RepRap Helios (SCARA)

[DonaldJ]

Nicholas Seward recently uploaded a couple of videos showing the printer in action.

https://www.youtube.com/watch?v=CJ2F-YjHe2k and https://www.youtube.com/watch?v=9_8dOvU_q6U&index=3&list=WL

Print quality is higher than I expected and this prlinter may be ideal for certain niche applications.  I'd like to see him print a big banana.

[Ryan Carlyle]

Oh yeah, I was meaning to post about this. It's a nice design. SCARA usually sucks because of the mass and weight of the Z stage slung on the end of the robot arm. But he's got a clever belt setup, basically a set of compound gearboxes made with belts, that uses stationary motors for all three motion joints. Higher gearbox ratios + giant printed belt pulleys + stationary motors = SCARA that doesn't suck.

Main issues come into play when you want to switch "arm modes" from the elbow joint bending one way to the other. You basically can't do that during extrusion moves. And the reachable area in general is a big C shape. So it's got a giant print area for lots of little things but a fairly small print area for each individual part. 

[Jetguy]

FYI, I just approved Nick's request to join. Hopefully, he'll jump in soon.

[Nicholas Seward]

Thanks for posting this here!  I would like to correct just a bit of misinformation from Ryan.  I can print an area about 4 times as big as a standard Prusa without switching arm modes.  This is great for big things.  Here is a link to a 450mm long print.  https://youtu.be/T8sIGOBbTHg   I can't agree more with all of his other points.

I will add that this will have trouble past standard default speeds.  This printer is good if you want big and portable but not if you want speed.

[Ryan Carlyle]

450mm? Ok, yeah, that's bigger than I thought, sorry for the misstatement.

[Nicholas Seward]

No worries!

[JasonB]

I'm not much of a software guy, but would it be possible for the motion planner to read ahead and choose shortest possible paths and arm mode changes per move?  Or would it be merely incredibly difficult?  I saw some moves that seemed a bit superfluous in the videos.

[Nicholas Seward]

The probing sequence will get better.  That is the job of the slicer.

I currently have G0 moves do machine moves instead of linear moves.

Could you tag the time where you think the superfluous moves occurred?  I would like to see if I can tighten this up.

[JasonB]

I noticed at the end of the arm mode change it seems to overshoot the next print location to compete the change, then move back out to the next position.

[Nicholas Seward]

Oh yeah.  I got that doing an optimized arc now.  There is now no harsh arm mode changes.  The problem was that the G0 command wanted to move in a straight line.  Now it see what the starting and ending machine coordinates and just interpolates between them.  It is kinda cool when I do rapids over a print it now leaves arcs instead of lines.

[JasonB]

Awesome!  Is there a github or repository for this code yet?  I like to at least try and understand things I see as magic and this certainly qualifies.  Weather or not I succeed is another matter.  I work around some industrial SCARA arms, but they only move in one linear axis and rotation, plus all our interaction is through a teach pendant that interprets simple direction commands to the proper joint motions.  

[Nicholas Seward]

I will publish on Github when I get v0.2 verified.

Thanks,

Nicholas Seward