Printbed or powder-coated sheet weirdly uneven  

  RSS
cordilon
(@cordilon)
Eminent Member

Hey guys,

since I got my Prusa i3 Mk3 late Nomvember, I had a couple of more or less annoying issues, but what probably made me scratch my head the most was the bottoms of my prints always being weird. Meaning: In some areas the lines were fused to a completely consistent surface and in other areas they weren't touching each other at all.
Of course I thought maybe I have to do some manual bed levelling, but today I finally had the time to print a large test pattern that covers most of the printbed, which tells a more complicated story.
And because the powder-coated steel sheet was my first suspect, I tried it on both sides. Here are the results:

(In the first image, the print was already coming off the cooling steel sheet. That's not part of my issue!)

If you zoom in, you can see a gradual change from a fused filament surface to gaps between the lines in several places. In the bottom right corner the nozzle almost scraped the steel sheet while left from the center most of the lines are seperate. Also both prints seem to show a very similar "layout" (so it's not the steel sheet?!).
As a lot of the problematic areas are near the center, manual bed levelling wouldn't help me at all and by now I'm practically out of ideas.
Here's all the things I tried/checked:

- did mesh bed levelling (obviously)
- removed steel sheet and cleaned it on both sides (with IPA)
- checked heatbed for debris and cleaned it just for good measure (with IPA)
- checked extruder for consistent extrusion + idler tension screws + idler door axle
- checked U-bolts of heatbed. They are snug, heatbed doesn't wobble
- checked all smooth rods for tight fit and cleaned them. All axes move smoothly
- went through Z-calibration multiple times
- went through XYZ-calibration at least twice (assemled the printer myself). Says all is perpendicular
- doublechecked and tightened all the screws, especially on the extruder and heatbed

The test pattern uses octagram spiral as infill to ensure equal line lenghts throughout the object and was printed with the same settings on both sides of the sheet:
Speed: 25mm/s @ 215°C nozzle and 60°C heatbed with Prusa PLA white. The file was sliced in slic3r PE 1.41.1 and printed on my Mk3 (self assembled kit) with firmware 3.5.0.
I'll attach the g-code and STL-file just in case I messed them up somehow or you want to try it yourself.

Could it still have something to do with the steel sheet or is it more likely to be the bed? Or something else entirely?
I'd really appreciate some help or suggestions!

Posted : 21/01/2019 7:51 pm
charles.h13
(@charles-h13)
Famed Member

It could be either the bed or the sheet. I had a similar issue to a black PEI sheet. I thought it was my bed or the sheet. I confirm it was the sheet thickness and an air bubble.

Posted : 21/01/2019 8:11 pm
Olef
 olef
(@olef)
Honorable Member

I had similar problems but was able to even things out by using 3x6x0.1mm steel shim washers on the bed mounting spacers. I purchased a decent 0.001mm DTI and printed this mount from Thingiverse - https://www.thingiverse.com/thing:3056465

It is a laborious process and very fiddly to get the bed spacers and shim washers in place but I found it worthwhile. I was able to get to a place where I am satisfied with my first layers wherever I am printing on the bed.

Awesome test pattern by the way, thanks for that. I may now print it and drive myself nuts again with bed levels! 🙂

Posted : 21/01/2019 11:14 pm
cordilon
(@cordilon)
Eminent Member


I confirm it was the sheet thickness and an air bubble.

Interesting! How did you find out, it was the sheet? and what exactly do you mean by air bubble?


I had similar problems but was able to even things out by using 3x6x0.1mm steel shim washers on the bed mounting spacers.

I thought about trying that, but as you can see in the photos, the lines are often fine above the mounting spacers but too far apart between them! So if I'm not wrong here, additional washers would only shift the problem and not solve it, correct?


Awesome test pattern by the way, thanks for that. I may now print it and drive myself nuts again with bed levels!

Glad you like it! I'm thinking about uploading it to thingiverse too, so that hopefully more people can smash their heads against the heatbed 😆

Posted : 26/01/2019 12:53 pm
charles.h13
(@charles-h13)
Famed Member

I was just saying that an air bubble is possible with the PEI sheets but not likely with the textured. I had a PEI sheet that I placed on the bed. I accidentally put an air bubble and had to remove it by nicking the sheet with a knife. I had similar results because the bubble created an uneven spot.

Posted : 26/01/2019 3:21 pm
chris.n6
(@chris-n6)
Reputable Member

To me the strangest part of this is the bed levelling is supposed to correct for this sort of thing. have you taken a steel rule to see if there is a low spot in the steel sheet or the carriage?

Posted : 07/03/2019 8:30 am
charles.h13
(@charles-h13)
Famed Member

I have has this same problem. I am still abel to print so I have not worried about minor imperfections in my prints and I avoid the front right corned if I can.

Posted : 07/03/2019 12:56 pm
cordilon
(@cordilon)
Eminent Member


To me the strangest part of this is the bed levelling is supposed to correct for this sort of thing. have you taken a steel rule to see if there is a low spot in the steel sheet or the carriage?

I'm pretty sure the bumps are too small to be detected by the standard mesh bed levelling. Haven't tried 7x7 mesh bed levelling yet, because I feel a little uncomfortable with flashing uncertified firmware.

UPDATE:
The steel ruler test was exactly what was suggested by Prusa support about a week ago. After taking the whole Y-carriage apart to make sure that the carriage itself is flat (and it is), I used a steel ruler and flashlight behind it in several places of the heatbed:

The pictures are actually screenshots from a short video I made. The bumpy areas might be more visible there.
So it turns out, the whole heatbed is uneven! The support employee supposed, I might have damaged it during assembly. So they were kind enough to send me a new one and this time I did the steel ruler test before installing it. And it had the exact same problem.
I haven't made a new video or printed my test pattern yet, but will probably do so in the next few days.
The problem seems miniscule but it can influence the adhesion quite a bit and apart from that I expect a higher standard from a machine in this price range and from a company with such a good reputation.

Posted : 09/03/2019 6:54 pm
chris.n6
(@chris-n6)
Reputable Member


The support employee supposed, I might have damaged it during assembly. So they were kind enough to send me a new one and this time I did the steel ruler test before installing it. And it had the exact same problem.
I haven't made a new video or printed my test pattern yet, but will probably do so in the next few days.
The problem seems miniscule but it can influence the adhesion quite a bit and apart from that I expect a higher standard from a machine in this price range and from a company with such a good reputation.

do you mean you had the same issue before you even installed it OR do you mean you had the issue after attaching it with screws to the carriage?
if you mean the latter, then i'm betting it's not the heat bed that is bent. either the standoffs are irregular height or the carriage itself has a problem.
realize these things are being shipped and a lot can happen during shipping.

Posted : 18/03/2019 3:56 am
cordilon
(@cordilon)
Eminent Member


do you mean you had the same issue before you even installed it OR do you mean you had the issue after attaching it with screws to the carriage?
if you mean the latter, then i'm betting it's not the heat bed that is bent. either the standoffs are irregular height or the carriage itself has a problem.
realize these things are being shipped and a lot can happen during shipping.

Sorry I haven't updated this post yet, I'm quite busy.
To clarify:

  • With my first heatbed I did the steel ruler test after a couple of months of use. The photos and video I posted are from that exact test.

  • After that I took the whole Y-carriage apart and checked if the steel carriage itself is flat. It is indeed totally flat and undamaged. Also measured the spacers with digital calipers and they only deviate by 0.01mm tops.

  • Then I got the second heatbed with which I made the steel ruler test practically right out of the box (=before installing it) and as I said: It has the same issues as my first heatbed, but probably in slightly different spots (still have to confirm that). I don't think it has been damaged by shipping either because it was wrapped in bubble wrap twice plus packed in a cardboard box.
  • Posted : 18/03/2019 11:58 am
    Share:

    Please Login or Register