Incorrect Z axis hight while printing

When I tell my printer go up by 10mm I measure ~10.05mm, I guess the steps per mm are good enough for now.
However, when I start printing the printer thinks it is higher on the Z axis than it really is. Resulting in prints that
are too low.

Description of the problem
I'm printing the 10x10x10mm test cube, here are my results:
X and Y are about 10mm.

First few tries: a 4mm high cube.
M114 showed that the printer was at z10.

After changing some settings in slis3r I get a 7mm high cube.
M114 showed that the printer was at z10.


After the last attempt (7mm high cube) I measured the hight from the printbed to the extruder, this was set a the zero point on my
digital callipers. Then I homed all the axis, placed z at 0 (G1 Z0) and raised Z by 10mm in Pronterface.
Measure the hight from print bed to the extruder again, 2.29mm difference!

I hope you guys can help me out.


Info on printer and software:
Prusa i3
Slic3r
Pronterface
Marlin firmware, development
Megatronics V3
Ubuntu 14.10

See attached files for configuration and sliced object.

I haven't checked your files.

My first thought is that the power to your steppers is not turned up high enough. If you are getting 10mm in z movement outside of printing, I'd have to think your z movement in your Configuration.h file or slicing program (or non print moves) is too fast.

have you calibrated your steps/mm for the z-axis?
[prusaprinters.org]

#define DEFAULT_AXIS_STEPS_PER_UNIT   {82.524592328, 82.524592328, 2026, 700}  // default steps per unit for Ultimaker

Have you checked the endstop offset options Or gcode? it seems maybe like a slicer problem. The endstop offset is to tweak the printhead starting height if the endstop cant be adjusted.

I've had the same issue with the prusa i3. Here's what can cause it and fix it:
- slipping z rod to stepper motor tubes. Replace it with a coupler. To test, mark motor and tube with cd marker. Lubricant can leak down the z rod.
- rod wire can have a deviation. I've used a 1200dpi flatbed scanner and counted the number of pixels for 100 turns and calculated what the z steps should be.
- calibrate the stepper motor voltage so that it doesn't produce noise when sitting still.
- use auto bed levelling as the z adjustment screw can be hard to set correctly. I use a 9g servo with microswitch with consistent results
- the z rod nut that goes into the x end and x motor mount is too loose sometimes. I use a small piece of rubber so it's stuck tighter in place. It can still come loose if the head crashes into the bed so no damage can occur. This helps auto bed levelling accuracy and consistency.
- calibrate the z offset when using auto bed levelling and measure the z height of one printed layer. I use a perimiter or skirt for this.
- make sure the z velocity is set to 3 instead of 5 which can cause missed steps. Make sure you set it in the correct place, depending on your configuration, within the firmware, recompile and upload, or using the display and control knob in control/motion. Make sure to store settings in control menu.
- lubricate all moving parts except the stepper motors, I think these are sealed anyway.
- calibrate the x axis being parallel with the print bed. I use a credit card slipped under the x smooth rod and on the bed with the 55mm side left and right. If it has about the same resistance you're golden.
- level the bed too of course, even when using auto bed leveling it's better to have it level in the first place. The bed, x rod angle and z offset can give an error of about 0.5mm, which can be the difference between parts just not fitting properly.

Hope this helps!

Good luck!

Thanks for all the advice!

Turns out my Z axis was turning too fast.
The printer didn't make the upward move in time resulting in squashed prints.

Getting close to printing a perfect 10x10x10 test cube, only 5% off now.

Man, you gave me the right hint!
- make sure the z velocity is set to 3 instead of 5 which can cause missed steps. Make sure you set it in the correct place, depending on your configuration, within the firmware, recompile and upload, or using the display and control knob in control/motion. Make sure to store settings in control menu.

I'm using two nema 17 low torque in my Z Axis, and the problem I had was not mechanical misconfiguration but MAX_FEEDRATE. The standard Z_MAX_FEEDRATE for Marlin is 5, but with that value many steps were lost. Finally got the right value to obtain 2.5 cms in my 5mm pryamid. Bye bye dwarf printings!

Quote
imqqmi
I've had the same issue with the prusa i3. Here's what can cause it and fix it:
- slipping z rod to stepper motor tubes. Replace it with a coupler. To test, mark motor and tube with cd marker. Lubricant can leak down the z rod.
- rod wire can have a deviation. I've used a 1200dpi flatbed scanner and counted the number of pixels for 100 turns and calculated what the z steps should be.
- calibrate the stepper motor voltage so that it doesn't produce noise when sitting still.
- use auto bed levelling as the z adjustment screw can be hard to set correctly. I use a 9g servo with microswitch with consistent results
- the z rod nut that goes into the x end and x motor mount is too loose sometimes. I use a small piece of rubber so it's stuck tighter in place. It can still come loose if the head crashes into the bed so no damage can occur. This helps auto bed levelling accuracy and consistency.
- calibrate the z offset when using auto bed levelling and measure the z height of one printed layer. I use a perimiter or skirt for this.
- make sure the z velocity is set to 3 instead of 5 which can cause missed steps. Make sure you set it in the correct place, depending on your configuration, within the firmware, recompile and upload, or using the display and control knob in control/motion. Make sure to store settings in control menu.
- lubricate all moving parts except the stepper motors, I think these are sealed anyway.
- calibrate the x axis being parallel with the print bed. I use a credit card slipped under the x smooth rod and on the bed with the 55mm side left and right. If it has about the same resistance you're golden.
- level the bed too of course, even when using auto bed leveling it's better to have it level in the first place. The bed, x rod angle and z offset can give an error of about 0.5mm, which can be the difference between parts just not fitting properly.

Hope this helps!

Good luck!