Clone prusa i3 or Original Prusa i3…
-
Thanks to all for your valuable input. I appreciate it. As far as kinematic go, what is the preferred design for those who have experience and would be in the market to build something new. I personally have only been into delta configuration since the beginning, with the exception of a Form 1+ SLA, but that's not a fair comparison, but am really interested in the pros and cons of alternatives such as the corexy/yz vs. Cartesian vs. My delta printers, in particular, which has the most ability to be fastest and most accurate in the long run?
-
For multiple filament printing, three options exist.
One is the single head, retract and replace; Prusa's MMU 4 color feeder, Promethius (i have it but not assembled).
Another is single carriage, multiple hot end; like I had on my BigBox Dual.
Lastly, is IDEX; Independent Extruders; X1 and X2; optional for BigBox owners.Each have their own set of issues and complexities. The biggest issue is alignment of the dual head, and not having the other head ram into the top layers of the current print.
This is my main reason for even considering the Prusa MMU option.
-
There is a 4th option: use a mixing hot end such as the Diamond.
-
….................................in particular, which has the most ability to be fastest and most accurate in the long run?
IMO, people in general seem to have a fixation about speed (or come to that a fixation that accuracy and speed are inversely proportional). I have seen many many designs or plans for people who want to build the ultimate "supercar" of the FFD printer world or on the other hand, the ultimate indestructible "offroader" that will for some reason that I have yet to fathom, be more accurate. The problem is that it all comes down to melting filament and squirting the result out of a nozzle. The speed that we can print at is ultimately determined by how fast we can melt the filament. What's the point of building a machine capable or printing at (say) 600mm/sec when we can only melt filament at the equivalent rate of (say) 150mm/sec)? Back to the vehicle analogy, why build a car capable of accelerating from 0-60 in under 4 seconds with a top speed of 200 plus mph when all we can use it for is driving in a 30mph speed limit in heavy traffic. On the hand, people build heavy duty machines and proudly call them "tanks". Fine but why do want a tank to do the school run? - you'll never wear out a Fiat 500 on that journey so it would be perfectly adequate.
I know nothing about Deltas but from what I've read, they seem more difficult to set up than Cartesians/CoreXY. I'd have thought that keeping the moving mass low would be more important with a Delta than it is with a Cartesian. If you want big in terms or XY, then it'll be easier to build a Cartesian/CoreXY than a big Delta. In theory, CoreXY have an advantage over "conventional" Cartesians in that both XY motors are fixed so the moving mass can be reduced and both motors share the load (other than for 45 degree infill). In practice, I'm not convinced that the extra complexity is worth it. Come to that, I'm beginning to think that the 3D community has become a bit too obsessed with saving mass. For sure it will enable faster acceleration but can the extruded filament keep abreast of that acceleration? One side effect of high mass that I have discovered is that you don't get "ringing". Mass tends to damp out vibration so it's not all bad.
Just the ramblings of an old man….........
-
With you on that Ian. Speed is a gimmick, sure if I could produce on my delta or corexy a part in 1/10 of the time that was just as good I'd be interested, but whilst I could print a bit faster its the laying down of accurate filament which suffers as the speed increases. Nothing you do to make the moving mass very light is going to change the fundamental properties of the plastic, and the fact you have to melt it fast enough (can hit the limit here fairly easily), and control its deposition accurately at high speed.
A new approach is needed, such as purpose designed high speed filament and multiple nozzle deposition - oh wait I just remembered its polyjet technology isnt it?
-
Thanks again guys for very valuable insight. Haha Simon, speaking of polyjet, would you happen to have a few spare hundred thousand dollars I can borrow to outfit myself with the HP tech you mentioned d above?
Thanks again guys.
P.s. -> hey Ian, would you mind if I essentially copied your machine?
Cheers.
Jonathan
-
…................P.s. -> hey Ian, would you mind if I essentially copied your machine?
Cheers.
Jonathan
No problem but I don't have a build log or complete bill of materials. The entire thing was designed using OpenScad. I'm happy to put all the OpenScad files into a folder on my Google drive and create a shareable link so that you can download them. Note that I'm not by any means a professional when it comes to OpenScad, so some of the files might not be pretty but they get the job done. I haven't got time to go through the machine and catalogue all the various bolt sizes and lengths, but you can work them out from the OpenScad files. Send me a PM and we'll take it from there.
Ian
-
Again, were getting away from the original post…
The diamond/gem/ruby tipped nozzle is NOT a solution for multi material (two or more filament) printing on the same object. What it's intended for is to handle any material without nozzle wear.
Has anyone first hand experience in using the Prusa Multi Material/multi filament solution?
-
Again, were getting away from the original post…
The diamond/gem/ruby tipped nozzle is NOT a solution for multi material (two or more filament) printing on the same object. What it's intended for is to handle any material without nozzle wear.
Has anyone first hand experience in using the Prusa Multi Material/multi filament solution?
No. We aren't talking about a tipped nozzle, we are talking about a hot end. This one http://reprap.org/wiki/Diamond_Hotend
3 inputs, one output. The main advantage over the Prusa set up is that all 3 filaments are loaded at the same time. So to switch between them, you don't have to retract one, then feed in another. Also you can mix them, so mixing Yellow and Red will give you Orange. If you take a look at my blog, you'll also see ways that you can print multi-colour parts without having to use wipe or prime towers. In a nutshell printing multi colour parts is faster and uses less filament than a switching method.
-
I'd be interested in those files also. Not sure I'm quite ready to build one but I'd love to look at how you did it.
…................P.s. -> hey Ian, would you mind if I essentially copied your machine?
Cheers.
Jonathan
No problem but I don't have a build log or complete bill of materials. The entire thing was designed using OpenScad. I'm happy to put all the OpenScad files into a folder on my Google drive and create a shareable link so that you can download them. Note that I'm not by any means a professional when it comes to OpenScad, so some of the files might not be pretty but they get the job done. I haven't got time to go through the machine and catalogue all the various bolt sizes and lengths, but you can work them out from the OpenScad files. Send me a PM and we'll take it from there.
Ian
