If you redesign the puzzle twice as large then I am sure many things can be achieved. Just upscaling the STL's by a factor of two might not be a good idea though because it will be really bad on the tolerances (with this many layers, many tiny gaps will add up).

Personally, I wouldn't consider FDM for a project of this complexity for a second. incredible's printer uses 0.25mm layers whereas Oskar describes TNO's SLS service with 0.15mm layers as barely usable (Shapeways offers 0.10mm layers).
At this price point, it pays to look at different vendors to find the best price. Maybe you could look into using a different printing technique such as SLA. SLA has the potential to provide much smoother pieces and even thinner layers.

Good. Only the mayor problem's. That means Barrack Obama might not deal with it.

You meant "major," didn't you?

I am looking at the pieces of you 17x17. Your design has a flaw. The flaw is pretty big also. I don't know what type of mechanism you used, but there is something wrong with the pice in the picture. If you used v-mech, it is clear that you did not make the rivet large enough. Therefore, it looks like to me that you have about 1-2 mm holding in the rest of the mechanism. This is a problem because if that 1-2 mm rivet is overlapped at any point, the entire edge will explode. A solution for this would to be to make the rivet a lot larger. Even if the rivet goes past the first or second piece (which it should), you can make the mechanism work by having the piece go out and around the mechanism. If I am to confusing for you, don't hesitate to ask for a digram. Hopefully this helped.

I personally think that the sphere core will help the most. That way, if the pieces try to move from where you intend them to move, they have one less axis.

V Mech doesn't work past 11x11x11, and verdes had an idea for a 12x12x12. This is a design oskar came up with and we call the Pagoda Mech.

he has his own 3d printer?

I understand where you going. And yes, I don't live in a cave. I can't believe you thought that I didn't know what a podaga mechanism was. I also posted earlier in what you quoted that the mechanism doesn't matter. The piece he showed is proof that there is a thin layer. This layer is only held in by (as you can see in the picture) be about 1 mm. This means if the mechanism passes over this tiny hair of a rivet, the cube explodes. This is just from the simple laws of big cubes.

Quiet a few members do. I may be mistaken, but I think RubixFreakGreg has one. Either that or I'm confusing someone else.

No, that would be VeryWetPaint (he has a SD3000).

what is the average cost of a printer? One with a good level of detail

I can get a 1600 dpi printer at the mall for about $100.

Seriously, the average 3D printer might cost $50k. That said, prices range from $15k to over $200k.
I know of three "affordable" printers (all around $15k with all bells and whistles). incredible has a uPrint which in my opinion, isn't all that suitable due to the low resolution but apparently can make puzzles.
A better choice might be the V-Flash (also around that price) which looks really promising to me but has never been tried for puzzles (I want to see if I can get a sample made some day).
Then there is the SD300 which is owned by forum member VeryWetPaint but that one doesn't really have much puzzle potential.

Alongside the ~$15k cost of the machine, expect to pay much more for maintenance, spare parts,... Build material can be especially expensive.
It costs approx $.30 per cm^3 for the uPrint. Compared to Shapeways' $1.50 this might seem good but keep in mind that hollowing out parts doesn't work and that you also need to account for support material.
V-flash resin costs around $.55 per cm^3. I think it allows hollowing but also requires supports to some degree.
The SD300 has the lowest cost of $.10 per cm^3 but the entire build volume needs to be filled with material so you can't compare this figure directly.
Of course there is the <$1k RepRap but that machine won't satisfy any puzzle building needs whatsoever.

This post doesn't have anything to do with the original topic, so if anybody wants to comment on household 3D printers, please do so in a new topic.

For those of you that missed it, I started a new topic on household 3d printers: viewtopic.php?f=7&p=220794

Claus,

I would be happy to make a new design for the 17x17x17, and scale it up while keeping the 0.2 mm clearance. Please provide me with a list of all the things that are wrong with the current design.
-Too small pieces
-Hooking of pieces with core (Photo? Video?)
-More ... ?

Thank you.

Oskar

Imagine doing an average of 5 with that beast, and if someone can solve it blindfolded, i'd build a rocket in my backyard, just to be safe as the universe implodes:)

Is there anyway to print just part of it, or a highly bandaged version so you can get a proof of concept going without the huge cost attached. It's terrible that someone spent so much on the first one and it didn't work.

Because I have no live, I stalk Oskar's Shapeways shop and I noticed this.

I beleive he will make a new thread soon for this living Monstrosity, and a Video.

Text and pictures are from his Shapeways Shop, I take no credit.

PS this design IS printible

http://www.shapeways.com/model/64058/ov ... tml?gid=rg





Think that it is just a screenshot? Think again.



More pictures can be found on his shop, just posted my favorites here.

This copied from his shop:

Over The Top is a world-record 17x17x17 Rubik's Cube. It makes quite pretty patterns. It can take many, many hours to solve. The puzzle will be first presented live to the world at at the New York Puzzle Party Symposium, Saturday Feb. 12th . A YouTube video will be published after that event.

The complete Over The Top 17x17x17 puzzle requires
-1x Set 1
-26 x Set 2
(Actually, you need only 24x Set 2, but it is strongly advised to order some spare parts.)

Please contact Oskar directly if you are interested in obtaining a fully colored and assembled sample of this puzzle.

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

Again, I take no credit, I just wanted to informally reveal this to everyone, and because he already has it public on Shapeways, I figure that it is okay to post here.

Oskar, if that's not okay for some reason just let me know.

EDIT: Oh and Oskar GREAT JOB! congrats on the amazing achievement.

Adam

Wow, that's amazing. So, a working prototype finally got made. Even though, although it's mechanically a 17x17x17, it's actually more of a babyface 15x15x15.
A momentous achievement, congratulations Oskar!
PS looking at the pictures, I really love the patterns the faces form! Remind me of modern art. Also, which of the 4 (IIRC) mechs did you settle on using?
EDIT: ok, I see what mech it is now. Stupid me.

Oh my.

YEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEESSSSSSSSSSSSSSSSSSS!!!!!!!!!!!!!!!!!!!! IT WAS FINALLY BUILT!!!!!!!!!!!!!!!!!!!!

i told you guys to stick with a cubic design.

I would totally buy one of these if i had the cash.

WOW, incredible, I never thought he'd get it working after the problems with the first version. Again, WOW!!

I guess a stickered version would be too much to ask!

So when will oskar make a youtube video of this? Also, he hasn't solved it yet right...

And it's only $2000! Doesn't that make it cheaper than the 12x12?

Also, when will we get dyed edges? I am glad that this is now bigger, so that the pieces are able to stay in and it is supposedly more stable (and easier to assemble). It's bigger than I would have thought though (monsterous, even). Seriously, this makes me want to cry. It shocks me. Tugs at me. Moves me...

PS. I wanna see someone speedsolve this.

Email from Shapeways

I beleive that a video will be made somtime after Feb 12th, as said earlier.

And I do not think he has solved it yes, but there is a partially scrambled picture at his shop.

I too love this cubic version over the pillowed version, and it seems that using shapeways gave much higher quality in the final build to allow this beast to come to life.

And it is solved like a babyface 15x15x15, I think stickering this puzzle would probably be THE STUPIDEST idea ever attempted by humans. ever. 1,734 stickers is way too many.

Maybe oskar could print the edges in 2 parts and dye them indempentantly, like a Dayan stickerless cube.

Again, fantastic job Oskar. I expect much higher order puzzles to come from this mech in the near future.

ALSO, do you think that this mech is mass-producable? do you plan to patent it?

Adam

why not just sticker the edges and corners? or spray paint them?

I wanna see someone break an hour on this thing. It took me 8 hrs on a simulator!

It would be easiest to split the pieces in two and dye them. The 'tiles' approach wouldn't be good, because as on the first prototype the tiles would probably fall off...
Also, is it even possible to sticker this thing? Each facelet looks smaller than a cubie on a micro 3x3x3!

im sure its possible. Id recommend tiles though, cause if i stickered it, id want it to last!

so how long did it take him to assemble this? I watched esquimat's 11x11 tutorial and i thought it was torture, oskar deserves a medal for his dedication.

So the petaminx is no longer the biggest twisty puzzle in the world (in # of pieces).

I was curious what the world record holders could do on the 17x17x17 so I threw the current world records into Excel, plotted them, and added a trendline (2nd order polynomial seems to fit and it makes sense).Looks like it would take the pros 30-35 minute But that's assuming this turns as well as the "smaller" cubes and isn't cumbersome.

I wanna see this thing pop during a speedsolve.

Wow, just wow, I think my mind exploded. . . school no longer seems important. . . just wow!

My heart's been going like 300 b/m for the last hour, but i think my cubegasm is starting to die down.

The Micros have smaller pieces!!

At least it makes deciding on my all time favourite puzzle a lot easier. Once I see one in black plastic with stickers of course.

o_O. Wow. That is phenominal! Great work. But I agree with Tony, there needs to be one with stickers.

send it to me il do it...ive got plenty of time on my hands as i work for norfolk county council(english cutbacks joke)....i need a reason to stay in starbucks for 3 days straight....
ps my god that cube is awsome...outstanding work....

Perfect!! I can't wait to see a video!!!!

- Greg

Completely
insane!

Sorry, the video has to wait until the New York Puzzle Party Symposium ("by invitation only") on 12 February, where I shall present the full story (including Claus' epic prototyping efforts) and the working puzzle. You can see more photos in my Shapeways Shop. You'll probably see more teaser info from Shapeways during the next few days.

Oskar

Wow!

It's the only way I can sum up my feelings without violating the "no cursing" standard. And I don't curse a lot-- it lacks originality and flair, but this reduced me to base thoughts and expressions.

Beauty and the Beast

We'll all be waiting with baited breath, Oskar. Congratulations!!

AMAZING. I can't think of any other way to describe it.

This is one pretty incredible achievement! That's all I can say now.
I would imagine there must have been a sponsor to print this puzzle.
Oscar, is that so? Is the sponsor going to own this puzzle? Just asking.

The first two prototypes have been sponsored and built in an epic effort by Claus Wenicker. Shapeways has kindly sponsored the third prototype, which is offered for sale at my Shapeways Shop, in exchange for some publicity. I assume that I can keep the prototype after my presentation in New York.

Oskar

We'll I have to change my pants now.


Does this make 3 17x17x17's? ( two functional and claus's first try)

WOW!!!! I must confess when I first saw that mech I didn't have high hopes. I couldn't be happier that I was wrong.

Any chance this might see some form of mass production? If this works in strong white and flexable I would be inclinded to think it would work even better in injection molded plastic. And go the way several MF8 puzzles have gone by using colored plastic and not stickers and you solve the issue of how to sticker this beast. At $2000 I don't see myself ever owning this beast get it down to $200-$300 and I'd LOVE to have one... even if I had to put it together myself.

Carl

I dont think it will ever go down to $200-$300 lol.. but I got the same hopes =)

Sharon

This makes v cubes look puny.

But I wonder which one is higher quality....

Oskar, I can proudly tell you that once that vid makes it's way on youtube, my first actions (after watching it of course) will be to favorite it (on all my channels).