \
|  cheap
| /

Typically, you can pick one edge of the triangle above. For a $200 machine, I would drop the fast requirement. Remember that robots can do work without humans so who really cares how fast they go.

I do have an idea that is fully Cartesian but uses no rails. It could really be made out of a bucket of 608 bearings and some scrap wood. You might need two springs (rubber bands could work in a pinch).

I don't have time to draw it up for a few days but it is a variant of the RepRap Tuga. Lay it on its back. Have the arms roll on the table itself instead of slide on rails. Put the bed on a box with rollers. Obviously, you will have to use your imagination with that "wonderful" description.


PostPosted: Tue Apr 21, 2015 3:54 pm
by scipione205
Hi Nicholas,

thank you for the suggestion. I will clarify about the speed.

By saying 'good speed' I don't mean 150 mm/s ( xD ) but something around 50. I know that robots can work without humans, but I can't wait to see printed a set of gear for tests... :D

I understood: the arms rolls in someway like a parallel robot, moving the end effector on X-Z, and the bed rolls on Y axis.

The springs are for preload?


PostPosted: Tue Apr 21, 2015 6:16 pm
by Nicholas Seward
CARTESIAN.png (62.93 KiB) Viewed 60591 times

This is my basic idea. It is 100% Cartesian but doesn't need rails. The effector is on the top most point pointing into the screen. You can see that this is made to slide along a table. Accelerations will have to be kept down to keep from losing contact with the ground. However, all the parts cost almost nothing and can be made from almost anything.

Feel free to ditch this idea. It has just been sitting around in my head and I thought this would be a good time to get it out.


PostPosted: Tue Apr 21, 2015 6:43 pm
by see3d
You need to hone the reasons for your goals better. Already a $200 retail 3D printer is being developed called Tiko. It is a Kickstarter project right now. The way they are achieving this is by simple design with cheap parts and high volume production. The cost of their parts is $70. I doubt they are the only ones trying to bring down the cost of 3D printing below $200. So, now that the psychological price barrier is broken, perhaps the thinking should be in terms of less than $150 (for a < 150mm^3 volume printer).

Another critical factor that is needed for an FDM 3D printer is a stable supply of power for days at a time. Most of the 3rd world does not have reliable electrical power. How does this figure into your goals?

What about the tools to design parts to be printed? Free browser based CAD tools are available, but you need to have long periods of access to the internet to use them.

I am not trying to dissuade you from proceeding, just challenging you to be as aggressive as you can imagine with your goals. Having RepRap as a goal is fun, but may involve a tradeoff with other goals.
Do you want to have fun with this?
Do you want to stretch your imagination?
Do you want to gain new skills?

These are personal goals, not societal goals. Often the personal goal must be included or nothing practical comes about other than dreams. Dreams proceeds action but does not equal action.

Engineering often revolves around how to design things within the practical constraints of the engineer's access to tools and materials. What supplies do you have access to within your budget constraints. Can you obtain some things for free? What tools do you have access to for designing parts, for software? For materials (electronics, mechanics, bearings, etc.)?

I think there is lots of room for new and unique low cost 3D printer designs.


PostPosted: Tue Apr 21, 2015 8:12 pm
by scipione205
see3d, yes I have also other concerns such as energy (not only for developing country, and I have some ideas to address that) and PLM (not only CAD or CAE or CAM, but PLM) software (I would like to start a project of a parametric CAD/CAE software like CATIA but fully open-source which could became a full PLM some day).

I want to free this technology and also the world, and doing that sure I want to learn, discover and have fun. I think that this machine can help people to work, innovate, discover etc.

I just saw the Tiko project. First of all, I was talking about a fully open source machine, and I don't think Tiko is open source (maybe in the future they could disclose some files).

I already thought about a delta, and now after the dinner I will study all the previous thread discussion about LISA here on sourceforge and also on the RepRap forum to understand some things.

I thought that a Delta without linear motion and with preload on the lead nuts could be cheap, simple but still precise and accurate (even if slow).

I don't trust the Tiko's linear motion design... I'm very afraid of plays and backlash, but I'm sure we could find a solution (hey, the humans just flew 3 times the speed of sound and landed on the Moon...).

I either considered a delta manipulator without linear rail nor leadscrew. Common stepper motors coul be used, with a reduction gearbox. The gearbox could be very cheap (gears printed by other printers) but still precise and backlash-free, using anti backlash gears. It's not that difficult, just a gear split into two with a spring... I've been in those gears for more than a year (not only spur, but also bevel, miter, and right angle helical) because I worked on a follow focus for more than a year... but only in SolidWorks and CATIA because I never had acces to a 3D printer to test my stuff..


PostPosted: Tue Apr 21, 2015 9:27 pm
by see3d
The Tiko is not open source. But that does not mean that we can not learn from them. For instance, they use a geared stepper that only costs $1, or $2 with a driver electronics in quantities of 5-10. It is cheap because they were designed to be made by the millions by automation for actuators in air handling. That is one of the ways in which I have found to make things less expensive -- cross purpose parts that are made in high volume for some other industry. You have to work within the limitations of those parts though.

Tiko designers say that they have figured out a method and have eliminated the play and backlash. They have shown it printing and the precision looks good. If you have gravity working for you in a delta, you can use it to get rid of play. I think that is what they did. Still, the motors are not the only source of play. The universal ball joints need to operate without much friction or play. I am not so sure their production design will work first time for that. I am not saying that a delta design is the best one, only that it has been used by many as the basis for a low cost design. I am sure other ways of doing a low cost mechanism have not been exhausted.

Check out Onshape for a free (for those who do open source designs), parametric CAD. Onshape is not open source, but it going on 100 million dollars so far to create it and support it. No OS project will be able to match that with the current OS development models. People would have to dedicate their lives to such a thing for free. That can only happen with a complete change in models, where the OS developers physical needs are supported by governments (i.e. society). The world is not currently headed that way, but it is a viable model. I happen to believe that an OS model along those lines with the added motivation of friendly competition could work better (faster innovation), than the private ownership business for profit model.

Some of the fastest innovation advances happen during war. Large segments are given the incentive to work together quickly to preserve their society. The race to space was similar -- without war. It is all a matter of getting priorities aligned and goals set. It takes directed resources to make advances. Strong governments can direct large resources to a goal. Large corporations can direct large resources, but not as much as strong governments. Wealthy individuals can direct some resources to a limited number of goals.

How do OS projects gain the kind of resources to compare to those? Just part time volunteer labor, or retired persons who have their needs taken care of? It is possible, but the scope of the project has to be limited to the reality of the resources.

One advantage of an OS community though, is the barriers to entry are much reduced, since the open source is available as the starting point for further innovation. That means that a whole society based on OS rather than patents and trade secrets can bootstrap itself much faster in overall "business" progress.


PostPosted: Wed Apr 22, 2015 8:14 am
by scipione205
Yes I really like the cross-purpose thing you said, however I didn't understood what kind of motor they used.

Yes the gravity can work for you against the play in some circumstances (like a preload), but only for travel which are exactly vertical (and I saw the Tiko carriage, there would be play also normal to the travel axis), and this only when there aren't such accelerations to overcome the gravity's one.

However, the backlash in a leadscrew is fairly simple... just two nuts with a spring, is not that difficult nor expensive. However I am really scared about the linear motion because I know that even in very high-end applications there is always play. I'm paranoid.

Regarding Onshape, I will discover it; I always liked Jon Hirschtick, I will check Onshape out and understand if it's free for us.

Regarding the cost of developing a non-profit opensource CAD/CAE/CAM, yes I understand that it will cost 100M$ or more, but please consider the Blender case: it certainly costed more than 5M$ (now I can't tell how exactly), and Ton Roosendaal raised the funds with Open Movie Projects: in every OMP the Blender Foundation paid artists and developers, so that new improvements and functionalities has been implemented. We could do the same with a PLM software.

I think that we will destroy the intelelctual property someday.


I was considering to get a loan from a friend to build an Ultimaker from scratch (becase he and other people are massively asking me to build them Filmmaking Equipment which I already designed - of course I will release everything as open source hardware :D ) , and I asked a quote for the linear shafts to IGUS (for me it's very hard to source linear rods and bushing, then I discovered igus).
I was surprised that they quoted me 14.59€ including VAT (which in Italy is 22%) for all the linear rods required for the Ultimaker Original+. So that I thought <<ok I will ask them for samples so I will not have to pay for linear bearings and bushing>>.

Then I discovered this:

Maybe in the future we could print ourselves bearings, bushings and lead nuts, using this "tribofilament".


PostPosted: Thu Apr 23, 2015 3:04 am
by see3d
They use a small 5V plastic gear stepper motors with about 2048 or 4096 neighborhood steps per revolution. Driving the motors in a bipolar way with 12V driver makes them much more powerful. It is not clear if they do this or just use the cheap 5V unipolar drive. Even that cheap way can be improved by using a 12V kick +5V drive. I would expect they would do at least that much. I designed those kind of drivers 40 years ago.

The new filament is interesting for wear parts, like gears or racks. The rest of the specifications will have to be checked out.


PostPosted: Thu Apr 23, 2015 9:49 am
by scipione205
which kind of motor is it? where do can I find it? I would need it for some experiments...