Notifications
Clear all

poor print quality  

Page 11 / 13
  RSS
Brigandier
(@brigandier)
Reputable Member
Re: poor print quality

Hey guys,

Had my best print improvement yet:

As you can see, I have it pretty much down to true ringing instead of the predictable vertical lines. Getting rid of that repeating line is thanks to RHDreambox, see my solution here: https://shop.prusa3d.com/forum/original-prusa-i3-mk3-f30/fine-tune-your-mk3-y-axis-t13851-s30.html#p65137

tl;dr, the Y axis motor mount was warped! My stepper motor was twisted inward, the backside was away from the metal frame by 1.5mm making the already warped Y stepper mount a pivot/vibration point as well as causing the pulley to be against the belt unevenly. Bad quality control on the Prusa printed parts...

Now to get my accel/jerk settings right, now that I can actually discern them from that freaking repeating line. 😀

EDIT: Before anyone mentions the corner bulge, this appeared when I swapped to IGUS. Going to swap in some better metal bearings when I get a chance. Other settings I have made include swapping to tooth'd idlers (little effect, if any). Belt twist was reverted back to normal, and tension back to normal, if anyone has been tracking my testing.

My MK3 Parts: [Bowden] [New Shoes] [TPU Micro Springs]

Posted : 08/02/2018 4:53 am
jediknight0
(@jediknight0)
Active Member
Re: poor print quality


EDIT: Before anyone mentions the corner bulge, this appeared when I swapped to IGUS.

I built my MK3 out of the box with IGUS, and I see the same thing. I think it was better when I ran with firmware with LA, but maybe I was imagining it.

Did you have a heck of a time finding the right tightness for the x-carriage with IGUS bearings? Too tight and the printer wouldn't pass the selftest. Too loose and the bearings slide around inside...

Posted : 08/02/2018 6:25 am
Brigandier
(@brigandier)
Reputable Member
Re: poor print quality



EDIT: Before anyone mentions the corner bulge, this appeared when I swapped to IGUS.

I built my MK3 out of the box with IGUS, and I see the same thing. I think it was better when I ran with firmware with LA, but maybe I was imagining it.

Did you have a heck of a time finding the right tightness for the x-carriage with IGUS bearings? Too tight and the printer wouldn't pass the selftest. Too loose and the bearings slide around inside...

Yes, X was very hard to get working with IGUS. Y was easier. I forgot about LA as well, seems more likely.

My MK3 Parts: [Bowden] [New Shoes] [TPU Micro Springs]

Posted : 08/02/2018 12:36 pm
pentrena
(@pentrena)
Eminent Member
Re: poor print quality

Does anyone know what these marks are that my mk3 does? it is printed at 0.1 layer and the perimeters at 35 mm / sec

Posted : 08/02/2018 10:42 pm
ed
 ed
(@ed-3)
Reputable Member
Re: poor print quality

Could the problem be with the slicing software? I have a calibration cube sliced in Simplify3D that prints really nice while the one sliced using Slic3r looks similar everyone else's. I'm going to post the gcode for the Simplify3D sliced calibration cube if anyone else wants to compare but at this point I'm seriously considering purchasing a copy of Simplify3D... Any thoughts?

Simplify3D 20mm test cube gcode... https://drive.google.com/file/d/1AfeG4CZ6xmGZrLxk4nLwzAY0kpLVR1fq/view?usp=sharing

I have no access to Simplify3D the gcode was given to me so I'm not sure what settings were used other than I guess it's possible to look at the file and see. Or do I just not understand the settings in Slic3r enough to set it up to get the results as those from Simplify3D. Ideally I'd prefer not to shell out $150.00

Left is Slic3r right is Simplify3D...

Posted : 10/02/2018 5:14 pm
stefan.p3
(@stefan-p3)
Active Member
Re: poor print quality

Left is Slic3r right is Simplify3D...
Image

mhhh, why can‘t I see the pics... all I see is just a box saying „image“. (on iPad)

Posted : 11/02/2018 2:48 am
Ed Tumbusch
(@ed-tumbusch)
New Member
Re: poor print quality


Could the problem be with the slicing software? I have a calibration cube sliced in Simplify3D that prints really nice while the one sliced using Slic3r looks similar everyone else's. I'm going to post the gcode for the Simplify3D sliced calibration cube if anyone else wants to compare but at this point I'm seriously considering purchasing a copy of Simplify3D... Any thoughts?

Simplify3D 20mm test cube gcode... https://drive.google.com/file/d/1AfeG4CZ6xmGZrLxk4nLwzAY0kpLVR1fq/view?usp=sharing

I have no access to Simplify3D the gcode was given to me so I'm not sure what settings were used other than I guess it's possible to look at the file and see. Or do I just not understand the settings in Slic3r enough to set it up to get the results as those from Simplify3D. Ideally I'd prefer not to shell out $150.00

Left is Slic3r right is Simplify3D...

I also noticed that S3d was providing nicer looking prints than Slic3r so I decided to look into why. I had been printing calicat ( https://www.thingiverse.com/thing:1545913 , so i could eventually give the prints to my daughter :P) and took a look at the gcode.

When printing things that are small, the slicer slows down each layer to make sure there is enough time for cooling. Simplify3D seems to slow down each of the features (outer perimeter, inner perimeter, infill, ...). Slic3r seems to slow down some features, but NOT the external perimeters. In the file that you provided, the external perimeter in a random layer from the middle prints at around 14mm/s.


...
; outer perimeter
G1 X115.240 Y95.240 F12000
G92 E0
G1 X121.410 Y95.240 E0.2339 F818
...

See that last line with "... F818"? That is the feedrate in mm/minute for the outer perimeter. On the same layer of slic3r gcode, I have observed the full 40mm/s (default) for the external perimeters, while other features are slowed down. In my experience, setting the speed of the outer perimeters in Slic3r to 20 gave similar results to S3D.

I'm not planning on printing that slow, but my guess is that if you had a larger print that didn't slow down because of the cooling, you would likely see the same artifacts from S3D.

Posted : 11/02/2018 6:05 am
Brigandier
(@brigandier)
Reputable Member
Re: poor print quality


See that last line with "... F818"? That is the feedrate in mm/minute for the outer perimeter. On the same layer of slic3r gcode, I have observed the full 40mm/s (default) for the external perimeters, while other features are slowed down. In my experience, setting the speed of the outer perimeters in Slic3r to 20 gave similar results to S3D.

I'm not planning on printing that slow, but my guess is that if you had a larger print that didn't slow down because of the cooling, you would likely see the same artifacts from S3D.

I noticed this behavior as well, but it's less about the speed and more about the acceleration. Been able to cut out almost all of the ringing that you can see in the above pics by dropping accel from 600mm/s to 100mm/s in Slic3r, maintaining the 45mm/s external perimeter speed. It just takes a bit longer to get there. Also if you are dealing with the vertical lines no matter what speed, check your Y axis motor mount. The stepper may not be against the metal frame.

My MK3 Parts: [Bowden] [New Shoes] [TPU Micro Springs]

Posted : 11/02/2018 6:16 am
Ed Tumbusch
(@ed-tumbusch)
New Member
Re: poor print quality


I noticed this behavior as well, but it's less about the speed and more about the acceleration. Been able to cut out almost all of the ringing that you can see in the above pics by dropping accel from 600mm/s to 100mm/s in Slic3r, maintaining the 45mm/s external perimeter speed.

Interesting, I'll give that a shot. I assume you just changed the accel for the external perimeters? Interestingly I just flashed your Linear Advance build. Without LA, do you get increased bulging in the corners with that slow of acceleration?

Posted : 11/02/2018 6:28 am
Brigandier
(@brigandier)
Reputable Member
Re: poor print quality



I noticed this behavior as well, but it's less about the speed and more about the acceleration. Been able to cut out almost all of the ringing that you can see in the above pics by dropping accel from 600mm/s to 100mm/s in Slic3r, maintaining the 45mm/s external perimeter speed.

Interesting, I'll give that a shot. I assume you just changed the accel for the external perimeters? Interestingly I just flashed your Linear Advance build. Without LA, do you get increased bulging in the corners with that slow of acceleration?

Yes, and yes. It's what prompted me to look into Linear Advance again. 🙂

My MK3 Parts: [Bowden] [New Shoes] [TPU Micro Springs]

Posted : 11/02/2018 6:31 am
Ed Tumbusch
(@ed-tumbusch)
New Member
Re: poor print quality


Yes, and yes. It's what prompted me to look into Linear Advance again. 🙂

Ok, now i need to stay up for another text print 😆

Posted : 11/02/2018 6:40 am
Brigandier
(@brigandier)
Reputable Member
Re: poor print quality



Yes, and yes. It's what prompted me to look into Linear Advance again. 🙂

Ok, now i need to stay up for another text print 😆

Will be interested to see your results, I'm not certain this was the best fix but it's working for me so far. I was banging my head against the wall trying to figure out why my external perimeters were going so fast on the XYZ test cube and came to the same conclusion that you did, Slic3r likes to keep the external perimeter speed fixed. I suspect this is so you get identical appearance all the way up the print.

100mm/s fixed my ghosting/ringing on the XYZ cube, but it's still lightly there on the curved faces of a benchy. I think it's something to do with the slicer thinking it can conserve momentum, going to try it even slower (say, 60mm/s) and see how that goes. Linear Advance seems to be working pretty well for me, hopefully it can compensate.

My MK3 Parts: [Bowden] [New Shoes] [TPU Micro Springs]

Posted : 11/02/2018 6:45 am
Brigandier
(@brigandier)
Reputable Member
Re: poor print quality

https://github.com/prusa3d/Prusa-Firmware/pull/475

Those of us with the vertical line issues, rejoice, and go support XPila with praise. Looks promising. 🙂

My MK3 Parts: [Bowden] [New Shoes] [TPU Micro Springs]

Posted : 17/02/2018 2:39 pm
thomas.h19
(@thomas-h19)
Active Member
Re: poor print quality

Hi Brigadier,

Would 'vertical line issues' be like in these pictures?
These images are a bit older and I've changed some settings and not reprinted this particular object again yet.

Thanks

Posted : 20/02/2018 3:55 am
Brigandier
(@brigandier)
Reputable Member
Re: poor print quality

The lines decreasing from the corners onward are classic ringing if you ask me. If I had to guess, is that the X axis? The second pic with the repeating lines definitely looks like the same issue I am having on the Y axis.

I haven't had time to play extensively with the microstep linearity settings, but it has me hopeful.

My MK3 Parts: [Bowden] [New Shoes] [TPU Micro Springs]

Posted : 20/02/2018 4:16 am
thomas.h19
(@thomas-h19)
Active Member
Re: poor print quality

These are perpendicular sides of the same object, if I recall correctly the top image with one window was parallel to the x axis.

Posted : 20/02/2018 5:52 am
Brigandier
(@brigandier)
Reputable Member
Re: poor print quality


These are perpendicular sides of the same object, if I recall correctly the top image with one window was parallel to the x axis.

I fought with this hard and can still see it when printing with glossy filament like PETG. Some things I found that helped:

  • Check the Y stepper mount. Mine was warped so that the stepper wasn't square with the frame of the printer. This caused a pivot point to vibrate at, and a tilted pulley.

  • If you can spring a few bucks for a toothed idler, go look at the "fine tune your Y axis" thread and pick out one of the several Y axis rebuilds. I am using the 16T version with the beefed up Y stepper mount.

  • Adjust Y belt tension. If this is vibration related, adjusting the tension may adjust the resonance one way or the other.
  • Can you try a calibration cube at default speed and slower speed (say, 15mm/s external perimeters) and post results? 🙂

    My MK3 Parts: [Bowden] [New Shoes] [TPU Micro Springs]

    Posted : 20/02/2018 3:54 pm
    thomas.h19
    (@thomas-h19)
    Active Member
    Re: poor print quality

    Thank you so much Brigandier, I will look into those. I have tightened up the Y idler a bit. I think that as it was letting me tighten it up quite a bit, feeling like the same tension the whole way, that the nut wasn't pulled all the way in.

    Unfortunately this is work and the boss doesn't really understand the machine and is very...not cool.
    He doesn't see the imperfections that I do lol, so I don't have much of a chance to fiddle.
    If I get a chance I'll do that cube.

    Thanks again.

    Posted : 21/02/2018 5:35 am
    thomas.a14
    (@thomas-a14)
    New Member
    Re: poor print quality

    Hello,
    I've read all through this thread trying to find the answer but for me at least it is not so clear:
    Have I to buy the MK3 or the MK2 keeping in mind that:

    [*]I don't care about print speed
    [*]I don't care about the noise
    [*]I don't care about the feature that resume the print after a blackout

    [*]I care about max print quality out of the box and print quality in general, I will not give up on this
    [*]I care about the removable plate but if I have to give up on some feature is this
    [*]I care about affidability

    Hope that someone that own both MK2 and MK3 can give me some advice.
    I know that is not a simple question to answer but I hope that the priority list helps.

    Thank you

    Posted : 08/03/2018 6:24 pm
    Fred G
    (@fred-g)
    New Member
    Re: poor print quality

    hi,
    Ive got the same problem with mk2s. Its not an firmware problem its, mechanical.
    My printer was fine no quality issue. But the igus was very noisy. So i checked this and i broke my y motor.
    i dessassembly and test with a new nema. everything is perpendicular y axes moves smoothly. I put the belt on and they are vibration (belt and pulley misalignment). The movement is not like before (before with the belt on it was difficult to move by hand but no vibration).
    My anet with 33030 profile have the same issue.
    I dont know how to fix it and it blows my mind. But i thing i need to print y-motor and y-idle parts to fix this.

    Posted : 10/03/2018 2:08 pm
    Page 11 / 13
    Share: