Welcome to the SeeMeCNC Community

why MatterControl only shows and work only whit the steps of z
My step motors need different steps.
I think it's because of different tension or something with the belts.

this are the numbers
#define XAXIS_STEPS_PER_MM 79.8935 //In this case gives me 0.4 millimeters less after moving 300mm
#define YAXIS_STEPS_PER_MM 79.9468 //In this case gives me 0.2 millimeters less after moving 300mm
#define ZAXIS_STEPS_PER_MM 80.1065 //In this case gives me 0.4 millimeters more after moving 300mm

I have put the values of x and y into z and and the towers give me exactly the millimeters I move.

Is there any way to make each stepper motor work with its necessary steps.

now I measure with my caliper 03/2018

This doesn't work like the way it does on a Cartesian machine... There isn't 1 stepper motor per axis, so adjusting the motor we call X to change the X axis, etc won't work.

Your steps per mm should be the same across all 3 of your stepper motors, and the only reason to change this is if your pulley tooth count or type of stepper motor changes. If you were 100% sure that you knew what you were doing, you could change this value ever so slightly to compensate for a stretched belt or something like that, but that's the only reason to do it.

If you've got an XY scaling error, look at your arm lengths. If you've got a Z scaling error, this is your steps per mm on ALL 3 of the motors.

hectori,

This frustrates me as well. Believe it or not, Marlin accepts different steps/mm for different towers. Repetier, in its infinite wisdom, won't let you set individual steps/mm for each tower. The only way I've found is to set the tension such that all the motors move the same amount.

Nice work on the measuring, I find it difficult to measure this figure accurately.

nebbian wrote:Believe it or not, Marlin accepts different steps/mm for different towers. Repetier, in its infinite wisdom, won't let you set individual steps/mm for each tower. The only way I've found is to set the tension such that all the motors move the same amount.

Are you sure Marlin has that option? Or does it accepts steps/mm in the X / Y / Z coördinate space. Naming the Delta towers X, Y and Z can be confusing, because it has no relation to the X, Y, and Z coördinate space. But I have never seen Marlin so I wouldn't really know.I think the only way to really perfect your calibration is to aim for equal belt tension. 8-bit controllers have enough difficulty to do the kinematics as it is, adding extra parameters won't be beneficial in that regard I guess.

Although I doubt that it means a significant difference. If you calibrate your endstops correctly, the effect starts to build from printsurface upwards, so it would only effect very high builds, and the effect will be smooshed over many many layers. Are you sure the measured steps differences are due to belt tension? They could be explained by other problems as wel, e.g. flex in the joints. I recently moved from the old melamine skates to the new injection molded ones plus ball-cup arms. Flex in the joints has been dramatically reduced. Also Due to this I retensioned the belts, by hand, a few times, and Duet auto cal does not see a big difference in endstop corrections after each retension. Therefore I think stressing about belt tension is overrated

DeltaCon wrote:

nebbian wrote:Believe it or not, Marlin accepts different steps/mm for different towers. Repetier, in its infinite wisdom, won't let you set individual steps/mm for each tower. The only way I've found is to set the tension such that all the motors move the same amount.

Are you sure Marlin has that option? Or does it accepts steps/mm in the X / Y / Z coördinate space.

Yes, I'm sure, because I very carefully measured the actual distance travelled by each carriage and the adjustments in Marlin correspond to the distances that each carriage moved. I was able to tune these values to get exact distances, just as the OP wanted. We're not talking large amounts here, we're talking much less than a percent of difference (just as the OP has measured).

If you don't need interference fits from parts printed in different areas of the bed, then you won't notice the difference. But I do need that precise accuracy, so it's something that I take a lot of care trying to achieve. Basically I need a 4mm cylinder to have less than 0.5% variation in size (that's 0.01mm), wherever it's printed on the bed. I've managed to achieve this (most of the time). Note that errors in X/Y space are magnified from errors in steps/mm on the towers, because the diagonal rods are closer to vertical than horizontal when printing in the middle of the bed.

I use this tool to calibrate X Y Z steps. this image shows the difference between repetier and Marlin calibrated the steps x tower. 0.25mm nozzle Hectori