[2.0] Skew Correction

[thinkyhead]

Based on #8623, #8204 by @Paciente8159

---

Concise Diff

Machine skew in the XY, XZ and YZ planes can be adjusted with command M852 I J K command.
Values are stored in EEPROM.

This is a re-commit of #8159 with the changes suggested by @Allted and additional mangling by @thinkyhead.
This addresses issue #3839 and #5116.

[CCS86]

Is there a discussion somewhere about the best ways to determine the skew values?

[thinkyhead]

@CCS86 The exact way to determine the skew values is documented in the configuration file. You just print a square, take some measurements, and input them as your side lengths.

/**
 * Bed Skew Compensation
 *
 * This feature corrects for misalignment in the XYZ axes.
 *
 * Take the following steps to get the bed skew in the XY plane:
 *  1. Print a test square (e.g., https://www.thingiverse.com/thing:2563185)
 *  2. For XY_DIAG_AC measure the diagonal A to C
 *  3. For XY_DIAG_BD measure the diagonal B to D
 *  4. For XY_SIDE_AD measure the edge A to D
 *
 * Marlin automatically computes skew factors from these measurements.
 * Skew factors may also be computed and set manually:
 *
 *  - Compute AB     : SQRT(2*AC*AC+2*BD*BD-4*AD*AD)/2
 *  - XY_SKEW_FACTOR : TAN(PI/2-ACOS((AC*AC-AB*AB-AD*AD)/(2*AB*AD)))
 *
 * If desired, follow the same procedure for XZ and YZ.
 * Use these diagrams for reference:
 *
 *    Y                     Z                     Z
 *    ^     B-------C       ^     B-------C       ^     B-------C
 *    |    /       /        |    /       /        |    /       /
 *    |   /       /         |   /       /         |   /       /
 *    |  A-------D          |  A-------D          |  A-------D
 *    +-------------->X     +-------------->X     +-------------->Y
 *     XY_SKEW_FACTOR        XZ_SKEW_FACTOR        YZ_SKEW_FACTOR
 */

[CCS86]

Doh! How did I miss that? hahaha

Thank you @thinkyhead

[CCS86]

In case anyone is interested, I modeled my own calibration square. I saw some areas to improve over the one linked:

• To avoid a bottom lip that needed to be carefully removed, I added a blended chamfer.

• I liked the planar faces for the diagonal measurement, but didn't like that there was such a sharp transition leading into them. I could see a ripple in that face, even when printing outer walls at low speed / acceleration. I added a large tangent arc as a lead in to that face.

• I also added a little peak on the Y+ side. This is helpful in getting the slicer to align the seams there.

I print it in Cura 3.1 with no bottom or top layers, 35% density triangular infill, 8 perimeter laps, a brim, and 25 mm/s outer wall speed:

Model:
Calibration Square.stl.txt

[thinkyhead]

Nice work! Is it essentially more sturdy, easier to measure from corner to corner, and so on? We should include it in the Marlin package with any other test/calibration G-code we think would help get settings dialed in (and to stress-test things like feedrate, acceleration, direction-changes, braking, etc.).

[CCS86]

Feel free to include it.

I also made a model for testing motion settings / planner throughput, that has been helpful. I print it as a spiral vase at different speeds/acceleration/jerk combinations:

MT3.stl.txt

[thinkyhead]

Wow, that is great, and a work of art besides.

As it happens we're in a total testing/debugging mode right now with Marlin, including adding more stress-tests like these for both developers to check code and end-users to check settings. I just started a MarlinTest.gcode script with just the most basic moves being tested so far. Even the most basic moves can expose bugs, such as the current (ridiculous!) last-move-not-joined bug.

The script uses M1 to put a prompt on the screen for each test, plus M118 to print headings to the console, and M114 D1 to print extra-detailed position reports (needs M114_DETAIL enabled). Later I'll put some of tests in a DEBUG menu that will also do internal and guided testing of Marlin and the machine to make sure everything is functioning properly. (Too many bug reports where users have a bad component.)

So, anyway, please contribute more ideas and tests! We'll be happy to use them and make them part of the test suite.

[CCS86]

Very cool. I think my thread here could be of some help in the testing/tuning:

#9219

[Roxy-3D]

So, anyway, please contribute more ideas and tests! We'll be happy to use them and make them part of the test suite.

Would it be appropriate to add a few line drawing GCodes to help test the linear advance stuff? It might be helpful for the people that use that if there was a standardized test. Maybe something that can draw a small sub-section of this pattern:

#9142 (comment)

[thinkyhead]

Would it be appropriate to add a few line drawing GCodes to help test the linear advance stuff? It might be helpful for the people that use that if there was a standardized test. Maybe something that can draw a small sub-section of this pattern:

There is such a test posted somewhere on the internets. Or, a script that generates the test G-code. We could include that.