28 avril 2016

Continuing the search for good print

So, I ws not happy with this horizontal banding and very small waves on sides. so I went through a systematic and  looooooong test processing,  changing parameters and  hardware configuration one by one.  Now I have got more than 30 samples....
There is a good test on thingiverse for that : http://www.thingiverse.com/thing:277394
Most  samples are printer at  70mm/sec minimum

To summarize all changes I did:

  • marlin parameters tuning, especially "Accelerations" and "Jerk"
  • bed and Heater temperature
  • filament flow extrusion rate  (80-120%)
  • extrusion speed, retract speed
  • infill 0 or 25%
  • infill overlap
  • perimeter after infill
  • two types of motor controller (DRV8825, A4988)
  • 2, 8, 16 or 32 bits motor substepping
  • install the print on dampers to avoid vibration resonnance from the table
  • belt tensioning
  • install motors on antivibration support
  • print on various places on the bed
  • reinstall bowden vertically inside a cable chain
  • add some grease to Z axis
  • check all screws
  • reinforce bed structure
  • check and reduce extruder filament compression
and finally:

  • replace sliders with bronze bearing tube by new slider with linear bearing LM8LUU
  • replace Z axis leadscrew TR8x8 by TR8x2 

See pictures below, the black one has been printer with a small compact printer (minifabrikator)

01 avril 2016

Horizontal banding

I just spent lot of time trying to understand why I got horizontal banding on new parts with oboXL3D printer. Initially it looks like this:
 To compare with another print coming from Kossel printer (the orange part)

After many tests and assumptions like bad threaded rod on Z or not aligned, accuracy of XY, bad filament, bad heater, etc... I finally found the solution, thanks to a post found on a forum.
That post explains that this type of banding may come from the regulation of head bed temperature.

So, I followed the instructions and did this:
1/ uncomment #PIDTEMP Bed in Marlin configuration.h, to activate true PID regulation of the bad, instead of basic regulation (as defined by default)
2/ add correct values Kp,Ki,Kd (as computed automatically by Marlin). 

And it gave this, (on the left before tuning, on the right after tuning):

Thanks to those posts: