What's wrong about my robot arm 3d printer ?

Hi guys,

this link is my robot arm first print
[www.youtube.com]

please watch it carefully, the print result is not perfect.
so please tell me what did I do wrong and hopefully you tell me how to fix it too.
you can give me additional ideas too if you want.

thanks

I am not sure what you expect from this kind of printer. I can see the technical appeal, but for precise printing it is maybe the worst setup i have ever seen. Long lever arms with high mass at the end and plastic connectors to the other parts. Two circular axis with the inherent numerical imprecissions.
For what it's worth, i think the result is actually pretty ok for such a setup.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

... read about difference in accuracy between "normal" and "parallel" Scara robots - you have a "normal" type with typical behaviour, a parallel Scara will behave different with much more stiffness and accuracy over the working area ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

@srek
I expect to get much larger print size with same material as delta printer.
How to make it better?

@vdx
I don't like pararel scara.
It's hard to make exactly the same length of arms, in the end hard to calibrate as delta.
I make calibration route very easy for this model and it quite accurate and only takes not more than an hour.

... but then you'll get all the "decreasing accuracy per length" issues of a "normal" (or serial) scara

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Hi vdx...
do you mean the more arm length the more inaccurate?
if that the case, isn't parallel scara also the same?
and it should be cured by use bigger gear ratio... right?

... the displacement-error is related to the overall arms length, but too to the distance from the base, what's even more problematic.

With the parallel scara the error with distance from the base is highly reduced, so the "error chart" is nearly constant over the complete working area.

There were some papers about this with pretty describing diagrams, but they're not online anymore ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

@VDX
I don't really understand what you mean with parallel scara reduce distance error. I think by nature parallel scara is a bit more stiff than single. but reducing distance error, I don't think it is.
I think in parallel, if I don't make the parallel arms really parallel, it would be an additional problem in calculating the error.

... no, it's more than that - a serial scara has only one "base" and suffer from any mechanical elasticity flexing the arm ... the more "stretched" it is, the more flexing through weight and dynamic forces ...

A parallel scara is not "geometrically parallel" - it has two arm bases with a distance of maybe 15% - 20% arm length and both arms are jointed in the tool center point.
This gives an enormous benefit in stiffness with stretched arms!

But you have to calculate the tool position with Inverse Kinematics, similar to a delta, to get this working with our normal XYZ positioning ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

So,single arm and parallel has different kinematics?

Maybe check here [www.youtube.com] to see what kind of mechanical effort you have to make to get a single arm Scara stable.
It will never be as fast as a delta or cartesic printer.
Single Arm Scara and Parallel have very different kinematics, i am not sure which is mroe computing intensive though.
In my opinion the single arm Scara printer combine the worst aspects of all printers, closely followed by the parallel.
Do not expect it to be precise, do not epxect it to be fast. If you encrease the build area the imprecision will get worse, similar to a delta, but quicker.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

... yes, they are different - but can be precise and/or fast too ...

[www.youtube.com]

[www.youtube.com]

When I was involved with the accuracy testing, Harmonic Drive assembled a small demonstrator to show the precision of their "harmonic gears" with a pretty nice and small setup with maybe 200mm overall footprint, 150mm square working area and sub-micron positioning accuracy

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Both those examples show machines with a Z height of a very few centimeters.
If you were to use a heatbed that moves in z direction i think it can work. The OPs idea of mounting the Z axis at the end of the arm will not be able to produce quality or fast prints.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Tx for the input.
I redesign my robot arm.
Z axis at the palm is a bad idea.

I believe scara arm can be fast and accurate.
My test print size is precise enough for me.