Mondrian
Release status: working
Description |
a modular cartesian RepRap
|
License |
|
Author |
|
Contributors |
|
Based-on |
|
Categories |
|
CAD Models |
|
External Link |
|
Here is what I wanted to try among the possibilities to make a Foldable RepRap.
I didn't pursued the XY-head + telescopic Z because I couldn't get my head around how to do it properly at this moment.
But after having seen the Create it gave me some new ideas, for another concept after the FoldaRap, it then evolved toward this cartesian but modular architecture, the folding plane idea will have another name :)
<videoflash>PEaP7HCyGPY</videoflash>
<videoflash>rwdRH7mGj-M</videoflash>
Specifications
- Printed Parts: 30
- Non-Printed Parts: +/-400 (counting every washer/bolt/etc)
- Material Cost: 100€
- Cost: 600€
- Printing Size: 200 x 200 x 200 mm
- Resolution : XY = 0.0125 mm ; Z = 0,00025 mm with m5 rods
- Accuracy : 0.1 mm
- Speed: 150 - 500 mm/s (print-travel, max speed tested yet)
Special features
- Modular design
- Based on standard 20x20mm aluminium extrusion (with 6mm slot)
- The frame is the linear guide for X, Y, Z axis
- Easy to build, with few tools
- 1.75mm Direct Driven Bowden Extruders (dual-head)
- small footprint : lenght of extrusion plus 40mm (240x240 for 200mm ; 340x340 for 300mm ; etc.)
Naming Convention
- Mx_xxx : Mondrian (non-foldable verion), version, print volume in X, Y, Z (in dm, add +/- 15cm for the footprint)
- then you can build things like : M222, M333, M763, or even a huge M999
To Do
- mini-cnc addon [66]
- bio-printing
R&D
adding a little engraving on the z-plate
new back/elec-cover, adding a panel-mounted usb plug, a hinged lid to access the electronic and a second spool holder, small holes to add a zip-tie to hold the cables coming from the head and a springed notch for the upper elec cover to lock in the back panel
need to be tested, a third revision of the z-pulley drive, if it works it will allow to remove the remaining M5 nuts and switch from MR105zz bearings to the more common 605zz. A spring effect in the printed part is taking care of removing any play in Z, but the part need to be adapted for the height of the pulley.
new plates arrangement, to ease color associations ;)
the latest prototype, now fully carterised :)
the endstops are being integrated in the printed parts
one of the corner can host the pcb webcam of a RaspberryPi
replacing the 40x40mm fan by a mini-blower
rebuilding, the latest version, in short height (for traveling with)
adding holes to allow access to the motor bolts even while the carriage is on the profile and the rack is in place (i.e. : you can remove the motor at any time)
adjustable play compensation
rack-n-pinion version, moar printed parts ! and the nice special thing is : it's a part shared with the FoldaRap Mendel design ;)
interesting contribution from Oussama-Zaz (FabLab Ris-Orangis)
3rd prototype with scissors (video [10]
playing with different lenght of scissors
also back on the coming 3rd try for the scissor Z version :)
a simple way to clamp the belt in the extrusion
nice print during the summer (although a corner was broken, had time to reprint one after)
First print !! (30-jun-2013)
quick change head (video [11])
not foldable yet, but still easy to carry ;-)
9th prototype born in the night of the 27-28 june, small and beautiful ^^
the integrated corner-extruder-idler (maybe better to have it separated)
9th iteration coming thanks to this reduced Z-slider ;)
wondering about different arrangements, moving the XY from the bed (z-motors at the base, or with the XY), or upside-down moving only the bed from the XY at the base
first and second print (extruder not tensioned, temp. too low)
easy leveling : lock the first, adjust the second and lock it with the pneumatic fitting (using an hex key)
after a short discussion with Thomas Landrain from La Paillasse of a possible usage for 10 nozzles
mounting points dimensions
for those who prefer other kind of nozzle, plotter/j-head/prusa/etc. :)
modular assembly of x-carriage and plugins
tries at lasercutting a test rack
or a prusa nozzle, or many other eventually (easier to adapt on this horizontal mounting than on the huxley/foldarap carriage)
if only the fan could be used also to cool the printed parts
a quick and dirty belt tensioning system :P
as I have one Z-motor and two Y-motors, I switched them (and the according pins in the firmware)
or many more, just kidding :P
just a thought, with a rack you can have several carriages running on it :)
new strategy, like rack&pinion but with belts and still the printed sliders, I think I'm on something good : [28] [29] [30] (200mm moves at 500mm/s and 3000mm/s²accel, while I was stuck at 100mm/s before)
much better than the previous try ! but not perfect yet. [31]
[32] [33] 100mm/s max, linear moves are good but diagonal ones harder (I'm considering to leave the h-bot and have 2 y-motors and 1 x-motor on the carriage if I keep the printed sliders)
the electronic side will need some tidying
details of the belt tension system (as on the kosselM, one loop in on the carriage, the other goes through it)
new iteration to assemble, and parts to reprint ^^'
after the small printed pulley + igus square bushing -> printed bushing + classic metal pulley
while I still have to assemble my last iteration and three other things to try, I'm also considering a no-belt approach (maybe more suited for huge XY tables) [35]
the nice detail is to keep the leadscrew in the hole of the flange ! (an M5 nut snuggly fit inside)
the previous and current prototype (I also switched to short nema17)
printed 10t2.5 pulley (10mm diam. ^^)
may try to use a roller based carriage [36]
and the resulting 6th iteration that's coming (also with nema17 and larger XY-sliders to try to remove the remaining play that I had with linear bearing pads (see video [37][38])
igus QJFM-02-20, nice square bushings (video [39])
forget about them, metal on metal will grind the aluminium they told me
after a talk with albill I wanted to try using bearing instead of pads, but 603zz in this case, not the bulky 608 we usually see on this kind of sliders :)
an idea of a different arrangement of the compact reprappro hotend, in which the fan may serve to cool the print at the same time
the 3 main plates (base, Z, XY), the Z extrusion are stored in the base plate for transport
now we have two independant modules (the XY and the cube-Z) really usable, lets put everything together !
the blue tape work surprisingly well at holding the motor down on the plate :)
5th iteration, the bearing pad corners are good but finally the Z-stage will be better when supported from both sides (and will be more prone to be scaled a little more before going to 3 leadscrews)
need to rebuild the XY, some play have been introduced when scaling up from 200 to 300mm
I'm not fully satisfied but it's usable to try a second drawing :)
4th iteration : back to the 2 points z-stage, nema17 and belt (ready to switch to trapezoidal leadscrew)
nice try, but the nema14 are too weak for the whole z-stage at this scale ^^'
current version in test, now based on 300mm extrusion (as we already have the little FoldaRap)
current version in test, now based on 300mm extrusion (as we already have the little FoldaRap)
new strategy for the base build, the electronic plate now serve to set the distance between the extrusions to make a perfect square
need to add something to hold them on the frame like on the Foldarap 1 :)
thing to integrate : having replacable builds plate to avoid manipulating the z-stage
third Z test : printed part could work, but need a very well built frame, will probably go with the UHMW pads instead
including a first drawing !! [43]
has been recorded of course ! [44][45] (need to eliminate some z wobble)
it will be probably powered by a RUMBA board
second test for a Z-stage
manual sliding video [46]
started some early research for a logo with a friend github/com, then found something like I wanted to do on thingiverse
at a point the 20x20 extrusion will start to flex a little and we may add some additional feet for a big plotter frame, but we could already scale a machine to a nice size as is :)
with 3 leveling point this time
another way to add the bed plate
will be the second test of the Z-stage
theoretical mounting of the extruder(s)
why a z-belt ? -> instant Z-lift (eliminating the need of springs), and if it don't work swaping for a leadscrew is easy.
I hope to keep it as compact for the foldable version
new corner-motor, enclosed
the intermediate and long term versions
with a bolt long enough, we could even fix a pad directly onto an extrusion
third iteration, a little sturdier
the corner allowed some flex in the cube (since the extrusion profile are not edge to edge), but the new one is much better :)
80/20 linear bearing pads, the solution taken from the video "MechBass", if printed parts aren't good enough
I like the simplicity of using m5 rods, but I want to try a real leadscrew just to experience the difference (or a belt-driven Z stage, as in the Printxel)
some researches for the Z, sliding on the four corners to avoid the Z wobble of the gantry a la makerbot2
few printed parts, and few type of parts (central symmetry powa)
I love the modularity of using extrusions instead of laser cut sheets (300mm extrusion version, similar to an Ultimaker), this one may host 4 hotends (if you place two reprappro dual mod on each side of the carriage)
intermediate iteration to validate the H-bot part of the machine, an easily scalable cube, or just replace some corners and you can make it foldable :)
(once hooked to the Foldarap000 mainboard) and after some tries, it moves as I want to !! ^^ videos [58][59]
fortunately there is only 4 possible combinations of true/false in the firmware, just try and see which one is good
comparing the footprint with the Foldarap000 (same print surface)
and a short video to show manual moves [60] (you really need to build a h-bot to understand how it works ^^')
the pulley need a short grub screw, or it will rub against the printed part sides
the 603zz are a perfect fit
there is a 2mm gap between the profiles
starting over... the corners use much less material this time.
the sliders are a critical part, for better wear resistance they may be coated with some ptfe tape or even machined in ptfe (if printed pla wear too quickly)
we need also a belt-tensioning system, maybe in the corners (at the motors or bearings corners), but I hope the zip-tie trick will be sufficient.
for this try the shape of the scissor is very basic, maybe a shape like a "S" could be better
ideally (for long term usage) we should add a ptfe insert as an axis and between the two scissors (something like a piece of tube or a machined part)
plus two short videos : folding [62] unfolding [63]
and there is a video too [64] :)
the xy platform may serve for other light application (mini laser engraver?)
H belt path for the XY, and below the one that synchronize the four Z-scissors
Exemples of size scaling
Small
Medium
From left to right (3310, 332, 333)
XL
A huge Mondrian, using 70cm and 100cm profiles
first print : the usual open-hardware logo
let's see if it's truly scalable ^^ (note : all the plastic parts are taken from the 300mm prototype, excepted for the red square braces)
a giant Mondrian ! (70x100x100cm)
External links, media/press, etc.
- Github mentioning an "awesome" project among others ^^
More inspiration
<videoflash>qpffEF9Jew0|320|240</videoflash>
<videoflash>Hc0ADpQ6FBg|320|240</videoflash>
<videoflash>FP2ciUBT6nY|320|240</videoflash>
<videoflash>uqW3t7c3u4M|320|240</videoflash>
<videoflash>_EWfmcWAz-M|320|240</videoflash>
- Direct sliding on the aluminium frame
<videoflash>5UYMnzXQEtw|320|240</videoflash>