Morgan Twin Arm SCARA robot

I thought I would post a link here to Quentin Harleys' "Morgan SCARA" robot. A really nice build and it prints very well.

Morgan SCARA

Now if we can just get rid of the linear rails for the Z axis we would be really on our way to a fully printable robot. I hope he posts the models soon as I am sure this will be a popular build.

Edited 1 time(s). Last edit at 03/01/2013 11:22PM by martinprice2004.

its very beutiful machine.
why not rails on Z axis?

@woo

Linear rails are very difficult to print and no one has yet achieved it. Removing them and using an alternative method would enable more of the machine to be printed.

martinprice2004 Wrote:
-------------------------------------------------------
> I thought I would post a link here to Quentin
> Harleys' "Morgan SCARA" robot. A really nice build
> and it prints very well.
>
> Morgan SCARA
>
> Now if we can just get rid of the linear rails for
> the Z axis we would be really on our way to a
> fully printable robot. I hope he posts the models
> soon as I am sure this will be a popular build.

Damn, that's awesomely innovative. I was wondering how he achieved so nice movement, but the coaxial setup allows using belt drives with large reduction ratio, and the vertical spacing of the arms helps with the horizontal stability. I also have to wonder, though, whether it wouldn't be simpler to mount the big pulleys right at the arm bases (maybe even print them in one piece with the arm).

I have been thinking about ditching the harmonic drives in favor of belt drives myself - it may be theoretically impossible for the harmonic drive design to reproduce itself since the inaccuracies in the drives get multiplied in the tool path... and a non-cartesian design which requires a cartesian bot to reproduce would be a really sad thing.

@ttsalo

Yes, similar thoughts struck me. The shafts from the pulleys are long and will twist a little even though the torque is low. The error is exaggerated by the long moving arms. The drive gears would be better placed at the top. I'm also not too keen on a rising table. There are a few benefits of a fixed bed which I would want to keep on my design. It was probably built this way to keep the Centre of Gravity low on such a tall machine..

My SCARA has separate pulleys, but they are attached to the arms with a dowel, but could be printed as one. The only downside of this is it is a larger part to print if you want to change the gear ratios. I think the best ratio will be something that will be done by trial and error (Speed v Accuracy) when there have been a few designs developed and we have a better understanding of arm length and working envelope.

The coaxial drive looks neat, but I think they would be more stable if they were separated out like your design.The spacing would have to be sufficient so the large pulleys didn't interfere with the shafts but they could be offset in Z so the spacing was just over half the pulley radius. the belts work really well on my design and Quentin is finding the same.

I have been CAD modelling a two arm design with pulleys similar to your thoughts . Here's what I have so far.

The motors would be better behind the vertical rails so they act to reduce the cantilever load on the rails as my first design, but my Z axis design would need a complete rehash to do this. You need to reserve the space either side as the arms can go right around to give a wider build envelope. This is why I place the motors in the middle.

I am still much keener to improve my single arm design (I think they look cooler) so I am a bit unsure of what to build next!

The best design is clearly a combination of the three designs we have on the table at the moment. Armstrong, Morgan and Helium Frog.

Perhaps we should call it the ArmMorous Frog (read Amourous Frog!)

Edited 1 time(s). Last edit at 03/06/2013 04:35PM by martinprice2004.

I built a direct drive setup back in January.
[plus.google.com]

What I found out from the experiment was that resolution was probably fine for my direct drive setup, but as I increased the speed my poor little motors just did not have the torque to hold the position with all the back forth movements required to rasterize a straight line.

Here is what happens when I go faster.
[plus.google.com]

Here is what the crazy rasterization looks like.
[plus.google.com]

Sometime soon I want to run the experiment on bigger motors and see if I can eek out a working direct drive setup before falling back to gearing down.

Hi there!

My drive wheels are at the bottom of the machine for two reasons:

1. They would intrude into the build area if directly below
2. If above the arms, the weight of the machine would be very high, and then you also have to devise some system to suspend and hold the motors, without interfering with the bowden cable...

The drive wheels are very large in order to make the resolution as high as possible. Twist in the shafts are limited because the 22mm pipe has a large diameter, countering twist, and the inner shaft is doubled up - 8mm threaded rod and a 15mm copper pipe combined.

Morgan is almost ready for public release... and is is busy making it's first child ;-)

Cheers,
Quentin

Morgan is now released and in the wild

[reprap.harleystudio.co.za]

hello.


so, i am little stupid, and will ask a question


// SCARA tower offset (position of Tower relative to bed zero position)
#define SCARA_offset_x 100 //mm
#define SCARA_offset_y -65 //mm

if i got it right, this are offsets from tower center to bed x0,y0 or?

#define SCARA_RAD2DEG 57.2957795 // to convert RAD to degrees
//#define SCARA_DEG2RAD 0.0174532
this one converts motor steps to degrees that first linkage does? whats right eguation? I assume thats simple eguation with steps converted to degrees, and then calculated in reduction ratio, but i rather ask.
thank you.

Edited 2 time(s). Last edit at 07/04/2013 03:08AM by woo.