D-bot wobble !! :-(

No support, I hope so this speed no miss steps, I really would like like to speed it up but keep it like this the time of measurement (I'm not sure about the DRV8825)

You could loosen the bolts holding the ballnut and let it let it float a little, you'll get some backlash and you will have to adjust z home, remove any z hop but it may show if its out of alignment

Measure on the first 20mm, move the dialler at 10mm and do the second part of the measures.

Is the data plotted the error in each step or the total error at each position? In the plot I posted it was error in each 0.1 mm step, positive error meant the step was longer than 0.1 mm, negative meant the step was shorter than 0.1 mm.

The data looks periodic to me. I don't know what would cause that with screws unless the screw threads are poorly cut, but that seems unlikely.

Is there any vertical play in the screws? If you grab a screw and try pulling it up or pushing it down, does it move? What do the bottom bearings on the screws look like? Do the screws/pulleys appear to wobble as the bed moves up and down?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Measures of 20mm, at 10mm I changed the position of the comparator, difference between theorical position and reality

Clearly see repetitive patern



Edited 2 time(s). Last edit at 06/01/2018 08:43AM by jackantubis.

to test only the mecanical parts, I installed a dialer to check the position of each turn (to be sure to have a very precise turn each time) at each turn I measure the position of the bed, all of the movement made by hand:



on the excel sheet the different measures


for this one is the difference between the theorical position calculated for 1 turn and the position of the bed



Edited 2 time(s). Last edit at 06/01/2018 11:07AM by jackantubis.

Quote

To O lamp:

You think the position of the leadscrews will change the linearity of the movement you whant to place the leadscrew in front position aligned with the linear rail ? with this configuration you can have tilt on the bed ? (OK the linear rail could handle the torque) in the actual position there is the minimum effort on the carriage

Never mind! I thought the screws were side_by_side with the linear rails, but they are not ( according to the last pic, you posted )

measure of mouvement on 10mm ask 0.2mm by software and read on the bed 1 measure on the left of the bed and 1 on the right to check if the screws are synchronised: the screws are not completely synchronised, 1 step decay sometimes.



Edited 1 time(s). Last edit at 06/01/2018 10:59AM by jackantubis.

Closed of the good solution :-)

I found the problem with belt and 2 leadscrews and 2 rails each sides:

Each step one leadscrew start before the other and next step the other leadscrews catch up the difference :-O you can this this in the previous post.

Now I changed the design, it's a first test, I'll improve the rigidity and all will be perfect :-)

Something closed to this design



Viewed on openbuilds:

[openbuilds.com]


Now my design ongoing:

I also installed diode smoother on each motor, the noise and the quality improvement is here but not big, need to be checked on next prints.

So you're going to switch to a cantilevered bed?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Yes strong cantilever bed, the main problem was mechanical, I simplified it and it works (I prefer 3 leadscrews but.... :-( ) in my measurements, each time I removed a part in the mechanical chain the les uncertainty I had.

Edited 1 time(s). Last edit at 06/05/2018 02:23AM by jackantubis.