RepRap Google Summer of Code 2010 Call for Ideas

Hey All,

For those of you unfamiliar with Google's Summer of Code (GSoC) it is a payed Google position offered to ~1000 students world wide to work with a open source organization as developers. The students are given specific tasks and a summer to complete them during which time they regularly meet with their open source group and google in working on their development task.

GSOC has an application deadline for organizations on friday. I for one would be very happy if RepRap got in on this, partially as a student (count me in!), and partially as a RepRap dev where I see taking on full-time world-class student developers for free as having very few draw-backs

So heres what we need: Ideas. Ideas, Ideas, Ideas (Developers... no wait, wrong speech.). Anything you want, need or think would be cool on a RepRap is exactly what we want. I've set up a wiki page for this and will be linking the GSoC application to this page so go - get your idea on there!! We have just over 48 hours!!!

Rep Rap Wiki: GSOC 2010

Edit: evidently I can't count time periods. It's due friday.

Edited 1 time(s). Last edit at 03/10/2010 11:59AM by theOtherRob.

Great! It would be wonderful if people could participate! I'll add a few ideas!

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

bump: still could use ideas up there! If your unfamiliar with the wiki just post your ideas here and i can copy them over!

Edited 1 time(s). Last edit at 03/16/2010 05:01PM by theOtherRob.

I added a few. Now to list them under my team thread in the Gada prize to take credit!

thanks rocket scientist! I'm going to keep tabs on this page as it seems to have allot of stuff we should put into the next gen RepRap (IMHO as a lowly wiki admin .

Unfortunately we didn't make the deadline for GSoC - this was very last minute and we gave it our best shot given the 48 hours we were left with but ultimately we just didn't have enough time left. Next year though I think the feeling is that we should start working on this a couple months in advance and have a strong proposal going into GSoC.

Thanks for everyone who posted ideas and all the people on the RepRap core team who helped in the effort towards GSoC 2010! See you again next year.. maybe a month earlier

I think we should keep the page on update it as time goes by. Is there some way we can convert it into a "I would like this" and "I just finished writing one of those"? That way people who are not so good at programming can make a request that should benefit the whole group, and others who are not good with hardware but want to contribute can look for open reqs for ideas on what to do.

We could also continue feeding it things we would like for next year's Google Summer of Code, should we get selected.

Speaking of modifying the code, I have only been on the forums a couple of weeks, but I have no idea yet where the source code is and how to help advance it. Perhaps that needs to be another wiki page, or do the current developers prefer that anyone who can't even find the source has no business modifying?

Mike

---

Team Open Air
Blog Team Open Air
rocket scientists think LIGHTYEARS outside the box!

Hey Mike,

A belated welcome to the forums! Code is certainly an issue for this project, there is an SVN and only recently Adrian set up a bug tracker.. basically if you want to submit code and eventually get commit access you should go through Adrian (he's friendly and has never eaten anyone that we know of).

SVN link in it's impossible to find place (agree we should have a more obvious link somewhere, not sure where, if you find a spot please feel free to add it in): [reprap.org]

also, Adrian Bowyer, our fearless leader (contact info is at the bottom): [people.bath.ac.uk]

Last thing, there's a development category on the wiki for everything from ideas to finished, working reprap mods that are large enough for their own page but I think your right, we need another page for small, one paragraph or less ideas so that people can easily read through the list and pick up a project their interested in seeing happen and then they can graduate to the development category.

How about adding backlash compensation to the firmware or the hardware?

-J

---

-J

True wisdom lies in knowing that you know nothing

I think we missed the deadline for the programme.

---

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

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

J-Dub, good suggestion!


The best way to handle backlash is in the hardware. There are special gears designed specifically to eliminate backlash in that gear stage anti-backlash gears. They have a pair of gears and springs that push them to rotate in opposite directions, causing them to 'grab a hold' of the other gear teeth. The more general approach that works on chain and belt systems as well as gear drives is to preload the drive with a constant force in one direction, causing all parts of the drive train to tighten up the slack in one direction. This can be a long, low tension spring on the back side of a chain or belt, or even drive wheel, pulley, sprocket, capstan, gear that is driven by a hanging weight.

If you cannot eliminate the backlash in hardware, then the usual software approach is to always move the axis with backlash in the same direction when doing precision work. For example, this might mean always laying down filament with the X-axis increasing and the Y-axis decreasing. But that does not allow for alternating the 'lay' in each layer, so that will not work well here. The next best is to have constants stored in a table somewhere the indicate how much to offset each axis depending on whether it is going forwards or back. I once wrote a self calibration program for an automated X_ray machine that inspected solder joints. To get the required accuracy (this machine had a 2 foot X and Y stage, a tank turret gear for a plus and minus 180 degree rotation stage, all mounted on a 0-45 degree tilt stage, all on a z axis drive that also moved about 2 feet up and down. The X and Y axes would shift more when tilted, and the direction of shift depended on the angle of rotation. The Z axis was not perfectly aligned with the X-ray source and camera, so going up and down also changed apparent X and Y. In short, every axis was affected by every other axis, and the equation to compute the image X and Y deltas used 39 equations of X,Y, Sin and Cos Rotate and tilt, and Z!

Mike

---

Team Open Air
Blog Team Open Air
rocket scientists think LIGHTYEARS outside the box!

That sounds like a fun project! Another idea that I thought would be nice is if there were an MDI (manual data input) section in the host software. One can dream can't they? I need to finish my lathe though, so I can get my extruder design completed. Then I'm on to my heated bed concept, so programming will have to wait.

Cheers,

-J

---

-J

True wisdom lies in knowing that you know nothing

Hey J-Dubs,

I hear that RepSnapper has this - if you can figure out how to find the download link (I've lost it now that his blog is no longer in English) then it should provide the G-Code console which I think is the same as this manual data input your talking about (windows only/linux with lots of hacking).

see also Wiki: Comparison of RepRap Toolchains