Notifications
Clear all

Print quality problem - overextrusion?  

  RSS
Jan Pichrt
(@jan-pichrt)
Member
Print quality problem - overextrusion?

Hi all, i have perfectly builded MK3 (perpendicular and all tests OK), but when I print something with flat top surface, the surface is it not smooth. It looks like overextrusion problem. But when I measured my extrusion, I found that it was set up correctly (measured 100mm and extruded 100mm).

Tested on sliced objects from Prusa Control, Prusa Slic3r and even with presliced files like Prusa logo on SD card. No changes in print profiles. All def.

What can be wrong here ❓ 🙄

Opublikowany : 18/12/2017 2:33 am
Paul Meyer
(@paul-meyer)
Honorable Member
Re: Print quality problem - overextrusion?

The interesting part is that the lower ‘top’ surfaces seem fine (especially in the first pic). Is your fan working? For the first picture I would think it was a cooling issue (fast layer). That wouldn’t explain the later pictures, though.

Suggestions: try it at lower temp and slower, see how it looks.

Opublikowany : 18/12/2017 2:53 am
Jan Pichrt
(@jan-pichrt)
Member
Topic starter answered:
Re: Print quality problem - overextrusion?

Yes, both fans working...

Opublikowany : 18/12/2017 3:29 am
Impruve
(@impruve)
Estimable Member
Re: Print quality problem - overextrusion?

The prusa logo doesnt look that bad....

Any other pictures from other prints?

Opublikowany : 18/12/2017 11:09 pm
Paul Meyer
(@paul-meyer)
Honorable Member
Re: Print quality problem - overextrusion?

Agreed. Try printing the benchy or other simple print that is pre-slice on the SD card that came with the printer, using the filament delivered. That will eliminate slicing issues and (hopefully) filament issues. If the benchy looks good and your own prints don't, it's time to look into your slicing.

Opublikowany : 18/12/2017 11:20 pm
gorkish
(@gorkish)
Eminent Member
Re: Print quality problem - overextrusion?

I am getting quite a bit of this too. The demo files printed pretty good (not perfect but definately no slicing issues), but my own files sliced with default settings at DEFAULT and OPTIMAL profies were absolute shit.

The first problem which I corrected was that the version of Slic3r distributed in the drivers 2.0.0 package (for mac did not check windows) did not support the conditional gcode used in the filament profiles, so in my own sliced files the linear advance (M900 K30 command) was never being set. From what I can tell this would affect everyone who is following the Mk3 instructions to the letter. Those taking the additional step of downloading Slic3r separately are probably OK.

Anyway, prints improved quite a bit when I upgraded Slic3r to the newest release from GitHub Releases (1.38.4). I also made some tweaks to the settings based on the settings used in the sample files. Chiefly this was adjusting external perimeter extrusion width to 0.42 and reducing infill overlap to 25%. I no longer have as much blobbing where my infill meets the perimeters which seems to help the most with walls that have a tiny amount of infill (like in your example) although they are still not very good.

The major remaining problem is that the nozzle is chattering as it catches on the infill while printing it. On cubic it hits it on the third and sometimes second angle passes. I cannot find anything particularly different between my settings and the samples that should affect this as I'm using the same cubic infill, same percentage density, extrusion width, etc. but I'm just having a terrible problem with it.

It does seem like overextrusion to some extent, but I am reluctant to point the finger at that because many other classic signs of overextrusion are absent. I initially brought up my printer with the 'buggy' firmware that was a little bit 'crashy' during the wizard, so i wonder if I missed something in the calibration. I did not see anything about calibrating the extruder, and I would presume this ought not to be necessary. My error counters are all pegged at 255 though which does not seem to be right.

I am just hoping that the firmware and profiles will get this all sorted out over the next few weeks.

Opublikowany : 19/12/2017 12:01 am
Impruve
(@impruve)
Estimable Member
Re: Print quality problem - overextrusion?

I tried printing two of these to see if they snapped together. https://www.thingiverse.com/thing:1622868

And they dont, at least not easy enough. I still didnt check the e-steps yet (anyone knows the default value? If not i will check with pronterface later) but maybe the default e-steps should be adjusted and calibrated??

Didnt see any option to change that on the printer itself.

Opublikowany : 19/12/2017 2:59 am
gorkish
(@gorkish)
Eminent Member
Re: Print quality problem - overextrusion?

I worked on this a little bit more and I have now gotten my settings tuned for my machine well enough that I am able to slice and print the sample models in a way that the quality exceeds that which the sample gcode produced.

Obviously getting Slic3r updated to 1.38.4 and onto the most current MK3 and filament profiles is critical since the profiles shipping with the 2.0.0 drivers package are both slightly out of date and also include the filament profiles with conditional gcode that doesnt work on the older Slic3r release that is distributed with it. That is step #1.

I was going to look at calibrating the esteps but when I asked the extruder for 100mm it was off by no more than a mm at most, which seems like an unnecessary adjustment. I did measure filament diameter at a bit of error - seemed to be between 1.76 and 1.78mm in diameter.

However rather than waste a bunch of effort to determine which setting was wrong, I did realize that really when everything is said and done, so long as the esteps/mm and filament diameter and nozzle size and linear advance and all such similar settings are all nominally correct, it is a waste of time to seek out what exactly is wrong since all of the fine tuning can be taken care of at once by adjusting the extrusion multiplier, and indeed that is all it took.


I tried printing two of these to see if they snapped together. https://www.thingiverse.com/thing:1622868

Thanks very much to leonardo.g3 for linking this model. Although it doesnt seem really to be a popular calibration sample, it is honestly very well thought out part to do this work. If you have calipers and print two rounds of 2 parts each you can rapidly converge an ideal extrusion multiplier for any new roll of filament in about 20 minutes by iterating your extrusion multiplier. I made a spreadsheet to easily do this using this model that I will clean up and release. I know calibrating the extrusion multiplier isnt exactly a new thing, but I do like this method as it can be performed regardless of xy dimensional calibration.

So my final conclusion was that in my case I was overextruding by about 6% and probably that is due to error accoumulated from filament diameter, extruder gear diameter, extruder tension, and nozzle size variance. My top layers are now almost as smooth as the bottom layers and much better than the flat top layers in the prusa and whistle sample gcodes.

Opublikowany : 20/12/2017 11:10 pm
kelchm
(@kelchm)
Eminent Member
Re: Print quality problem - overextrusion?

FWIW, I've always found the need to tune the extrusion multiplier based on the type of filament. In the past it hasn't been unusual for me to use an extrusion multiplier between .90 and .92 for PLA. PETG I usually run much closer to 1.0.

Opublikowany : 21/12/2017 12:07 am
Impruve
(@impruve)
Estimable Member
Re: Print quality problem - overextrusion?


I worked on this a little bit more and I have now gotten my settings tuned for my machine well enough that I am able to slice and print the sample models in a way that the quality exceeds that which the sample gcode produced.

Obviously getting Slic3r updated to 1.38.4 and onto the most current MK3 and filament profiles is critical since the profiles shipping with the 2.0.0 drivers package are both slightly out of date and also include the filament profiles with conditional gcode that doesnt work on the older Slic3r release that is distributed with it. That is step #1.

I was going to look at calibrating the esteps but when I asked the extruder for 100mm it was off by no more than a mm at most, which seems like an unnecessary adjustment. I did measure filament diameter at a bit of error - seemed to be between 1.76 and 1.78mm in diameter.

However rather than waste a bunch of effort to determine which setting was wrong, I did realize that really when everything is said and done, so long as the esteps/mm and filament diameter and nozzle size and linear advance and all such similar settings are all nominally correct, it is a waste of time to seek out what exactly is wrong since all of the fine tuning can be taken care of at once by adjusting the extrusion multiplier, and indeed that is all it took.


I tried printing two of these to see if they snapped together. https://www.thingiverse.com/thing:1622868

Thanks very much to leonardo.g3 for linking this model. Although it doesnt seem really to be a popular calibration sample, it is honestly very well thought out part to do this work. If you have calipers and print two rounds of 2 parts each you can rapidly converge an ideal extrusion multiplier for any new roll of filament in about 20 minutes by iterating your extrusion multiplier. I made a spreadsheet to easily do this using this model that I will clean up and release. I know calibrating the extrusion multiplier isnt exactly a new thing, but I do like this method as it can be performed regardless of xy dimensional calibration.

So my final conclusion was that in my case I was overextruding by about 6% and probably that is due to error accoumulated from filament diameter, extruder gear diameter, extruder tension, and nozzle size variance. My top layers are now almost as smooth as the bottom layers and much better than the flat top layers in the prusa and whistle sample gcodes.

Happy i could help.... Lets see that spreadsheet when its done hehe =)

Opublikowany : 21/12/2017 2:41 am
gorkish
(@gorkish)
Eminent Member
Re: Print quality problem - overextrusion?


Happy i could help.... Lets see that spreadsheet when its done hehe =)

My calculator sheet using this calibration object is now posted

Opublikowany : 21/12/2017 5:36 pm
Impruve
(@impruve)
Estimable Member
Re: Print quality problem - overextrusion?



Happy i could help.... Lets see that spreadsheet when its done hehe =)

My calculator sheet using this calibration object is now posted

Nice work.

Let's hope it helps everyone dial in the Extrusion Multiplier =)

Opublikowany : 22/12/2017 1:22 am
GertL
(@gertl)
Reputable Member
Re: Print quality problem - overextrusion?

Nice initiative. Need to try this.

---
Gert

Opublikowany : 22/12/2017 9:03 am
Share: