Calibration Cube XYZ is off, how to adjust Axis length

duetino · 25 Sept 2017, 07:22

Hi,
I have a Tevo Little monster and printed the testcube (https://www.thingiverse.com/thing:2145703), because all of my prints where not really fitting together. I know why now, since I never paid attention on how the WCS was aligned (on a delta it does not matter I thought)
As i measured all the sides and get the following reading:

X: 20.2
Y: 19.99
Z: 20.01

As you can see the Y and Z Axis have an acceptable deviation, but X is far off limits.

My Question is where and how can I feed some correction factors into the machine? gcode?

dc42 · 25 Sept 2017, 07:36

The easy way is https://duet3d.com/wiki/G-code#M579:_Scale_Cartesian_axes. But I suggest you measure a print larger than 20mm.

duetino · 27 Sept 2017, 05:51

Thank you I will print a upscaled hiphop uptodate cube to get accurate settings. Do you think that 100x100x100 is the right dimension?
all the best

duetino · 27 Sept 2017, 12:08

Hi, thanks for the response and which leads me to the following questions:

@dc42:

The easy way is https://duet3d.com/wiki/G-code#M579:_Scale_Cartesian_axes. But I suggest you measure a print larger than 20mm.

… that that does not so directly apply on deltas or am I unright and the wiki does not know the duet3d controller?

https://duet3d.com/wiki/G-code#M579:_Scale_Cartesian_axes here is says …

But this does not work so easily for Delta and other RepRaps for which there is cross-talk between the axe

please let me know before I print a monster cube

fma · 27 Sept 2017, 13:35

I also noticed such deviation on X axis. I didn't dig yet, but is changing the scale the right answer?

I mean, it can be an extrusion problem, so tuning scale on a 100x100 square won't fix the problem on a 20x20 square.

So, I suggest you first try to see if it is really a scale problem, by printing a 100x100 square (100x100x2 cube, as you don't need to calibrate Z; even a 100x2x2 should be ok, if you only want to check X), and see if you get something like 101x100x2 (10% more on X).

fma · 27 Sept 2017, 13:36

In fact, you can first try to measure the head move with a good ruler, without extruding. 10% of 200mm is 2mm, so easy to see.

duetino · 30 Sept 2017, 21:01

Hi, I'm talking about a tenth of a millimeter on X, which is way off, since Y and Z only have a hundredth deviation

dc42 · 30 Sept 2017, 21:06

M579 can be used to correct axis scaling on any type of printer. The "this does not work so well for a delta printer" bit refers to adjusting the tower steps/mm instead as described in the preceding sentence.

You should also check that the X and Y sides of a large cube or square are straight, not bowed inwards or outwards.

fma · 1 Oct 2017, 07:07

Duetino, if your problem is really a scaling problem, when moving 200mm you should measure 202mm, which is easy to see. If you still have 0.2mm more, this means this is not a scaling problem! In this case, if you correct the scale for 20mm, a 200mm print will be something to 198.2mm…

So, first, move your axes as far as you can (200mm), and measure with a ruler.