Hybrid Polar-2d Delta Geometry - New design?

I wanted to run this idea past you guys and see if it is indeed something novel or if I just haven't found it yet.

My main concern was reducing the amount of linear rod required.

The Green circle is a standard rotary stage. I'm thinking belt driven with a continuous belt to allow 360 rotation without need for unwind.

Red and Blue blocks are travelling on 2 linear rods and are individually driven by 2 steppers. Either 2 belts or 2 leadscrews just slightly off-center for each block.
The Blue arms create a 4-bar linkage/parallelogram such that the end effector can stay horizontal regardless of position. The red arms connect to the midpoint of the bottom blue arms.
Red and Blue arms cover a 2d plane that runs through the center of the rotary stage and should be perpendicular to the plane of rotation.

Is anything glaringly wrong with this?

Should I invest in something like a smoothieboard to handle the calculations?

Thanks!

neat, yeah think that's new. Would need to look into it more but i think that would have a really great foot print size to build size ratio.

Think it should work, though you may have to place those arms kind far away from the platform and make the two carriages kinda far apart from each other. maybe not, i'd build a moch up of those arms, would be very interested to see them.

Edited 1 time(s). Last edit at 03/01/2016 04:55PM by DaGameFace.

The sketch looks pretty simple, but the real thing would need three vertical towers like a delta to achieve decent stability.
Also the linear guides and parallelogram rods would have to be free of play.
Maybe a more triangle shaped parallelogram would make a stiffer and simpler carrier. ( "Double wishbone suspension" was the term I was looking for )
You could use standard Traxxas ball ends then.

The rotating bed would be difficult to heat.

OTOH, why not take the next step and put three of those arm on one rotating bed? The collision zone would be very small ( only the center of the bed, where the hotends* could get in contact )
Each arm is controlled by it's own controller and the printed part is split in 120° pizza pies before slicing.

* New hotend would look like a triangle shaped E3D with the filament throat not in center of the heatsink, but on the very edge of it. ( Even smaller collision zone )

Edited 3 time(s). Last edit at 03/02/2016 04:26AM by o_lampe.

The build envelope is kind of a cylinder then it comes to a tip as the upper slider hits its upper limit.

@O_Lampe : I'm not really understanding what you mean by a triangular parallelogram. I didn't show it, but the upper 'rods' would actually be connected to each other, such that they don't skew. same with the bottom rods. the objective being that they don't move left-right with respect to their plane of motion.
The Blue balls in my drawing will be bearings, so their motion will be constrained to a single axis as well. I'm thinking the whole device will pre-load itself as gravity will constantly be pulling the end effector down to negate much of the backlash.

Yes, I had been thinking of a small, compact bowden-type hot end to reduce the moving mass.

In the future, I was actually thinking of doing a completely heating build enclosure and surrounding it in foam. The turntable needs a circular opening which would be easy enough to insulate. The vertical motion only needs 2 parallel slots in the wall, and the filament tube can either run along the arm or from a small hole in the top or side.

Edited 1 time(s). Last edit at 03/02/2016 12:45PM by jongscx.

Rotating beds are usually painful to get the center aligned properly. Maybe you should use a sliding bed like printrbot simple.

Your YZ axis idea is nice! I would just suggest you replace the bottom two rods by a hinge-like plate, this would make your Extruder carriage way more stable.

Quote
lhartmann
Rotating beds are usually painful to get the center aligned properly. Maybe you should use a sliding bed like printrbot simple.

Your YZ axis idea is nice! I would just suggest you replace the bottom two rods by a hinge-like plate, this would make your Extruder carriage way more stable.

When you say aligning the center properly, do you mean making sure the extruder goes directly through it? Could that be fixed with a slotted hole where the extruder attaches so it can be finely adjusted left-right?

Edited 1 time(s). Last edit at 03/02/2016 04:04PM by jongscx.

Yes, that's what I mean. A rotating bed is completely possible, just make sure to design something that allows fine-tuning of the YZ alignment.

Moving the extruder through a slotted hole sounds doable, as long as you don't move it by hand. Using a couple of fine pitched screws pushing the extruder against each other should provide reasonably good positioning.

Screw ===> extruder <=== Screw

See attachments for the triangle shaped "parallelogram". Where parallel refers to the upper and lower arms only.

Thinking through the mechanical requirements, it would be best to move the arm in z-direction with a lead screw. ( slow but strong )
Moving the effector to and fro the bed center would require a belt drive.
Maybe you can integrate this stepper and belt into the arms basement? This would save extra linear bearings and offers better accuracy.