Wally for Marlin

Wally for Marlin

Postby mateusz.p » Tue Mar 18, 2014 2:26 am

My name is Mat and I am new to this forum, I am software engineer for over 8 years and I am interested in 3d printing for almost 1 year. Build 2 printers off the open source world (wallace and smartrap) and I love Wally design.
That all for introduction, I am interested in Marlin/Sprinter support for Wally.
I should assemble Wally XY parts soon and start working on Marlin update for inverse kinematics.
So far I got basic math model work out, from 3 circles equations:
    1. first arm base point and length of arm as radius
    2. second arm base point and length of arm as radius
    3. hotend point and length of arm as radius
From that I will calculate points position in elbow points and then calculate angles of final arms - that are driven by motors.
I will add sketch and graph in few days with some example equations and math logic.

Nicholas, do you have some math model that you was basing on when creating preprocessor script? I feel it would be useful to take a peak before I start getting Marlin update on the way.
I looked into python code, however I never used it, which makes harder for me to get full understanding of its logic.

Posts: 7
Joined: Thu Mar 13, 2014 5:51 pm
Location: Chicago Area, IL

Re: Wally for Marlin

Postby Nicholas Seward » Tue Mar 18, 2014 3:04 am

Inverse Kinematics

There are a few reasons this is going to be a challenge.

1) There are several trig and inverse trig functions that are needed.
2) I have calibration curve for the z heights.
3) I do bed tilt/level compensation in the processor.

I feel that all of that can be overcome with clever programming and lookup tables but it is far from trivial.

(FYI: There is a known bug in segmentize. Something is wrong with the coordinate transformation.)

For the record, Wally would be better matched to Smootie firmware running on an Arduino Due with RAMPS-FD or for an even better match to a LinuxCNC running on a Beagle Bone Black with a BeBoPr. Delta machines just barely get by with Marlin on AVRs because the inverse kinematics are so simple.
Nicholas Seward
Posts: 738
Joined: Mon Nov 25, 2013 10:41 pm

Re: Wally for Marlin

Postby mateusz.p » Thu Mar 27, 2014 4:21 pm

Thank you,
I am half way through and discovering that much more computations needs to be done.
Not sure if Mega 2560 can handle it - but I give a try, at least it will do it slowly.
At least I got my wally XY parts printed and mounted - so I can run tests on real device.
Posts: 7
Joined: Thu Mar 13, 2014 5:51 pm
Location: Chicago Area, IL

Return to Wally

Who is online

Users browsing this forum: No registered users and 1 guest