Hi all!
Today i faced a strange problem.
I wanted to check my Printer calibration and printed a Calibration pyramid.
I should have 50m by 50mm by 50mm.
In x and y direction calibration is almost on the spot ; )
but …
In Z direction the first pyramid layer has about 9 mm
the second has 8 mm
the third 4 mm
the fifth 8mm
WTF…?
Seems that my Z axis is quite buggy.
What can i try to fix this?
I already set the VRef of the drivers to .55v
Reduced acceleration to 200mms² what did the trick for X and Y
But what happenedn to my Z axis now?
Any Help appreciated.
cheers
Christian
Just a guess but is the project being sliced properly?
Yes, the slicing is ok.
I managed to gett better results by reducing the Z axis jerk value in Marlin Eeprom.
Way better but still not completely correct.
I think the stiff coupler between Z motor and screw is the caus as it makes the whole thing wobble around.
A flex coupler is already ordered and i will try it later tis day if it arrives.
Maybe that will work better.
What puzzles me is that tis issue didn’t show up till yesterday.
I did a lot of test prints that were all quite ok.
Okay, some news.
I now got my brand new flex cupler for the Z axis mounted.
Result :
-
Z axis wobbling completely gone!
-
Z banding almost gone too. (better results will only be possible with a trapezoid spindle i think)
-
Z axis accuracy now also nearly spot on and no lost steps anymore!
I think the problem was the holding torque the wobbly threades screw caused when it was stopped.
With my actual setup the Z motor doesent even move within its fittig, while before it was wobbling around like millimeters!
I will test going back zo original Z axis stepper firmware settings, as i assume they will work now too.
I installed a trapezoidal spindle, and you are correct that it solved my z-axis banding problem. I also installed a flexible coupling.
One note of advice to those working on their z axis. It is very difficult to constrain the threaded rod (or trapezoidal spindle) in three places and not experience binding of the rod. The binding increases the force required to rotate the rod or spindle, and can even cause motor skips. I opted for “play” in the follower nut so that it does not cause the rod to bind.
Backlash is not a concern because the z axis always approaches a new layer from the same direction. But too much backlash will cause the first layer thickness to be off because the home position can only be approached from above.
[quote=“artcominio”]I installed a trapezoidal spindle, and you are correct that it solved my z-axis banding problem. I also installed a flexible coupling.
One note of advice to those working on their z axis. It is very difficult to constrain the threaded rod (or trapezoidal spindle) in three places and not experience binding of the rod. The binding increases the force required to rotate the rod or spindle, and can even cause motor skips. I opted for “play” in the follower nut so that it does not cause the rod to bind.
Backlash is not a concern because the z axis always approaches a new layer from the same direction. But too much backlash will cause the first layer thickness to be off because the home position can only be approached from above.[/quote]
I think, using a flex coupler, binding should not be a problem.
You might have to loosten the motor mount screws for allowing the motor to adjust itself with the threaded rod.
Then tighten it up again and it should work ok.
Assumed the Top mount for the Z bearing and the carriage mount for the nut are not way off in alignment.
If i switch to a tapezoid spindle i will get one with a separate loose nut and make an adjustable mount for the carriage.
So anything can be alingned straight.