[Delta] Individual steps per mm settings for X, Y and Z-axis

dec

I'm setting up a delta printer. I want to calibrate the steps per mm options but unfortunately in delta mode only the ZAXIS_STEPS_PER_MM option is used.

I think this behaviour should be changed to gain a better printing quality.

Repetier

No, all 3 motors MUST have the same steps per mm so only one setting is needed. Math does not work with different settings.

dec

Why should that be?

The math uses the unit mm, I assume. Z-axis movement means the same amount of mm for each tower but each tower could have a different steps per mm ratio. At least that's the way I would do it

Repetier

We use normally integer math so we convert position into steps allowing us faster sqrt of integers and results are directly positions in steps. See transformCartesianStepsToDeltaSteps in motion.cpp,

If you have the same microstepping/belt/pulley you have the same steps. Deltas are symmetrical, so why should one want to use different resolutions?

dec

Of couse, ideal and 100% identical motors, pullys, belts etc would result in an even movement on a delta. But when I measured the movements of all 3 steppers there was some difference.


I was just curious if the behaviour of the firmware is normal and intendet.

Repetier

Ok, at least I can say behavior in firmware was intended.No complains about that the last years so I guess not identical behavior is more the exception.  

What difference do you get? Should not be more then your substepping. I think motors can not always hold motor at exact substeps but you will always be inside the right full step. Different tensioning might extend one belt a bit more whcih could be fixed with identical tensioning.

dec

It's a bit tricky to measure, so I'm measuring repeatatively. One motor had an offset of 0.2mm per 100mm. I rearanged the carriages (another issue) and the belt tensions. So I'll recheck, maybe the offset can now be neglected.

MShievitz

Dec, the confusion comes from the conversion from cartesian to delta math and the axis names. I find it easiest to think of the towers as "A", "B" and "C". I wish that they had been named that way. It makes it easier to explain. "To move the x-axis, towers a,b and c must move the same amount, in the same direction and have the same limit, or the platform will skew" explained this way, it's easy to understandnd why they must all be symmetrically configured.
Good luck.

covert

Came here looking for the same solution, Being able to individually change the steps for mm on a per motor basis.I'm having issues with the Z level between the towers being lower than at the towers.

This thread says the error is a relationship between steps and Diagonal rod length and explains a iterative process to address it.

http://forums.reprap.org/read.php?178,492382

I checked my steps and found two towers to be out a little (80.5 steps/mm).

Repetier

If the belts have the same tension they have the same steps per mm if pulleys etc. are of good quality. So tighten the other tower the same and set 80.5 for all towers.