Hello,
First post here. There has been a thread going on for a short while on the "Delta robot 3D printers" Google Group titled "Delta mechanism variant" by Billy Zelsnack. Billy started it to explore using different simple arm mechanisms to position X,Y with a fixed Z plane. A lot of ideas were brainstormed by several contributors over the last couple of weeks. One complete concept grew out of those sessions for a Tri-Polar Bot. X,Y, and Z axis are all along radial coordinates, rather than straight. This concept does not fit the Delta definition, so one contributor (Nicholas Seward of Simpson fame), suggested that the discussions should be moved to this forum. So, here I am (Dennis Brown aka see3d) to introduce Wally (yes, it is named for a geneticist, but on a first name basis). Wally is a wall bot. Everything is referenced of a back board. Nicholas also thought that Wally would lend itself well to a RepRap design. I put together a paper model of the arms to see how they would work, and then did some schematic drawings of a side view, and a view looking up from Yoda's point of view. They are attached here:
[attachment 16179 SideZ32013-06-21at9.21.16PM.png]
[attachment 16180 TopView2013-06-22at12.09.12AM.png]
[attachment 16181 Paper32013-06-23at8.12.07PM.png]
Wally evolved out of some prior design ideas. The arms look like a simplified early Simpson prototype. The fishing line drive came from other recent Delta and Cartesian bots. The wall bot idea came from some recent Delta design brainstorming. The crazy math to translate coordinate systems was shown to work for Delta bots. I have not identified any bot crazy enough to use a 4 bar linkage for the Z yet, but that can't be new. There was a lot of cross pollination to conceive Wally.
The idea was to make Wally as simple and low cost as possible while maintaing a decent build volume and precision. Simple and build volume can be seen from the concept. Low cost is assumed due to only rotating bearings and no fancy linear actuators or precision lead screws. It will take a detailed design, a working prototype with BOM, and real 3D prints to verify the cost and precision are there.
Details about the concept:
> Built off a rigid back board. The board can have ribs attached to make it more rigid. The material is TBD based on prototype needs.
> Wally could be designed to fold fairly flat for storage or portability. Think suitcase with a handle on the side. Legs could fold out to make it free standing, or it could be attached to a wall.
> The design can be scaled. Perhaps a 12" backboard width at the small end. That might give more than a 6" cube build volume. A 15" backboard width might give something on the order of 10" Diameter, or an 8" square, or a 10" x 7" Rectangle -- all with a 10" height. I think this would be a good design center.
> All axis are driven from stepper motors through a Spectra fishing line drive on preferably smooth pulleys.
> The two X,Y arms are driven by rotating the elbows. This gives a build area that is limited by an arc from the shoulder with a straight elbow to the hot end.
> The ratio of the stepper drive to large elbow pulley determines the resolution and torque. 0.025mm or better with a reasonable ratio with a 16-32x microstep.
> The Z axis is a 4 bar linkage that keeps the bed flat as it is rotated up and down through an arc. Gravity lowers and a stepper winding a string raises it.
> The build platform moves through an arc as the Z is changed. The X,Y Arms have to compensate for the displacement in the Y axis (perpendicular to the back board).
> A convenient place for a spool is on the lower back. Filament would be fed up through a Bowden extruder and over the top and down to the hot end.
If I missed some important point, please ask about it. I am doing this summary from memory of 5 pages of posts in Billy's thread.
Other brainstorming ideas:
An end stop for both arms and possibly a table leveling Z probe might be able to be designed with just one microswitch located on the effector.
It might be possible to have a hot end tool changer. If the hot end is held with magnets on the effector, the hot end could be parked along the back wall to be picked up as needed. A strong hold magnet at the backboard could be magnetically adjusted to hold or release a tool. They would each have to have their own Bowden extruder. This would allow for multiple extrusion heads that stayed in the same center of arm rotation, and did not require any reduction in build area.
With a universal holder design, things like a Laser Diode head, a light duty rotary cutter head, a magic marker, or a clay extruder could be accommodated. To offset some of the weight of a heavy tool, a counterbalancing string could be added from above. How practical these things would be would have to be tested in a prototype.
Next steps:
The next thing needed is to have real MCAD design work done to verify all the X,Y, and Z movements and build volumes. That is not my realm.
Then some design work on the arms to make them into 3D printable form.
Then building a prototype arm assembly to test the ideas in real life.
Then likely change the design to make it work better.
Then do the same for the Z axis.
Finally put all the pieces together and put the juice to Wally to make him alive!
My contribution to Wally was to pull the various pieces together into a complete printer concept.
I would like to build Wally at some point, but need an experienced 3D bot maker to lead the way (or a group of collaborators).
I am currently waiting for delivery of my first 3D printer kit from the RigidBot KickStarter.
First post here. There has been a thread going on for a short while on the "Delta robot 3D printers" Google Group titled "Delta mechanism variant" by Billy Zelsnack. Billy started it to explore using different simple arm mechanisms to position X,Y with a fixed Z plane. A lot of ideas were brainstormed by several contributors over the last couple of weeks. One complete concept grew out of those sessions for a Tri-Polar Bot. X,Y, and Z axis are all along radial coordinates, rather than straight. This concept does not fit the Delta definition, so one contributor (Nicholas Seward of Simpson fame), suggested that the discussions should be moved to this forum. So, here I am (Dennis Brown aka see3d) to introduce Wally (yes, it is named for a geneticist, but on a first name basis). Wally is a wall bot. Everything is referenced of a back board. Nicholas also thought that Wally would lend itself well to a RepRap design. I put together a paper model of the arms to see how they would work, and then did some schematic drawings of a side view, and a view looking up from Yoda's point of view. They are attached here:
[attachment 16179 SideZ32013-06-21at9.21.16PM.png]
[attachment 16180 TopView2013-06-22at12.09.12AM.png]
[attachment 16181 Paper32013-06-23at8.12.07PM.png]
Wally evolved out of some prior design ideas. The arms look like a simplified early Simpson prototype. The fishing line drive came from other recent Delta and Cartesian bots. The wall bot idea came from some recent Delta design brainstorming. The crazy math to translate coordinate systems was shown to work for Delta bots. I have not identified any bot crazy enough to use a 4 bar linkage for the Z yet, but that can't be new. There was a lot of cross pollination to conceive Wally.
The idea was to make Wally as simple and low cost as possible while maintaing a decent build volume and precision. Simple and build volume can be seen from the concept. Low cost is assumed due to only rotating bearings and no fancy linear actuators or precision lead screws. It will take a detailed design, a working prototype with BOM, and real 3D prints to verify the cost and precision are there.
Details about the concept:
> Built off a rigid back board. The board can have ribs attached to make it more rigid. The material is TBD based on prototype needs.
> Wally could be designed to fold fairly flat for storage or portability. Think suitcase with a handle on the side. Legs could fold out to make it free standing, or it could be attached to a wall.
> The design can be scaled. Perhaps a 12" backboard width at the small end. That might give more than a 6" cube build volume. A 15" backboard width might give something on the order of 10" Diameter, or an 8" square, or a 10" x 7" Rectangle -- all with a 10" height. I think this would be a good design center.
> All axis are driven from stepper motors through a Spectra fishing line drive on preferably smooth pulleys.
> The two X,Y arms are driven by rotating the elbows. This gives a build area that is limited by an arc from the shoulder with a straight elbow to the hot end.
> The ratio of the stepper drive to large elbow pulley determines the resolution and torque. 0.025mm or better with a reasonable ratio with a 16-32x microstep.
> The Z axis is a 4 bar linkage that keeps the bed flat as it is rotated up and down through an arc. Gravity lowers and a stepper winding a string raises it.
> The build platform moves through an arc as the Z is changed. The X,Y Arms have to compensate for the displacement in the Y axis (perpendicular to the back board).
> A convenient place for a spool is on the lower back. Filament would be fed up through a Bowden extruder and over the top and down to the hot end.
If I missed some important point, please ask about it. I am doing this summary from memory of 5 pages of posts in Billy's thread.
Other brainstorming ideas:
An end stop for both arms and possibly a table leveling Z probe might be able to be designed with just one microswitch located on the effector.
It might be possible to have a hot end tool changer. If the hot end is held with magnets on the effector, the hot end could be parked along the back wall to be picked up as needed. A strong hold magnet at the backboard could be magnetically adjusted to hold or release a tool. They would each have to have their own Bowden extruder. This would allow for multiple extrusion heads that stayed in the same center of arm rotation, and did not require any reduction in build area.
With a universal holder design, things like a Laser Diode head, a light duty rotary cutter head, a magic marker, or a clay extruder could be accommodated. To offset some of the weight of a heavy tool, a counterbalancing string could be added from above. How practical these things would be would have to be tested in a prototype.
Next steps:
The next thing needed is to have real MCAD design work done to verify all the X,Y, and Z movements and build volumes. That is not my realm.
Then some design work on the arms to make them into 3D printable form.
Then building a prototype arm assembly to test the ideas in real life.
Then likely change the design to make it work better.
Then do the same for the Z axis.
Finally put all the pieces together and put the juice to Wally to make him alive!
My contribution to Wally was to pull the various pieces together into a complete printer concept.
I would like to build Wally at some point, but need an experienced 3D bot maker to lead the way (or a group of collaborators).
I am currently waiting for delivery of my first 3D printer kit from the RigidBot KickStarter.