Delta Robot Style 3D Printer

[Hamp856]

 I see that the Duet 6HC can drive 6 stepper motors. I am wondering if you could build a 3D printer more along the lines of a Delta Robot. You would have 3 Steppers driving the 3 arms and a 4th stepper move the entire arms (3) assembly up and down the Z axis? Is this doable??

See the attached picture of a Delta Robot. Imagine this whole assembly being driven up and down the Z axis with a Ball Screw/

I have not seen a machine like this before. I think it would be a very cool way to do it.

[Vincent Chernesky]

It appears to be similar to the Nelu.  You already have Z travel with the arrangement you propose, having the entire assembly translate would give a larger Z volume.  The long moment arms and high torques on the stepper motors may lead to control issues and reduced resolution at the print head.

It would be cool to see in operation.

[Konstantin Uvarov]

Check the videos of this guy from Russia... he has a working version of that type of delta printer and also shows the whole process of building it.

https://www.youtube.com/watch?v=6oefrWdS_3M

Also on thingiverse:

https://www.thingiverse.com/thing:3489886

--
You received this message because you are subscribed to the Google Groups "Delta robot 3D printers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to deltabot+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/deltabot/af42bd46-31df-4e31-b8ce-62ea719ba9bb%40googlegroups.com.

[David Crocker]

It would certainly be possible to drive that kinematics from a Duet 3 or even a Duet 2, although a modification would need to be done in the RotaryDeltaKinematics class in RepRapFirmware to use the 3 arms just for XY movement and use the separate motor for Z movement.

What advantages do you see for such a configuration, compared to linear delta? I can see a few disadvantages:

- Rotary deltas need very high resolution in the arm positioning. This means you have to use motors with a reduction gearbox, or alternatively use belt-and-pulley reduction. A gearbox on the motor will introduce backlash, resulting in imprecise positioning.

- Rotary deltas are difficult to calibrate. Even if you get the 3 upper arm lengths identical and the 3 lower arm lengths identical, and you know their values exactly, there are 9 factors to calibrate. This is because the 3 upper bearings may not be in an exact equilateral triangle and may be at different heights, and the 3 homing switches may trigger at slightly different arm angles. We don't yet have an auto calibration mechanism for rotary deltas, and I haven't done the maths to see whether it would be practical to do based on Z probing alone.

- Rotary deltas have to use segmented kinematics because of the trigonometric calculations involved in the inverse kinematics; whereas linear deltas casn use non-segmented kinematics because the most complex operation needed is a square root.

If you do decide to go ahead with this, and you have some C++ coding ability, I can help advise you on changing the kinematics code.

[Hamp Stevens]

Thanks, David,

I very much appreciate the information. My understanding is the the Duet 3 can drive a Rotary Delta without having a separately driven Z axis. That config would work for me. 

I have built 2 CNC machines. This is my first stab at 3D printer.  I like the Delta style (the are just neat to watch). I will probably build a  Linear Delta first then move on to a Rotary.

One thing is for sure. I'm no C++ coder. I have a rudimentary knowledge of Microsoft Basic. I know enough to be dangerous.

Thanks for your input. I'm really excited about this project.

Hamp

--

You received this message because you are subscribed to the Google Groups "Delta robot 3D printers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to deltabot+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/deltabot/ff053c24-9333-4815-b3dc-2f7ef1007ffa%40googlegroups.com.

[Michael Anton]

I don't recommend it, given the problems that were encountered with the FirePick Delta PnP machine that was attempted.  You can read about the saga here: https://groups.google.com/forum/#!forum/firepick.  There are a lot of sources of error in this design.  Even the commercial rotary deltas that use servos, don't achieve much better than 1mm precision in a 1x1m build area.