M556 parameters?

Kulitorum

What are the parameters for M556? - I can only find one description and I fail to understand it.

I have a machine that's skewed. If I move 4900mm on the X-Axis, the Y is off by 23mm.

So I tried M556 S4900 X23 P1 ( to make it compensate on the Y-axis? - the P parameter is really never explained)

But it does not go as expected - no real difference.

Some documentation says that the X parameter is in Tangent, others in mm (maybe)

Can you help me understand the parameters?

Thank you.

infiniteloop

@kulitorum You mean this here: M556: Axis skew compensation. That's the ultimative reference, regardless of what others say.

the P parameter is really never explained

Well yes, it is.

BTW: if your Y is 23 mm off (over 4900 mm - is that true?), you really should adjust your frame.

EDIT: Did you read this: Orthogonal axis compensation with M556?

Kulitorum

@infiniteloop Thanks for the help - yes, I read it all.

You think the P parameter is explained? - Pnnn means what? Is P2 an option?

Yes, it's 23mm over 4900mm I will fix the frame, but right now I have half a house printed and the printer no longer aligns with the print for some unknown reason

It gets better with M556 S100 X-1 P1 and even better with X-1.9 but X-2 throws it way off. I don't get it.

Also the documentation only referes to using the printed test piece, which really does not apply for 12x20m long houses - it does not really explain how it works.

I hope @DC42 will try to explain tomorrow, and maybe update the documentation for others to read later.

Thank again.

infiniteloop

@kulitorum

the documentation only refers to using the printed test piece, which really does not apply for 12x20m long houses

So true! Sorry that I could not imagine such dimensions.

Pnnn means what? Is P2 an option?

From the documentation:

Apply XY compensation to Y axis instead of X

To me, this looks like a flag which, if set (to 1), applies the counter-skew to Y instead of the default X.

It gets better with M556 S100 X-1 P1 and even better with X-1.9 but X2 throws it way off.

Did you mean X2 or X-2?

Kulitorum

@infiniteloop X-2 of cause

infiniteloop

@kulitorum

X-2

That’s really weird. In what way does this „throw it off“? Another thing to note is the S value: according to the documentation,

The S parameter is the length of a triangle along each axis in mm.

Printing a house, 100 mm are way off - try 4900 instead. But I have to admit that my approach is speculative: being unable to apply the calibration tool as described in the above link, I'm left with pure guessing …

Sorry, but time for an expert to chime in. @Phaedrux, do you know who could help with this?

dc42

@kulitorum said in M556 parameters?:

So I tried M556 S4900 X23 P1 ( to make it compensate on the Y-axis? - the P parameter is really never explained)

That should work. The X parameter divided by the S parameter is the tangent of the difference between the the actual XY angle and 90 degrees. P0 (the default) causes any skew between X and Y to be compensated by adjusting the X coordinate by Y * tanXY; whereas P1 causes the XY skew to be compensated by adjusting the Y coordinate by X * tanXY.

If you go to the Object Model Browser in DWC and expand move->compensation->skew you can see the tangents that have been set.

The S parameter in M556 must be at least 10 but there are no limits on the values of the other parameters.

[Edit: I have made some corrections to the above since I first posted it.]

dc42

PS - I tried entering your original command:

M556 S4900 X23 P1

and the object model shows this:

which looks correct to me.

Kulitorum

@dc42 Agreed, it just does not seem to make any difference on site. That being said, the printer is in Saudi Arabia and I'm in Denmark, so I do not have hands on myself. We already have people on site and are sending another guy down there to help debug it.

Thank you for the input.

dc42

@kulitorum which firmware version are you using?

Kulitorum

@dc42 On most machines we are running Release 3.3 - never machines are 3.4.1

droftarts

@dc42 I’ll add what you have written in your post to the appropriate pages, for a clearer description of function.

Ian

Kulitorum

@dc42 Cool, it works now - dont know why it didnt before, but now it does. Thank you for the help.