I've been trying to document Tommelise so that others can duplicate it or even hack the design. A bit problem I've been having is finding a cad package that can actually dimension parts. I'm gritting my teeth because I see that when you all documented to Mk II the couple of drawings of things like the extruder barrel and the PTFE thermal barrier were dimensioned and were apparently drawn with QCAD.
I looked up QCAD and discovered that the "community" version of QCAD is restricted to the Linux community or to those bright enough to do a build for Java in Windows, neither description fitting me. I can get a licensed copy of QCAD for $30. Gloom.
That gets us back to another issue. In the Mk II documentation the AoI files for the Stratasys made parts were made available. Indeed, although they took forever to load they included not only the end product of boolean operations but also all of the bits that got you to that point.
At that point it struck me that if you make a design "open source" by making just the STL files available, it is still very difficult to hack the design with new ideas or for new uses. It's pretty difficult to take dimensions off of a solid that is the end result of half a dozen boolean operations. Indeed it is damned hard to even rotate or translate it to new positions unless you've been very careful how you've put it together in the first place.
Should there be some sort of standard that says that "open source" designs should be documented so that they're easy to hack?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Well, we've said it before: what we need is a good open-source CAD system. QCAD is 2D, so is OK for quick sketches, but not much use for 3D design. Then we have AoI and Blender, both of which are turning into CAD systems, but aren't there yet.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I think keeping things simple, and sharing 'small' items, will go a long way... I've had a lot of luck with AoI, but the key is keeping it simple, and keeping a 'history' of what you did...
I've scrapped everything I did before in AoI, but I've started working on the reprap version of my lego XYZ stage, and I'm keeping it simple. The first thing I've come up with is what I called a 'tear drop rod'; essentially it's just the reprap tear drop logo, put to use. I've got an AoI file with two main items, and all the pieces that went into making them. The two items: Tear Drop Rod Hole (the tear drop shape, in the form of a long rod) and the Tear Drop Rod (same thing, only I've trimmed the pointy tip off; can't make a point finer than the extrusion head, but it's still 'bigger' than a circle so the rod won't allow twisting.)
I've messed around with the involute profile gear script thingy, but I haven't actually made anything with it yet. I'd really like to get some 'standard sized' 'rack' and 'pinion' gears.
My goal here is to make/use parts in my reprap design that are generic enough that other designs will be using them as well. I'm hoping to make a standard sized 'tear drop rod' connector that's the right size for both strength in what it does, and being a standard screw/bolt size so it can be seamlessly swapped out. Similarly, I want to be able to use the same pinion gear on all my stuff; especially with the kind of wear that I expect the teeth to take, it'd be nice to have a few spares on hand, without having to have a few spares for each instance. (I'm also working on trying to figure out how I can easily replace the 'rack' on my reprap without rebuilding entire stages... Again with the consistancy and the wear on the teeth)
I don't think defining a 'standard' is going to be very productive in this sort of 'open source' atmosphere; but perhaps a 'rating' system. A number or letter or grade or something that signifies how well organized the AoI file is. A hand crafted mesh won't really have the parts that went into it, but an extruded gear might. The MkII plans did seem to have all the steps that went into them, but they weren't organized very well. I've taken my AoI files further int that I not only name all the pieces, I also organize them in a hierarchy structure to signify what went into what. I don't expect everyone to produce AoI files as nicely as mine, but I do expect everyone to be able to work with my AoI files, and make modifications/etc to them.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I was planning on picking up AOL as a first step in 'reprap for dummies', but if Blender is a good solution I might start with that instead. I've been eyeing blender for a long time but have been frustrated by the steep learning curve.
However, for the longest time I've been doing 3d work in the freely available Amapi.
Amapi is effectively abandonware and though not very stable has a very pleasant interface.
Every tool is dimensional (several dimensional settings available) and you constantly have an axis system shown while working. There's a plethora of export and import options.
Cons are low stability, no linux version (has mac though) and it being discontinued... also it's not open source.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
It appears that this French company has simply made an earlier version of their 3D package available for free. Thre is some confusion because if you click the company link on the Tucows download page for Amapi you get an American firm.
I'm extremely wary of Blender as a candidate for CAD-design.. Blender came out of a project the target of which was something like a Maya, not an AutoCAD.
I've used it & found it, well, frankly, depressing.. AoI, for all but those who know Maya or are looking for Maya-like software, seems a superior option.. albeit one that has failings too; namely, BeanShell as a scripting language, poor support for CAD-like work (which one can fix if you care for Java/Beanshell (which I don't)). Archi/Arqui is extremely immature, even if it is a project that I admire..
Obviously, if a user's RepRap project is not for commercial gain, then one can use any s/w one likes ('low price' Education or Gratis 'Testing' Editions, presuming the necessary features are enabled) for developing work for the moment.. sure, it is probably not Big Corp's desire, but who cares? One is designing for non-commercial (& educational) purposes..
Transport of files across, some future day, for use on a completely 'free software' platform is, obviously, possible...
So, everyone who is reprapping, just sign on for a community college course on the CAD software of your choice.. =D!!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I used it to create a 3D model of what our second floor would look like in order to provide it to the contractor and to get the necessary building permits. It worked great. Since that went so well, I used it to decide how we were going to lay out our basement when doing some finishing work there, too.
It is certainly possible (and not at all difficult) to create CAD models in blender. Though I haven't tried creating any RepRap models for a while, when I was playing with the spline curve algorithms for the embedded controller a few months ago, I remember thinking I was able to generate the exports needed by the reprap software.
If you wanted to try using Blender, I'd suggest going thru 2 or 3 tutorials on it to get a good basis for the commands and tools you'd need. http://en.wikibooks.org/wiki/Blender_3D:_Noob_to_Pro would probably be a good start.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I wonder if this is the time to talk about it?
I've been trying to document Tommelise so that others can duplicate it or even hack the design. A bit problem I've been having is finding a cad package that can actually dimension parts. I'm gritting my teeth because I see that when you all documented to Mk II the couple of drawings of things like the extruder barrel and the PTFE thermal barrier were dimensioned and were apparently drawn with QCAD.
I looked up QCAD and discovered that the "community" version of QCAD is restricted to the Linux community or to those bright enough to do a build for Java in Windows, neither description fitting me. I can get a licensed copy of QCAD for $30. Gloom.
That gets us back to another issue. In the Mk II documentation the AoI files for the Stratasys made parts were made available. Indeed, although they took forever to load they included not only the end product of boolean operations but also all of the bits that got you to that point.
At that point it struck me that if you make a design "open source" by making just the STL files available, it is still very difficult to hack the design with new ideas or for new uses. It's pretty difficult to take dimensions off of a solid that is the end result of half a dozen boolean operations. Indeed it is damned hard to even rotate or translate it to new positions unless you've been very careful how you've put it together in the first place.
Should there be some sort of standard that says that "open source" designs should be documented so that they're easy to hack?
Well, we've said it before: what we need is a good open-source CAD system. QCAD is 2D, so is OK for quick sketches, but not much use for 3D design. Then we have AoI and Blender, both of which are turning into CAD systems, but aren't there yet.
"Parts is parts..."
I think keeping things simple, and sharing 'small' items, will go a long way... I've had a lot of luck with AoI, but the key is keeping it simple, and keeping a 'history' of what you did...
I've scrapped everything I did before in AoI, but I've started working on the reprap version of my lego XYZ stage, and I'm keeping it simple. The first thing I've come up with is what I called a 'tear drop rod'; essentially it's just the reprap tear drop logo, put to use. I've got an AoI file with two main items, and all the pieces that went into making them. The two items: Tear Drop Rod Hole (the tear drop shape, in the form of a long rod) and the Tear Drop Rod (same thing, only I've trimmed the pointy tip off; can't make a point finer than the extrusion head, but it's still 'bigger' than a circle so the rod won't allow twisting.)
I've messed around with the involute profile gear script thingy, but I haven't actually made anything with it yet. I'd really like to get some 'standard sized' 'rack' and 'pinion' gears.
My goal here is to make/use parts in my reprap design that are generic enough that other designs will be using them as well. I'm hoping to make a standard sized 'tear drop rod' connector that's the right size for both strength in what it does, and being a standard screw/bolt size so it can be seamlessly swapped out. Similarly, I want to be able to use the same pinion gear on all my stuff; especially with the kind of wear that I expect the teeth to take, it'd be nice to have a few spares on hand, without having to have a few spares for each instance. (I'm also working on trying to figure out how I can easily replace the 'rack' on my reprap without rebuilding entire stages... Again with the consistancy and the wear on the teeth)
I don't think defining a 'standard' is going to be very productive in this sort of 'open source' atmosphere; but perhaps a 'rating' system. A number or letter or grade or something that signifies how well organized the AoI file is. A hand crafted mesh won't really have the parts that went into it, but an extruded gear might. The MkII plans did seem to have all the steps that went into them, but they weren't organized very well. I've taken my AoI files further int that I not only name all the pieces, I also organize them in a hierarchy structure to signify what went into what. I don't expect everyone to produce AoI files as nicely as mine, but I do expect everyone to be able to work with my AoI files, and make modifications/etc to them.
It appears that Blender recently got a dimensioning script. Check this out.
http://www.alienhelpdesk.com/index.php?id=38
It was contributed to the Blender effort by Render Reality, LLC as sort of a public service.
http://www.renderreality.com
The guy at alienhelpdesk also has a great little tutorial on using the DAVID 3D scanning routine as well.
http://www.alienhelpdesk.com/index.php?id=46
and...
http://www.alienhelpdesk.com/index.php?id=48
LOL! I guess I'm going to have to try out blender. That means installing Python on my dual Xeon. :-)
I was planning on picking up AOL as a first step in 'reprap for dummies', but if Blender is a good solution I might start with that instead. I've been eyeing blender for a long time but have been frustrated by the steep learning curve.
However, for the longest time I've been doing 3d work in the freely available Amapi.
Amapi is effectively abandonware and though not very stable has a very pleasant interface.
Every tool is dimensional (several dimensional settings available) and you constantly have an axis system shown while working. There's a plethora of export and import options.
Cons are low stability, no linux version (has mac though) and it being discontinued... also it's not open source.
Amapi isn't exactly abandonware. You can see their website at...
http://www.eovia.fr/
It appears that this French company has simply made an earlier version of their 3D package available for free. Thre is some confusion because if you click the company link on the Tucows download page for Amapi you get an American firm.
http://www.daz3d.com/
As best as I can figure these people used to carry Amapi in the States but have since abandoned it.
Guess I'll have to get off my ass and into 'legalspace'... Blender it is.
I'm extremely wary of Blender as a candidate for CAD-design.. Blender came out of a project the target of which was something like a Maya, not an AutoCAD.
I've used it & found it, well, frankly, depressing.. AoI, for all but those who know Maya or are looking for Maya-like software, seems a superior option.. albeit one that has failings too; namely, BeanShell as a scripting language, poor support for CAD-like work (which one can fix if you care for Java/Beanshell (which I don't)). Archi/Arqui is extremely immature, even if it is a project that I admire..
Obviously, if a user's RepRap project is not for commercial gain, then one can use any s/w one likes ('low price' Education or Gratis 'Testing' Editions, presuming the necessary features are enabled) for developing work for the moment.. sure, it is probably not Big Corp's desire, but who cares? One is designing for non-commercial (& educational) purposes..
Transport of files across, some future day, for use on a completely 'free software' platform is, obviously, possible...
So, everyone who is reprapping, just sign on for a community college course on the CAD software of your choice.. =D!!
I've used blender for about a year or two now.
I used it to create a 3D model of what our second floor would look like in order to provide it to the contractor and to get the necessary building permits. It worked great. Since that went so well, I used it to decide how we were going to lay out our basement when doing some finishing work there, too.
It is certainly possible (and not at all difficult) to create CAD models in blender. Though I haven't tried creating any RepRap models for a while, when I was playing with the spline curve algorithms for the embedded controller a few months ago, I remember thinking I was able to generate the exports needed by the reprap software.
If you wanted to try using Blender, I'd suggest going thru 2 or 3 tutorials on it to get a good basis for the commands and tools you'd need. http://en.wikibooks.org/wiki/Blender_3D:_Noob_to_Pro would probably be a good start.