From RepRap
Revision as of 20:01, 21 September 2012 by VikOlliver (talk | contribs) (Other Printed Parts)
Jump to: navigation, search
Crystal Clear action run.png
RepRap Out Of FabLab (RROOFL)

Release status: Development

Rroofl axes complete.jpg
A Mendel variant designed specifically to be made in a FabLab
CAD Models
External Link

"Don't like society? Print a new one." - Vik Olliver http://diamondage.co.nz


This RepRap/RepStrap variant is designed to be produced in a FabLab, and yet to be sufficiently modular to produce using a variety of local manufacturing processes. It is rather unfriendly to the lone user in its present configuration, as it (a) will make use of all the available CNC-type devices in the workshop in parallel and (b) doesn't work yet. It sprang out of a 2-day session organised by User:VikOlliver and about a dozen hard-working volunteers at FabLab8 in Wellington, New Zealand. They are invited to add their names to the list when they find out about this page :)

Rroofl extruder clearance issues.jpg
Actual build volume currently envisaged 230mm x 230mm x 150mm. In reality there are some collision issues limiting X movement by 50mm or so above Z heights of more than about 60mm. The top of the arch needs to be thinner, the void in the side panels more generous at the top. Perhaps bringing the entire control panel (and thus stretching the void) towards the nearest narrow end?

In the picture left, the extruder gear nearly colliding with the trapezoidal hole on the upper left. That whole side needs to be moved over, and the top M8 rod closest to it removed. If we then move the remaining rod further away, we should be able to let quite a large extruder rise up into the central gap.

Design Objectives

  • Fabricate mechanical hardware in 1 person-day.
  • Completely Open Source
  • As many parts as possible to be able to use more than one fabrication technique.
  • All routing cuts at least 6mm wide.
  • No little extra bits to bolt on - minimize part count.
  • If you do use bolts and threaded rod, either M3 (or 1/8") or M8 (or 5/16") please.
  • If a zip tie or woodscrew does the job instead of 8 screws, 4 nuts and 16 washers, use one.
  • Common parts only.
  • Allow for lots of variation in hand-cut parts like the rods.
  • Modular design that can be mass-produced, and modules used in other projects.
  • Control systems capable of being scaled to drive other workshop equipment.
  • Runs off 12-14V DC to be friendly to small, local, power generation systems.
  • There is but One True Bearing: the 608Z "skateboard" bearing.

Current Status

Rroofl axes complete.jpg
Rroofl mechanical assy1.jpg
Rroofl awaiting x axis.jpg
X Axis uses the carriage from Simpleton prototype slightly hacked by a nice lady at Fab8nz, a NEMA17 and 3 x 608 bearings as corner pulleys. Tension in belt will draw under-constrained Z sliders onto Z guide rods. The unconstrained sliding portion of the carriage needs to clip more firmly onto the X rail and have slots at the idler end to allow the X rails to be pinned in place. End stop microswitch held in place by a single 2mm zip tie.

Y axis fully assembled. 2mm zip tie used to hold end stop switch in place. Uses 5mm zip ties in the frame to hold both NEMA17 drive motor and the idler. Needs to be measured and the cutting files updated.

Z Assembled, currently using non-spec. gears. Sockets for 608 bearings were not deep enough, the waterjet cut holes in the frame to support Z brackets need to be +0.5mm diameter. The Z motor bracket needs to allow the motor to drop down an extra 3mm to mesh properly with the large driven gear.

Yes, we know we'll get more even carriage movement with the motor driving the trans-axle direct. This is being bravely sorted out by Hamish.

There is no complete Z endstop in the design at present. Hamish built into the Z motor bracket some form of adjustable microswitch holder integrated with it. It holds the Z endstop statically on the Z motor bracket, and add a length of M8 screw thread to the underside of the X motor bracket. The latter pokes the former when the X carriage descends too far. Vik is designing this.

The deposition bed is a simple piece of plastic sheet stuck down with double-sided foam, so it needs no fancy holes. Y carriage is in 2 identical parts joined with M8 rod. Y belt secured to carriage with zip ties, which also tension it. It is important to clean the bed's underside before sticking it down.

It looks like we can drop all the holes in the lower part of the frame (and the top edge of both short ends) by about 15mm to increase our build height a bit. If we take another 5mm off the top edges we can claim that back for build height by dropping the height of the Y carriage a bit.

The circuit holding panel needs to move up and away from the build area, and needs a hole matrix for mounting Arduinos etc. on.


Prototype rroofl electronics.jpg
Electronics are assembled and working well. Stepper motors are Sparkfun NEMA17's and some beefier ones scrounged from Massey University, Wellington. Drivers are a mix of EasyDriver V4's and Makerbot V3.3's and connect to an Arduino 1280 Mega using mostly commercial jumper leads. Connectors and lead lengths are a nightmare, so we need to rationalise this into something that suits the inventory.

There are no fixed or special electronics, though they must be designed to run of car batteries and donkey-powered generators etc. Not only does this make the printer capable of running on homebrew power, it ensures that they will continue to run in the event of natural disasters. Just put an automotive electrical lighter plug on the power lead and you can run it in the car. Do remember that the spring goes under the fuse in those things, not on top of it (otherwise the spring melts).

One option being explored here is the use of an inexpensive breadboarding panel as a universal parts interface and bodging board. No soldering required, it encourages tinkering, and simplifies things when you blow a pin on your Arduino - just move the connector and recompile. OTHO a cat can wreak havoc with it. The Du Pont jumper/connectors seem resistant to normal vibrations and movement.

Zero endstops are all implemented with microswitches. a 10K resistor pulls the relevant input low, closing the switch makes it go high. There are no max limit switches.

Stepper motor STEP & DIR pins must not be longer than about 200mm.

When cable paths are established, we need mounting holes for cable ties so we can tidy them up.

A stepper motor module capable of driving 2A and made from FabLab inventory is being designed. Hopefully with sane connectors!

Spiral cable tidy is one option for constraining the cabling, and it is at least reusable.


Please link or add your files here, folks.

Guide to the filenames and intended arrangement of the Z-axis

Note that the Y axis brackets are replaced with zip ties in the prototype. If it works, why not? It looks likely that we will need to brace the frame across the top. In the prototype this is done with M8 threaded rod.

File:Rroofl Z-axis Assembly-Blender257.zip The archive included here contains the assembly model for checking gear meshes and interference. This model has been used to create the STLs listed on the image. An option for the Z-axis driven end is included. Stand-in geometry for switches, motors, and gears have also been used to help in the design process. - updated 15th September 2012 by User:BouncyMonkey

File:Rroofl Z-axis STLs.zip The image above will guide you to the appropriate model contained within this archive. - updated 15th September 2012 by User:BouncyMonkey

Note that the tops of the M8 Z drive screws should not be overly constrained, so the bearings allowed for will seldom be necessary.

Other Printed Parts

Rroofl y carriage complete.jpg
File:Y bed runners.scad The Y runners. It would be great if someone could vector this to cut on a laser Print two of these and join with a couple of 140mm lengths of M8 studding, nuts and washers. Ease the frame into shape on the rails before fully tightening. Y Slider legs can be compressed with a bolt or zip ties if it is necessary to level the bed. The actual deposition bed is attached by using small double-sided foam adhesive pads.

Rroofl y carriage belt.jpg
The Y belt is attached with zip ties; Make loops in the belt ends and secure with zip ties. Thread another zip tie through the small holes in one side of the leg of the Y carriage, pass it through the loop in the belt and then use to tension the belt to your requirements. Note that if you leave a large loop of zip tie this introduces unwanted springyness, so you'll have to cinch it with a zip tie or two.

(The following need an update 'cos the Z guide rail and drive screw spacings changed) - (15th September 2012) Hopefully the new Z-axis models above have returned the guide rail and drive screw spacing to their original distances.

File:Rroofl.inc Common measurements. Based on Simpleton's include file so lots of crap in it.

File:X axis bits.scad OpenSCAD file for the x-axis parts. Needs Rroofl.inc.

File:Printed rroofl stls.zip Contains printable STLs used in the prototype X, Y & Z axis.

Rroofl mendel prusa adaptor.jpg
File:Rroofl mendel adap.scad

Basic Rules of Engagement

Any rules or warnings that should be established before using the Rroofl can be included here. For example, a safety related tip would definitely make a good addition to this section (thanks for the reminder, OHM).

  • Do touch, but touch responsibly.
  • During assembly, look out for sharp bits of wire in the edge of the belt. They hurt.
  • Do not ever unplug a stepper motor with the power on. It kills the driver.

Ongoing Development

Pronounced roofle, like rifle with an oo in it.

Tell me again why I have two M8 rods running across the top???

Well, if I recall they are a hang-over from the original edition we were using as a basis. While they give some good rigidity to the top end of the frame, the plate for the electronics will help give that, and allow us to remove at least one of them. The top end z-axis bearing holders could be held on by a pair of M8 bolts each rather than the threaded rods. One thing that springs to my mind is that depending on how far we are able to get the x-carriage to travel upwards we may have interference between the object being printed and the electronics back-board. In my view it is probably worthwhile redesigning the end panels of the frame to be square. This would allow us plenty of room for the x-carriage, the printed piece, and still give us the opportunity to give it a good firm structure. I could talk to the local FabLab and see if we can wheedle them into cutting the redesigned ends in that so nice and space-agey looking acrylic they gave us in the conference.

Use a zip tie looped over each of the rod's idler ends to allow the rods to be held in permanent compression. Otherise the Z axis will bind.

Finished X carriage belt post needs to be longer by about 3mm.

Sculpt lower edges slightly to reduce contact with bench.

Rroofl first light.jpg
This is how it started at Massey University's FabLab during Fab8nz.

See Also