Open Source Ecology - User contributions [en]

Sketchup

2013-06-13T17:19:54Z

Jake Boening: /* OSE Case */

SketchUp is a 3D modelling program for a broad range of applications such as architectural, civil, mechanical, film as well as video game design — and available in free as well as 'professional' versions.

* http://en.wikipedia.org/wiki/SketchUp
* http://www.sketchup.com/

=OSE Case=
While Sketchup is not open source - meaning - one cannot simply take the source code and make improvements or add features as needed - it is free and widely accessible around the world. The wide accessibility allows wide-scale collaboration to happen. The problem that OSE had in the past with respect to open collaboration on CAD - was that if a user created a design in a certain CAD package, most other collaborators were not able to open it because they did not have access to the same software. This problem ran deep - because even if people did have the same software, and it was created on an older or newer version of the program - the CAD file would still not open properly.

The most common issues regarding widenspread collaboration on CAD that OSE faces are:

#'''Lack of Software''' - A collaborator does not have software to open or view a file
#'''Cross-Platform Compatibility Issues''' - CAD Software does not exist on all platforms
#'''Backwards Compatibility Issues''' - CAD file does not work with older versions of same software

One route to addressing this is using Sketchup. While not open source, it is easy to use and it is widely accessible. The solution path for OSE is to create an open source equivalent to Sketchup.

Even though Sketchup is less powerful than fully-featured CAD programs, OSE claims that it is more important to have wide access be the main selection criterion for the CAD software choice - because that allows development to happen faster. We have seen this in practice: many collaborators working piecemeal on Sketchup create design faster than a single dedicated designer on a professional package. For example, it is easier for 10 people to create fabrication drawings via Sketchup than for 1 person to use a professional CAD package which has a dedicated Fab Drawings generator. Further, recent research by DARPA on collaborative CAD work has shown that N designers can generate a design in a time less than 1/N the time it would take for a single designer to generate the same design. There is empirical evidence that collaborative design takes less cumulative time than one-person design.

Therefore, it is encouraged that any OSE collaborator becomes familiar with Sketchup. Our team will share designs via Sketchup as much as possible. Basic literacy in Sketchup - in terms of being able to open and navigate designs at the very least - is required to be an effective member of the OSE development team. Most of the existing 3D designs of OSE - such as Ironworker Machine, tractor, brick press, pulverizer, Power Cube, house designs, brick rollers - are available in Sketchup.

This does not preclude the use of other packages, such as [[FreeCAD]], and many designs for OSE have been done in FreeCAD already. The choice of the preferred OSE CAD package may change as the open source CAD solution landscape changes with time. Sketchup is the generally-used common platform of OSE in the meantime.

=Further Discussion=
[[FreeCAD]] is another option, but FreeCAD currently does not allow for the creation of assemblies or fabrication drawings. FreeCAD has a more difficult to use interface. Sketchup boasts the ability to create simple shapes, extrude them, and modify corners, adges, and faces. OSE is currently considering collaboration with [[Sunglass.io]] to develop an open source equivalent to sketchup. Why not start with FreeCAD? According to [[Kaustuv Biswas]] of Sunglass, the software architecture of standard packages such as FreeCAD, Autodesk, Solidworks, and others - comes from a paradigm based in the 1970's - where a powerful and complicated CAD kernel was reqiured. Today, Kaustuv states that we can design a much lighter CAD kernel - and make all the heavy analysis and computation on CAD models happen via add-on modules. Thus, the design of the modern CAD platform can be much more modular and lightweight, and that would be the motivation for creating an open source equivalent of Sketchup from scratch - while building on the same open source modules, as needed, that the more complicated packages use.

=Links=
*Sketchup in Wine - [http://fablabtruck.nl/nl/projecten/138-sketchup.html]
*[[Sketchup_on_Linux]]
**Exporting STL files from Sketchup running in wine - [http://fablabtruck.nl/nl/projecten/138-sketchup.html]
*Makerbot on Sketchup in Wine - [http://wiki.makerbot.com/google-sketchup]
*[[Sketch Up Tutorial]]
*[[Sketchup Cube Test]]
[[Category:Software]]

HydraFabber

2013-06-13T17:11:39Z

Jake Boening: /* Functional Spec Questions */

See description of HydraFabber in this video -

<html><iframe width="560" height="315" src="https://www.youtube.com/embed//EXau4L6Ghk8" frameborder="0" allowfullscreen></iframe></html>

HydraFabber - after [[HydraRaptor]] - is a portable, multi-headed CNC fabrication machine that can do 3D printing, circuit milling, and laser cutting. It is intended to be a portable machine that can be brought to [[OSE Design Sprint]]s for prototyping. During the sprints, we design machines in Sketchup, and then prototype them by printing out modular parts with a 3D printer and laser cutter. The laser cutter cuts out flat sheets that are folded to make 3D beams and other parts, and these beams form 3D frames and other 3D parts. Then the circuit mill functionality can be used to prototype circuit boards, such as controllers for automated machines.

=Specifications=

==Functional Spec==

* 200x200x120mm mininimum build volume
* Quick change heads with accurate repeatable alignment
* Complete system includes robust carry case
* At most 10 minutes for a competent user to set up or prepare for travel

* Printer
** One material at a time
** heated bed to at least 140C
** reliable extruder
** hot end to at least 245C
** Glass print surface for flatness
*** Optional tape over glass for ABS adhesion
** Fan to cool print
** at least 60mm/s practical print speed (more is better)

* Laser
** 1.5W laser cutter diode

* Circuit mill
** depth of cut control relative to surface of PCB
** easy tool change
** effective quick change PCB holddown

* Folds up into a portable suitcase so it can be brought to design sprints and other events

===Functional Spec Questions===
'''General Comment: study of open source industry standards is our best bet. It appears the 3D printer part is all go. Shapeoko has sold 1500 of their routers - so let's assume they work well enough. Search for laser diode paper cutters on the web. Combine RepRap, Shapeoko - and come up with a design. I could foresee that the Router Module is just 2 reinforcement plates for the sides of the machine to provide rigidity. Perhaps use Shapeoko as base, add a 3D printer head or laser diode head to convert to the other 2 functions? Then we have structure all worked out. Does Shapeoko appear to get us the right accuracy?'''

* Is 1.5w optimum? '''See discussion at [[SLS_Wax_Printer]]. Requirement is anything that cuts paper.'''
* What wavelength? '''That which cuts paper, may have to work with black paper'''
* What laser radiation safety measures (1.5w is plenty to blind)? '''Good question. Let's research industry standards'''
* Fixed laser + moving optics or moving laser? '''Replacement for extruder head on the triple machine'''
* What feed rates would be ideal? '''Need to look at specs of laser diode or other projects'''
* How accurate do we need to be for circuit milling?'''Same performance as Shapeoko'''
* What is our target track pitch? '''What is this'''
** Are we aiming for through hole or SMT? If SMT, then how small (just SOICs or QFNs and BGAs)? '''Same as Shapeoko'''
* Is a Dremel-type spindle good enough? '''Appears so'''
** Noise? '''What kind of noise?'''
** Speed (RPM)? '''5,000 - 35,000'''
** Accuracy (TIR)?
** Mass - is it too heavy?
** Size - is it too big? '''No.'''
* Are there any reasonable off the shelf alternatives to the dremmel? '''Don't know. First prototype with dremel appears sound. Based on Shapeoko.'''
* How hard/expensive would it be to make a good small open source spindle? '''I wouldn't worry about it in the first iteration.'''
* What feed rates would be ideal? '''See Shapeoko'''
* What cutting forces are we going to get? '''I would say 2 lbs of force'''
* How do we hold down the PCB? '''Paper Clamps in first version?'''
* Will the dust from the PCB milling interfere with the 3D printing? '''We will have to get data on that.'''

=Process=

== Quick Route to Functional Prototype ==

Almost any RepRap will fulfil the printer part of the functional spec. The need for quick-change heads without recalibration requires a new head mount that will position the head very repeatably each time it's mounted. A kinematic mount is under development for this purpose.

https://github.com/leodearden/kinematic-mount

[[File:2013-06-09-kinematic-mount-scad-03.jpeg]]
[[File:2013-06-09-kinematic-mount-scad-04.jpeg]]
[[File:2013-06-09-kinematic-mount-scad-05.jpeg]]

The more rigid RepRap the better. A Mendel90, MendelMax, ORD Bot, or TAZ would be particularly suitable.

== Complete design process ==

Start with Window 7, presentation page 2 - concept design.

#Do [[Systems Engineering Breakdown Diagram]]

* Software
** CAD
** CAM/slicing
** Device control
** firmware
* Control Electronics
** Embedded microcontroller
** I/O
*** motor drivers
*** high current digital drivers (fans, heaters, etc)
*** digital inputs (eg: limit switches)
*** analog inputs (eg: thermistors)
* 3 axis positioning robot
** X
** Y
** Z
** Quick change head mount
*** mechanical
*** electrical/electronic
**** ID for head available to controller
** Quick change bed (reversible?)
** fold down mechanism
* heads
** print
** laser
** mill
* beds
** print
** laser
** mill

Diagram to follow.

#Potentially do an interface design that shows how modules fit together
#Then embed the original i3 RepRap STL via an [[STL Viewer]] in the wiki

=Printer Reliability=
*Leo's RepRap Kit had at best 4/5 print success on 8 hour overnight prints. Drift in mechanical calibration, failure of print adhesion, nozzle blockage, filament feed slip, and filament spool tangles are the main cause of failures in long prints.
*For short prints, once print parameters are established for a particular object for a particular printer with a particular material. Once you are dialed in, ~1% failure rate for prints of 1 hour. This is for unambitious - ie, safe zone - prints. Wall thickness safe, fairly thick layers, not too fast, etc.

=Links=
*[[CNC Circuit Mill Spindle]]
*[[Leo Dearden]]
*Opencreators Korean printer - [http://cafe.naver.com/makerfac]
*[[Suitcase 3D Printer]]
*[[Lulzbot]]
*[[RepRap Kit]]
*[RepRap Prusa i3|http://reprap.org/wiki/Prusa_i3]

HydraFabber

2013-06-13T17:11:09Z

Jake Boening: /* Functional Spec Questions */

See description of HydraFabber in this video -

<html><iframe width="560" height="315" src="https://www.youtube.com/embed//EXau4L6Ghk8" frameborder="0" allowfullscreen></iframe></html>

HydraFabber - after [[HydraRaptor]] - is a portable, multi-headed CNC fabrication machine that can do 3D printing, circuit milling, and laser cutting. It is intended to be a portable machine that can be brought to [[OSE Design Sprint]]s for prototyping. During the sprints, we design machines in Sketchup, and then prototype them by printing out modular parts with a 3D printer and laser cutter. The laser cutter cuts out flat sheets that are folded to make 3D beams and other parts, and these beams form 3D frames and other 3D parts. Then the circuit mill functionality can be used to prototype circuit boards, such as controllers for automated machines.

=Specifications=

==Functional Spec==

* 200x200x120mm mininimum build volume
* Quick change heads with accurate repeatable alignment
* Complete system includes robust carry case
* At most 10 minutes for a competent user to set up or prepare for travel

* Printer
** One material at a time
** heated bed to at least 140C
** reliable extruder
** hot end to at least 245C
** Glass print surface for flatness
*** Optional tape over glass for ABS adhesion
** Fan to cool print
** at least 60mm/s practical print speed (more is better)

* Laser
** 1.5W laser cutter diode

* Circuit mill
** depth of cut control relative to surface of PCB
** easy tool change
** effective quick change PCB holddown

* Folds up into a portable suitcase so it can be brought to design sprints and other events

===Functional Spec Questions===
'''General Comment: study of open source industry standards is our best bet. It appears the 3D printer part is all go. Shapeoko has sold 1500 of their routers - so let's assume they work well enough. Search for laser diode paper cutters on the web. Combine RepRap, Shapeoko - and come up with a design. I could foresee that the Router Module is just 2 reinforcement plates for the sides of the machine to provide rigidity. Perhaps use Shapeoko as base, add a 3D printer head or laser diode head to convert to the other 2 functions? Then we have structure all worked out. Does Shapeoko appear to get us the right accuracy?'''

* Is 1.5w optimum? '''See discussion at [[SLS_Wax_Printer]]. Requirement is anything that cuts paper.'''
* What wavelength? '''That which cuts paper, may have to work with black paper'''
* What laser radiation safety measures (1.5w is plenty to blind)? '''Good question. Let's research industry standards'''
* Fixed laser + moving optics or moving laser? '''Replacement for extruder head on the triple machine'''
* What feed rates would be ideal? '''Need to look at specs of laser diode or other projects'''
* How accurate do we need to be for circuit milling?'''Same performance as Shapeoko'''
* What is our target track pitch? '''What is this'''
** Are we aiming for through hole or SMT? If SMT, then how small (just SOICs or QFNs and BGAs)? '''Same as Shapeoko'''
* Is a dremmel type spindle good enough? '''Appears so'''
** Noise? '''What kind of noise?'''
** Speed (RPM)? '''5,000 - 35,000'''
** Accuracy (TIR)?
** Mass - is it too heavy?
** Size - is it too big? '''No.'''
* Are there any reasonable off the shelf alternatives to the dremmel? '''Don't know. First prototype with dremel appears sound. Based on Shapeoko.'''
* How hard/expensive would it be to make a good small open source spindle? '''I wouldn't worry about it in the first iteration.'''
* What feed rates would be ideal? '''See Shapeoko'''
* What cutting forces are we going to get? '''I would say 2 lbs of force'''
* How do we hold down the PCB? '''Paper Clamps in first version?'''
* Will the dust from the PCB milling interfere with the 3D printing? '''We will have to get data on that.'''

=Process=

== Quick Route to Functional Prototype ==

Almost any RepRap will fulfil the printer part of the functional spec. The need for quick-change heads without recalibration requires a new head mount that will position the head very repeatably each time it's mounted. A kinematic mount is under development for this purpose.

https://github.com/leodearden/kinematic-mount

[[File:2013-06-09-kinematic-mount-scad-03.jpeg]]
[[File:2013-06-09-kinematic-mount-scad-04.jpeg]]
[[File:2013-06-09-kinematic-mount-scad-05.jpeg]]

The more rigid RepRap the better. A Mendel90, MendelMax, ORD Bot, or TAZ would be particularly suitable.

== Complete design process ==

Start with Window 7, presentation page 2 - concept design.

#Do [[Systems Engineering Breakdown Diagram]]

* Software
** CAD
** CAM/slicing
** Device control
** firmware
* Control Electronics
** Embedded microcontroller
** I/O
*** motor drivers
*** high current digital drivers (fans, heaters, etc)
*** digital inputs (eg: limit switches)
*** analog inputs (eg: thermistors)
* 3 axis positioning robot
** X
** Y
** Z
** Quick change head mount
*** mechanical
*** electrical/electronic
**** ID for head available to controller
** Quick change bed (reversible?)
** fold down mechanism
* heads
** print
** laser
** mill
* beds
** print
** laser
** mill

Diagram to follow.

#Potentially do an interface design that shows how modules fit together
#Then embed the original i3 RepRap STL via an [[STL Viewer]] in the wiki

=Printer Reliability=
*Leo's RepRap Kit had at best 4/5 print success on 8 hour overnight prints. Drift in mechanical calibration, failure of print adhesion, nozzle blockage, filament feed slip, and filament spool tangles are the main cause of failures in long prints.
*For short prints, once print parameters are established for a particular object for a particular printer with a particular material. Once you are dialed in, ~1% failure rate for prints of 1 hour. This is for unambitious - ie, safe zone - prints. Wall thickness safe, fairly thick layers, not too fast, etc.

=Links=
*[[CNC Circuit Mill Spindle]]
*[[Leo Dearden]]
*Opencreators Korean printer - [http://cafe.naver.com/makerfac]
*[[Suitcase 3D Printer]]
*[[Lulzbot]]
*[[RepRap Kit]]
*[RepRap Prusa i3|http://reprap.org/wiki/Prusa_i3]