XY plane not parallel to working surface/calibration issue when drawing

[virtualritz]

uArm Serial No.: UARM-0505170439

Firmware Version: 3.1.9

Operation System: macOS 10.12.5

uArm Controlling Method: uarm Studio

I am doing a first test with one of my two Swift Pro units that arrived yesterday. I'm using the universal holder with a ballpen to draw something. What happens is that the uarm seems to move in Z as well, not just X & Y.

I adjust the zero point so the pen has contact with the paper.

When I start drawing, the pen initially doesn't even have contact with the paper.

As the drawing proceeds and X increases, the pen has too much contact with the paper, the arm has trouble moving and the shape gets distorted.

It look as if the XY plane is not parallel with the surface that the arm is on, but tilted by maybe one or two degrees around the Y axis.

Here is how the result looks. As you can see, in the region where the pen has contact with the paper w/o too much/too little pressure, the result is good.
https://goo.gl/photos/pGFurMvkHmPepyU46

Is there a way to fix this with some calibration procedure?

If this can't be fixed, I can't use the arm for 3D printing (shapes will get warped) or laser engraving as well (laser will get out of focus as arm moves in X).

Cheers,

.mm

The issue I describe seems similar to the one @derwentx is describing/showing a video of in https://forum.ufactory.cc/t/uarm-not-calibrating-position/692/26

[jfdzar]

I am having exact the same problem! Does anyone get to a solution yet?

[derwentx]

Hey @virtualritz
I'm having the same issue with mine, and just about everyone I've asked on the forum is having issues with Z-allignment and calculating position. I dug through the firmware and reversed the calibration process so that I could re-calibrate mine, but that only partially alleviates the problem for me. Even the numbers Tony (CTO) gave me reflect the same issue: https://forum.ufactory.cc/t/movement-out-of-limits/865/6

Tony has confirmed that the firmware still has some bugs to iron out, so hopefully we will see an improvement soon. It is my intention -- once I have enough spare time -- to rewrite and test the position calculation trigonometry in my own fork fork and submit that to engineers. I'll post here once I have something to show. I'm confident that I'll be able to get positioning working correctly accurately soon, since the whole reason I bought this device was to use it as a 3d printer

[virtualritz]

I actually bought it to use it as a 3D printer, laser engraver and calligraphic drawing aid. But all three require predictable movement.

Speaking of printing, do you know if someone if working on a fan mount? That would also improve precision once the Z-axis stuff has been ironed out (https://www.youtube.com/watch?v=7LVu3Ir10UQ).

I will test with my 2nd Swift Pro at home today (yes, I bought two!), to see if it has the same degree of freedom on the X axis as the one I have on my desk at work.

.mm

[AlmostThere]

I have it too, but so slightly that the 3d print is fine.

Why don't you just put something under the arm base to correct it?

[tony]

Hi all , We just finish the instruction of re-calibration. And we also find out that during the factory calibration there might be some situation which affect the accuracy of calibration. Please try to calibrate it again as the following instructions. THANKS
https://drive.google.com/drive/folders/0B-L-tCvknXU9ZXBweVlYRXd5VHM

[hoodel]

I can level the XY plane by slipping a couple sheets of paper under one side of the arm. Unfortunately, my XY "plane" isn't planar, but is a bit conical... with the effector (print extruder) going toward -Z with increasing radius. This is not a big deal with very small pieces, but over a few inches of radius change, I go from "can't extrude" (against the platform) to "won't adhere" (too far away from the platform).

Question: Will this calibration procedure correct a conical XY "plane?"

Addendum: Actually, I think it slipped a cog or two. I fought it to get it loose from a workpiece, and forgot to turn it off.

[tony]

Yeah, you'd better try the calibration. The radius might be caused by the factory calibration.

[hoodel]

Calibration corrected the conical XY plane issue. Thank you.
I added an extra "M2401 C" step at point C, before doing the "M2401 B" at point B. I don't know if that did anything other than making me feel better.

[virtualritz]

Hey @tony,

after calibration the issue is not entirely gone. When drawing, I can now get the shape drawn with a soft felt pen. But the pen still has too much pressure at C and too little at A. The shape is thus still slightly distorted, closer to C and the lines become noticeable thinner or have missing bits, around A.

If I use a ballpen, the arm will have to do a lot of work because of the high pressure at point C. Basically, the XY plane is still tilted; this time downwards away from the arm.

I did the calibration for point B only and then again, for points A, B & C. I have two uArm Swift Pros. The behaviour is exactly the same. As I get away from B, the pen is pushed down too much.

So this must be something wrong in general. If it was specific to one of my arms, I would only see the XY plane issue with one after calibration. But it is exactly the same with both arms.

.mm

[tony]

Hey @virtualritz， you could use the blockly Moveto function to test how big the difference is the Z in point A B C. Usually, it should be around 0.5mm after the calibration. So could you please do a test like that to see if the calibration works fine for you? Thanks

[virtualritz]

I didn't test but I would be it will be 0.5 more or less because that's what I am seeing.

0.5 is too much to do any drawings with a hard pencil or fountain pencil and will also create issues with focussing of the laser.
Detailed laser work will not be possible unless the arm is close to point B.

Is this going to be improved to be much less than this, e.g. less than 0.1mm, in a future update?

.mm

[tony]

We are designing a new pen holder with spring in it which would solve the drawing issue with hard pencil. And it's possible to make the 0.5 much smaller in the future, as well. Thanks

[Hjelt]

It was discussed in the FB group that there would be calibration coming also to A and C points. If that is correct then I assume that may fix the issue. Any news on that @tony?

[serdan]

Have the same issue. The difference between C (-48.7) and A (-50.3) is almost 2mm after the calibration. That makes drawing impossible even with soft pen.

Minor issue is arm also slightly inclined to the right if you compare intersections of the B radius with the calibration list border: right intersection has (-49.24), left has (-49.79).

[tony]

I think you should try the calibration guide which would improve the accuracy.
https://drive.google.com/open?id=0B-L-tCvknXU9ZXBweVlYRXd5VHM

[tony]

Hi Hjelt, sorry for the late reply, since we are dealing with some different issues from our customers, so the process is a little bit slow for each. We will keep you informed about it. Thanks

[serdan]

Hi @tony, that is actually after the calibration. The situation before was slightly worse (2.5mm diff).

[tony]

That's weird, did you upload the 3.1.16 first? And did you plug in the power before calibration? Show me some picture about your steps might be helpful. Thanks

[virtualritz]

Hey @tony, this is not weird. It should be a known problem to you guys by now.

I'm seeing exactly the same issue @serdan is describing, with 3.1.16, after calibration, with both of my uArms!

The xy plane is still not well enough aligned.
It makes uArm swift Pro useless for something as simple as drawing and even more so for engraving (laser goes out of focus) or 3D printing (unless you use a tiny working volume in xy).

Please fix this.

.mm