Welcome! Log In Create A New Profile

Advanced

Rostock : Layer shift

Posted by Repier37 
Rostock : Layer shift
February 21, 2016 11:08AM
Hi everyone,

As you can see on the picture attached, i have a problem with my rostock.
The part is supposed to be a cylinder, but you can notice that at some point, the layer start shifting a bit, and then coming back to the correct position.

This issue is happening at each print, on different part, always in the same direction (the shift seems to happen opposit to my X tower), so I think it's a mechanical problem.

I tought that my pulley was slipping on my X stepper axis so I've tighten again the pulley and I still have the problem.
I have tried printing at lower speed, lower acceleration, lower jerk without any change.
I don't feel any "hard point" on my smooth rod when moving the carriage by hand.

I don't think that backlash is the problem, as my infill and my perimeters are touching each other without any "infill perimeter overlap", so i am guessing that I have very low (if no) backlash. But I might be wrong

So if anyone have any idea, please tell me smiling smiley
Attachments:
open | download - Sans titre.jpg (564 KB)
Re: Rostock : Layer shift
February 21, 2016 12:10PM
My guess is that it is missing a small number of steps (perhaps just one microstep) during travel moves, and that the travel moves are not always in the same direction. This could be caused by the motors not being able to accelerate fast enough (so try lower acceleration or jerk settings, or higher motor current if appropriate), or by a firmware bug. What electronics and firmware are you using?



Large delta printer [miscsolutions.wordpress.com], Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Rostock : Layer shift
February 21, 2016 12:29PM
I am using ramps 1.4 + repetier firmware 0.92
My acceleration settings are :
Acceleration : 50mm/s²
Travel acceleration : 60mm/s²
max jerk : 20mm/s

I think these are quit low already don't you think ?

I've calibrated the stepper current using this guide : [reprap.org]
and raise it a bit after (1/8 turn)
Re: Rostock : Layer shift
February 21, 2016 01:09PM
Quote
Repier37
I've calibrated the stepper current using this guide : [reprap.org]
and raise it a bit after (1/8 turn)

Those instructions for setting motor currents are easy for novices to use, but will often result in the motor current being too low to support good acceleration and precise positioning. Here is a better approach:

1. Check your motor specifications to see what the rated current per phase is.

2. Choose a current setting that is between 50% and 85% of the motor current rating, and not so high that your drivers will overheat.

3. Establish whether your stepper driver modules have 0.05 or 0.1 ohm sense resistors, hence calculate the Vref voltage needed to achieve that current (for most drivers, Vref = 8 * sense_resistor_value * desired_motor_current).

4. Use a voltmeter to set the voltage on the pot wiper to that Vref.

A less accurate method:

1. Do what you have done already.

2. If the pots are not all at the same position, turn up the 2 lower ones to the same position (I am assuming all 3 drivers and all 3 motors are of the same types).

3. Disable idle hold mode in firmware, home the printer to energise the motors, and leave for a few minutes. Check that the carriages are still held in position (to make sure idle hold mode really has been disabled), then check the temperatures of the drivers and motors. If the motors are cold and the drivers are not too hot to touch, turn up the pots a little, and repeat. You are aiming for the motors to run a little warm but not hot, and for the drivers to not be overheating.

Or use electronics with software-selectable motor currents, then you can just set the currents to the desired value.



Large delta printer [miscsolutions.wordpress.com], Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Rostock : Layer shift
February 21, 2016 01:13PM
PS - I think your acceleration settings are rather high. Try setting both accelerations to 15mm/sec^2 and jerk to 10mm/sec. If that improves the print, set the motor currents to a more optimal value (see my previous post) before you try increasing them again.



Large delta printer [miscsolutions.wordpress.com], Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Rostock : Layer shift
February 22, 2016 08:22AM
Ok, great informations !! Seems to be a way more accurate calibration method smiling smiley
I'll try that ASAP and will keep you informed
Re: Rostock : Layer shift
February 22, 2016 01:22PM
The fact that it does it the same way on multiple prints make me want to agree with you that it's likely mechanical.

Motor currents and acceleration settings might make you miss steps, but I don't expect it would be so repeatable.

So I'm trying to think of what might cause a non-linearity in your vertical motion. I don't think it's a gear or bearing that's out-of-true, because the oddity would likely happen more often (like, whatever distance it takes for your motor to make a full rotation).

But what if one of your belts isn't straight? I'm thinking maybe the carriage is too far in or out, or to one side or the other. I could see that causing two things:
  1. The tension on the belt would increase as you got closer to either end, as the angle got more acute in comparison with where it wants to be.
  2. The steps per mm would shift as there's an angle where the belt should be straight.

It may not be what's happening, as I think that might cause more of a curved slant in the print, but it's worth looking at.
Re: Rostock : Layer shift
February 22, 2016 03:04PM
Quote
maso
The fact that it does it the same way on multiple prints make me want to agree with you that it's likely mechanical.

On reflection, I think you are probably right and it has nothing to do with missed steps.

Does this machine use linear guides or wheeled carriages? If it uses the typical 3-wheel carriages, then I think a variation of the width of the slots in the extrusion that the wheels run in could cause the carriage to rotate a little as it moves up the tower, which would cause this type of problem.



Large delta printer [miscsolutions.wordpress.com], Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Rostock : Layer shift
February 22, 2016 04:55PM
Thanks for your help.
I am using LM8UU on inox rod. It's quite a "standard" rostock, I have just made a few changes on the wood frame, but as I have been able to make a correct vertical print a few weeks ago I think my frame is ok.

On the 3 tower, when there is a change in the carriage direction I can see the belt turning a bit on itself (something like 5° maybe). If this is making you think of something...

I didn't had time to print again yet, but I will check my pulley again tomorrow as it's seems that it's the most common solution for that kind of problem.
Re: Rostock : Layer shift
February 25, 2016 01:50AM
So, as it's seems that the problem was mechanical, I've run a print with high acceleration to accentuate the problem (or to brake the printer). I've put some marker on my pulley and my stepper axis to check any displacement between them. And there was a displacement.
I've tightened my pulley again, run another print, it was ok at first and then I saw a displacement again. The screw was getting loose (due to vibration I guess).
So I removed the screw, put some super glue on it, and tightened it again. And It seems like the problem is solved.

Thanks everyone smiling smiley

Edited 1 time(s). Last edit at 02/25/2016 10:35AM by Repier37.
Re: Rostock : Layer shift
February 25, 2016 03:27AM
Glad you solved it! Do your motor shafts have flats on them?



Large delta printer [miscsolutions.wordpress.com], Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Re: Rostock : Layer shift
February 25, 2016 05:37AM
Yes, they have, and I think that's why the shift was oscillating. The screw was just loose enough to allow a movement on the flat part but wasn't loose enough to allow a complete rotation of the pulley.
Re: Rostock : Layer shift
February 25, 2016 10:33AM
Yeah, I'm glad you solved it too!

It's funny--would've been way more obvious if the shafts hadn't had flats.


And it gives me one more thing to watch for when I see weirdness somewhere.
Sorry, only registered users may post in this forum.

Click here to login