Print quality issues - banding
Hi everyone,
After putting together my i3 MK2 kit and printing a few basic parts I start noticing some banding on the larger prints.
I printed a cylinder and a cube to test this out and here's what I'm getting.
Notice the vertical straight ridge on the cylinder (red arrow) and the banding (blue arrows).
The first layer on this particular model is a bit squashed because I was testing a new value for the Z offset but I get similar results on prints with better first layer.
The vertical ridge seems to be happening at the beginning of the perimeter.
The banding is also very visible on the test cube.
Not sure if this is related to the print issues, but I noticed that the right Z stage threaded rod is jerking towards the frame whenever is changing the layer height. At rest, the right stepper rod is leaning towards the top bracket, compared to the left rod which is more centered.
I've uploaded a video here:
https://drive.google.com/file/d/0B4reMI0d3sFaT2JETE5OTEgwRWM/view?usp=sharing
Both models were sliced with Slic3r (on Ubuntu) using the 0.2 NORMAL profile from the latest Mac sw download.
I only changed the bed temp to 60C to increase print bed adhesion.
During XYZ calibration I got "XYZ calibration all right. X/Y axes are slightly skewed." which sounded OK to me so I haven't spent time yet to fix the slight skew yet. The printer sits without wobble on the original printed feet (with pads) on a sturdy table.
Any tips on what should I be looking at for fixing this problem ?
Thanks!
Re: Print quality issues - banding
And here's the cube I was talking about, looks like it wasn't attached on the first post...
Re: Print quality issues - banding
First thing I would check is the tension of the extruder springs.
with filament inserted, the springs (only the spring) should be 13mm.
beyond that, maybe someone else has more ideas.
Re: Print quality issues - banding
OK, I made some progress in troubleshooting my print artifacts.
Some of the things I tried:
1. Checked the extruder springs length based on @mavu's suggestion. Turns out they were a bit short of 13mm but not by much.
I made sure they are 13mm exactly but this didn't make any difference. However, I think your hunch that this is somehow related
to extruding consistency was correct!
2. I noticed that the smooth rods of my vertical stage are sticking out a bit and the left one was ~1mm lower in the top bracket.
I knew that I haven't managed to push the rods all the way into the motor brackets during assembly but at the time it looked good enough.
I loosened the tension on the X stage and pulled the rods out of the brackets, cleaned up the lower bracket mounting holes using a 5/16" drill bit
and put everything back together. I took my time aligning the top brackets and measuring the height between the two brackets on both sides to
make sure they are equal. In the end the smooth rods are now flush with the top mounting brackets.
I rerun XYZ calibration and now I get " XYZ calibration ok. X/Y axes are perpendicular". Yay!
However, a new test print of the Cylinder of Doom showed just the tiniest of improvements. So small in fact, that in I didn't even bother to take pictures because I didn't think it would show up well enough.
Assorted musings on the Z stage:
- after aligning the Z stage much better, the right threaded rod is still almost touching the bracket but I don't think that's a problem.
It looks like by design it has enough room to wiggle around and it's not actually putting any pressure on the bracket.
- The jerks I noticed are happening during the "lift Z" step, which is 0.5mm on my Slic3r default settings.
I don't think this is the source of the vertical ridge artifact (see below).
- I realized that what actually stops the Z stage are the nuts of the top brackets and not the brackets themselves!
This is a bit surprising because the final orientation of the hex nuts can add a little bit to the height difference between the left and right columns.
Probably insignificant, but I haven't noticed this at first.
3. I changed the extrusion speed to 80% thinking that slower is better... no difference whatsoever, in fact this particular test print was worse than most.
4. After staring at the wavy prints for a while I realized that the banding is periodic and most likely due to inconsistent extrusion between layers.
I thought about thermal variation because I print in my garage and the ambient temp can drop rapidly during the evening. In any case, without an enclosure there's not much I can do about this because the controller seems to be doing a decent job at keeping the temp within 2degC of target which should be OK.
I then realized the extruder has to work extra hard sometimes to pull the filament off the spool and when I checked the spool holder it turns out the spool was not rolling smoothly on the original holders! The spool holder brackets fit very tightly on the frame and it's easy to press them together enough to make the spool hard to roll!
After I spread apart the holder brackets, the spool rolls much easier and I get much better looking prints!
A spool holder upgrade (one of the designs using bearings) is next on my list!
See attached images for some examples. The vertical banding is still present but with much smaller amplitude and the prints feel much smoother on the finger. For now I'm content with this print quality, hopefully it stay stays consistent!
5. Now, the vertical "ridge" did not disappear! It's somewhat less pronounced after the latest fixes but still present.
It happens at the perimeter seam but not everywhere throughout a more complex model . See for the example the shark print where the seam disappears at one point. On the test cylinder it's still present from top to bottom.
Also, it's actually a small notch instead of a raised ridge on the model, which makes me believe it has something to do with under extrusion at the end of the perimeter.
I found the retraction length parameter in Slic3r and I'll play with it next to see if it makes any difference. Most likely this effect is also filament dependent. I don't remember seeing the seam very pronounced when printing with the orange Prusa PLA.
In any case, I can live with this print quality for my hobby uses.
If anyone has any ideas for further improvements, in particular about the perimeter seam artifact, please let me know!
Re: Print quality issues - banding
Attaching two more pictures of the shark because I think it turned out quite nicely!
Re: Print quality issues - banding
never thought that the spool holder will have such a visible effect at the print. thanks for figuring that out...
dem inscheniör is' nix zu schwör...
Re: Print quality issues - banding
I also tried to remove your red marked seam. No way. From my tests it is NOT the retraction.
The fact that it disappears on the shark only means (guesswork) it is inside the mouth now.
I also played with other parameters, but so far no luck,
Wild guess again:
Fact: It is caused by the layer change,
Guess: There the extruder moves a right angle "inside" the object and as we know we cant print exactly sharp corners and also the start of a perimeter will not be that sharp, so there meet two perimeters with a round corner leaving a seam.
Carsten
My Prints: https://www.prusaprinters.org/social/15695-carsten/prints
My Employer: https://make-magazin.de
Re: Print quality issues - banding
Hi, did you try to change the value "lift Z" to 0 or 0.15 in printer settings ( 0.15 is like all the Gcode on the SD with your prusa) ? with Prusa Slic3r it's set to 0.5...
And did you try to change the perimeter starting point ( print setting > layer and perimeter > advanced >Seam position) to random, cause nearest is set, but nearest is for angle, if you print a circle all the retract point are to the same position ~....like your shark
Re: Print quality issues - banding
Hi, did you try to change the value "lift Z" to 0 or 0.15 in printer settings ( 0.15 is like all the Gcode on the SD with your prusa) ? with Prusa Slic3r it's set to 0.5...
And did you try to change the perimeter starting point ( print setting > layer and perimeter > advanced >Seam position) to random, cause nearest is set, but nearest is for angle, if you print a circle all the retract point are to the same position ~....like your shark
Great suggestions! I stopped printing the calibration objects for a while 😀
I tried 0 for "lift Z" but the print stopped after the second or third layer (consistently, I've tried multiple times).
I didn't know the SD objects use 0.15 Z lift, I'll have to try that setting as soon as possible!
I'll be traveling for a few days but will be trying this over the weekend.
Re: Print quality issues - banding
Hi, did you try to change the value "lift Z" to 0 or 0.15 in printer settings ( 0.15 is like all the Gcode on the SD with your prusa) ? with Prusa Slic3r it's set to 0.5...
And did you try to change the perimeter starting point ( print setting > layer and perimeter > advanced >Seam position) to random, cause nearest is set, but nearest is for angle, if you print a circle all the retract point are to the same position ~....like your shark
Great suggestions! I stopped printing the calibration objects for a while 😀
I tried 0 for "lift Z" but the print stopped after the second or third layer (consistently, I've tried multiple times).
I didn't know the SD objects use 0.15 Z lift, I'll have to try that setting as soon as possible!
I'll be traveling for a few days but will be trying this over the weekend.
Any luck with solving your banding issue?
Re: Print quality issues - banding
Hi there,
I'm having the same issue here with the seamings. This is not bothering me that much, but I would like to improve their looking...
I saw in another forums that overlapping the beginning with ending (from 0.2 to 0.8mm) may smooth it, but I can't find such parameter in Slic3rPE, it may be a firmware tweak.
Any improvement with other kind of settings?
Cheers.