first thoughts on RepRap, and some questions

Hi all

First of all... RepRap is a very inspiring project!! I'm trying to resist buying parts right now, until I actually know how to build it

I have a very fundamental question before I get into this, though, and that is... why do most of these things go for a flatbed-plotter approach? Especially for fabbers, rather than 3d printers, wouldn't it make more sense to have a 6-dof robotic arm, that can just pick up tools from a set, point them at any angle, and use them? Wouldn't it make more sense to have an arm that can squeeze a glue-gun type tool to create shapes, then pick up a chisel and route out channels, etc.?

Secondly, I just want to also register my interest in a kit: I'm mostly a programmer/sysadmin, with an interest in 3d modelling. The electronics of printing PCBs and soldering in all the components is a bit beyond my skills, I think, even though I do own a soldering iron, and was introduced to one early in life!

Welcome aboard!

To answer your question about cartesian robot versus robotic arms:
I think this is a decision that has been taken on the grounds that a cartesian robot is a lot easier to build from scratch than a robotic arm with 6 degrees of freedom.

It may be true that a 6dof robot is a lot more versatile and works well over several orders of magnitude in scale as opposed to a cartesian robot and offers a solution to solve the crucial issue of positioning accuracy and the usage of multiple heads, but it is also true that it does this at a very high price at this moment in time.
If we can get people aboard that can tackle the issue of designing a 6dof robot at a low cost, it would be absolutely fantastic and embraced by many of us in a heartbeat.

But for now, the balance of the accuracy/cost equation pends towards the 3d cartesian robot.

Most machines of this type are cartesian machines. While there is a lot of flexibility in a 6DOF robotic arm, using the same level of technology, there will also be at least a magnitude increase in power requirements and a magnitude decrease in positioning accuracy and at least doubling the cost in materials alone. Having to suspend motors out on the arm components, plus the overhanging hardware will put a lot of strain on both the drives and the mechanics. The inverse kinematics required to position the head accurately are not insignificant. In industry, for instance, only recently have they started using robotic arms for milling very special complex shapes. It's been normal cartesian industrial mills with special rotary heads that have done this traditionally, and still do most of it, due to lower cost, higher speed, and higher accuracy for a given price.

Another reason for not doing a 6DOF arm is that this is an additive project. The part you are building will either stand on its own or need support structure no matter what robot system you use. Going back to your example, it's as simple as designing in the channel in the first place and will actually be stronger given the layered design. If we want to intermix in other materials, it'll be no harder to add a head changer to the Darwin design than it would be to make one for a 6DOF robot.

Now, I do hope to use the RepRap to build the structural components of a small robotic arm eventually, but we have to get the basics working first.

Yeah, I can understand it being much harder to construct. Fair enough -- I'll look forward to seeing the object library entry for a robot arm someday in the future then

For now, I'll get back to figuring out the basics of this Thanks for the welcome!

hey, welcome to the project!

the others touched the 6DOF idea, but let me assure you we are working on the kit situation as fast as possible. we will definitely have pre-printed boards available, and soldering them up is quite easy. the soldermask + silkscreen included makes it a pretty simple task.