Polar coordinate programming

I have designed a printer that uses a combo polar/linear XY mechanism and a conventional Z axis all on a Delta chassis design. I have the option of moving the bed on the Z axis, or leaving it fixed as per a conventional delta, and letting the XY mechanism travel along the Z legs. Either case the bed doesn't rotate, the XY rotates per the drawing. Since this will be an enclosed chamber printer, I'm leaning toward moving the bed and keeping the XY stationary so drive motors and components are fixed outside the heat chamber.

My question to the group stems from my lack of knowledge on polar coordinate plotting and integration of that into the conventional 3D print process chain. Is anyone familiar with a software package that exists to manage these movements that can be configured for Smoothieware? I have the Smoothieboard hardware already so I'd like to stick with that.

The mechanical design allows end to end movement of the print head through the 0,0 center coordinate, but I'm thinking some fixed distance past center is really all that is needed to give full print coverage across the bed. Any thoughts on that? Any other issues come to mind?

Thanks for your feedback...any suggestions or critique is appreciated. TP

Your design has nothing like a DELTA

X linear (mm), theta (polar rotation (deg or Radians), Z linear (mm)

How do you plan to rotate your X gantry, seems like a very complex rotary frame


Do you intend to use a normal slicer or are you going to code a custom slicer?

I don't know of firmware that will do what you want --- ready to custom code -- or heavily change Marlin?
Do you plan a pre processor to generate move commands
If you intend to do it in real time --- plan on a fast processor on your control board.

What are the advantages of this design over X,Y,Z or delta design?

Thanks (I guess) on your input Ray... Seems strange that your first comment is decidedly negative considering that you know nothing about the design other than what I offered. Considering also that the Polar XY mechanism described above, will bolt directly to the large delta chassis I originally designed for Berrybot hardware I purchased, the commonality between this and a conventional delta design is significant whether you realize it or not.

Beyond that negativity...I intend to use whatever editing/printing software tools I can make arrangements to locate or commission that will do the job. Hence the point of my original post.

The advantages....? Reduced cost, increased print volume, broad scalability, and a distinct price point advantage given the totality of available features over existing market offerings.

Thanks again....TP.

I am sorry
I am an old geezer and will never be politically correct,
and say what I see based on 30 years of engineering and 64 years of life.
Yes I know nothing about your design except what you presented in your attached diagram.
with that you asked "Any thoughts on that? Any other issues come to mind?"

My idea of a delta is like this



Please show me how your design is a Delta.

How your gantry is attached to delta arms

A delta uses three vertical drives to steer conjoined arms to position extruder in X, Y, Z

As I see your design --A X gantry rotating around
using distance from center and rotation angle as a positioning system
but I suppose I don't have enough information provided to comment!

NP Ray... We are the same age so I know the feeling. I appreciate your comments.

I'm using a delta chassis; but not the arms and effector. The frame uses OpenBuilds V-Slot vertical 6020 X 1500mm extrusions much as you've shown, with hexagon rather than triangle 6020 V-Slot horizontals at the top and bottom . The six 60° outside corners are 2020 extrusions I bought from Musimi. I'm using powder coated MDF between all the risers for the infill panels and to make the structure rigid. I plan to use 6mm clear Lexan for the full height door, framed with 4020 V-Slot extrusions. All that remains as I originally planned for delta printer hardware I bought for that design.

I'm departing from a conventional delta on this design, and replacing the effector and arms with the polar XY rotational assembly. That assembly will be CNC machined and put together as a self-contained unit that bolts to the top hexagon horizontals. Both X and Y axis are rotational, and driven by opposing Nema 17s using 6mm interlocking belts for the X, and 9mm interlocking belts for the Y. Distance, angle, and speed are the XY positioning variables. The Z axis will consist of a 2020 extrusion hexagon platform attached to the three 6020 risers. The 600mm diameter heat bed will bolt to the top of that. I plan to use one or more Nema 23s for the Z axis wired parallel to external stepper controller(s). I'm still debating on the best way to mechanically connect the 3 riser carriages to the motor(s).

I have what I need to build, but I want to find someone to work with me to put the polar coordinate firmware package together before I do. Arthur Wolf over at Smoothieware offered his assistance so hopefully we can work that out soon. If not I need to find someone else with the appropriate expertise that can. Thanks...TP.

Here are some drawings of my design for perspective...

My first comment stands -- NOT a delta!
Mechanism geometry and control code not close to delta.

What will the red platform rotate on? What will rotate it.

Closest I have seen to it is rotating a dome on an astronomical observatory
bunch of golf balls in a ring and wheel friction drive
Not the accuracy you'll need.

From what I have seen and done with polar coodinates
I would replace your design with a Core XY type mechanism in the hexagon

But if you must,

Look at
[forums.reprap.org]

Look at:
Polar and Cartesian Coordinates Using G15 and G16
[www.cnccookbook.com]

Check out Concept Forge
[forum.conceptforge.org]

Nick Seward really good at geometry calculations

Ok...thanks for the links. I'll follow up with those and see if I can track those guys down from 2010. BTW....I said in my original post I was using a delta chassis...I am and it is. I think I mentioned I'm using motors and belts. The platform will ride on bearings machined into the housing for precision spacing and fit. Thanks...TP.

I am also interested in this approach but have not gone past thinking. There needs to another name for this geometry other than polar. The math is pretty simple but the arduino suffers when much computation is required. That is why most serious delta users have gone with 32 bit controllers such a duet and smoothy. The 32 bit software was based on marlin but was completely refactored. You should be able to add other geometries although I don't know how. I have a duet but have done nothing with it.

Would suggest doing a search on ebay for lazy susan bearings. VXB has very nice 17" ones with several hundred pound capacity for less than $75. I am sure that these are intended for more than a lazy susan on a table.

Roger

Have a look at these designs: http://www.thingiverse.com/thing:128700 and http://reprap.org/wiki/R_360

Looks like the same approach you like to take.

Some firmware for this type of printer: https://github.com/kory75/Marlin_360

Note that these printers suffers from issues around the center of the circle.

Good luck with your build!

Edited 1 time(s). Last edit at 12/16/2015 03:03PM by wingmaster.

Thanks wingmaster....I appreciate you sharing this info.

Hey Wingmaster.

The guy on the PiMaker page on thingiverse - is he not the creater of Polar3D at [polar3d.com] ?

Hello from Germany!

Im actually do the same with my Prusa I3 / Mendel90 (link in signature).
replace the Y axis with a turn table because i need realy round prints and not prints that just look like a roud thing.

Work principe of a carousel lathe
simspeed, did you build it? or anyone else?

Edited 1 time(s). Last edit at 10/05/2016 09:59AM by Kenny-F-Powers.

---

Prusa Air 2
Merlin Hotend (passiv und verlängert)

Eigenbau X300 Y500 mit verlängertem Merlin Hotend
BIQU Magician
JGAURORA A5
Anycubic Photon
--------
Eine Kiwi am Tag deckt den menschlichen Tagesbedarf an Kiwis

I've returned to thinking about polar's, a question I had on the polar 3D thread never got answered,
but the other night I was thinking about it, and the thought was that if you can get the bed, or arm to rotate only 90deg maybe you could do half the moves X, then the Y, or combo of that so still kind of Cartesian, bed might have to shift axis too...still wrapping my head around it.

Edited 1 time(s). Last edit at 10/06/2016 09:47PM by MechaBits.

I have been toying at some point with a similar concept but have left scara&polar aside for a while. It's an interesting concept I would like to re-consider at some point. I still have the CAD, here attached is my design I explored back a year or so...
The linear part of the XY gantry should only be able to move half way of the diameter to cover the whole print surface. You should look into the R-360 3D printer that was open-source (with open sourced sofware too) and which had a linear Z, a linear X and a rotary Y just like you need. Maybe you can really use the software as is, I see no difference in geometry.

Edited 1 time(s). Last edit at 12/14/2016 06:43PM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff