I am trying to understand what could cause a misprint after several layers. It acts as if the extruder head bump to the PLA after a while. Thus moving the parts and then, it fails because it is not printing at the right location anymore. layer offsetenter image description here

The expected result, for one "pyramid" is the following: enter image description here The whole model is the cube gear: Model of the geared cube

I checked the extruder multiplier coefficient parameter. The width of the filament is ok: 1.75 mm. The Y axis seems to be properly calibrated, and if it was the problem, I would expect it to fail before anyway. But I don't have any idea of what could cause that aside from that. And the problem seems to general and "quite" repeatable based on this print and some others (I do not guarantee the same height of failure, but...).

I got a better result by printing layers by 0.3 mm with a 0.4 mm nozzle. But this failed print was with a 0.2 mm layer.

My configuration: I have a prusa i3, with a bowden extruder. It has a slight multiplication of the extruder size, because during my calibration, I noticed that not enough material was pushed. I also calibrated my Z axis as carefully as I could. So what could cause the problem?

During my prints, I also have a problem of small strings everywhere, that I didn't manage to solve completely with the hotend temperature, nor the withdrawal parameter set to 2 mm. And I don't have a fan blowing at the end object (I plan to add one eventually). And the bridge speed is supposed to go at 60 mm/s and non print move at 100 mm/s, thus fast enough I think to avoid strings.

  • $\begingroup$ The image you have added is not very clear, could you add an image from another perspective and show what the print needs to look like (i.e. the model)? What you describe is a problem called "layer shifting", but the image does not show that effect clearly. $\endgroup$
    – 0scar
    Commented Jan 23, 2019 at 6:29
  • $\begingroup$ Did you check whether the print shifted on the bed? Adding a brim or a raft (and adjusting the Z-axis zero) can help if that's the case. $\endgroup$ Commented Jan 23, 2019 at 16:30
  • $\begingroup$ @0scar : I added the asked information. I also tried another print with a 0.3 mm layer height instead of 0.2 mm. And it worked. Thus the layer height plays a big role. I added the information in the paragraph. $\endgroup$
    – Emile D.
    Commented Jan 24, 2019 at 3:06
  • $\begingroup$ @CarlWitthoft Yes, the print did move. It seems that over layer stacking, the nozzle started hitting the print. Thus moving everything. Whenever I have a difficult print such as this one, I tried printing a bigger "fundation" layer. And indeed, it helps for stability. But it doesn't solve the core problem. I try to understand how to solve that part. But as a workaround, it is something. $\endgroup$
    – Emile D.
    Commented Jan 24, 2019 at 3:12

1 Answer 1


Your printer basically has 2 problems;

  • stringing, and
  • layer shifting

The second problem is most probably the result of the first issue. Excessive stringing leads to a lot of material outside the print object, when this clutters, the nozzle could get caught up and cause hindering of movement of a certain axis to result in a shifted layer print as the printer is not aware of this shift, it will continue the print from the shifted reference point.

From your second experiment you see that a larger layer height does not show excessive stringing and the print has therefore not been hindered by caught-up material to finish the product without layer shifting.

It is key to solve the stringing issue; stringing is the result of filament being extruded by pressure build-up in the nozzle that "leaks" out of the nozzle on travel movement of the print head. You will probably not see this happening when you print a single product; it is typically seen when printing multiple objects on the build plate. This usually implies that the printer need to be properly tuned. It is up to you to find the correct settings; test print objects, like calibration prints, help you find these correct settings.

Important print parameters to fight stringing are:

  • retraction, (do you retract enough prior to movement?)
  • coasting, (do you stop in time with extruding to make use of the overpressure in the nozzle?)
  • travel speed, (do you move fast enough to minimize the amount of filament leaking out?)
  • print temperature, (do you print at a temperature where the filament doesn't get too fluid?)
  • print part cooling, (do you have sufficient cooling?)

Once you optimized this, the stringing should be gone and the change of layer shifting should have reduced.

  • $\begingroup$ I will complete with details later today. But about the coasting: I didn't see any parameters for it in slic3r. Is it? For the print temperature, I am quite sure that is the right one. I already solicited your help about it in the past. But for the print part cooling, I certainly lack of it. I miss a fan that I plan to add. $\endgroup$
    – Emile D.
    Commented Jan 24, 2019 at 16:46
  • $\begingroup$ However, I had some strings too one the "succeeded" part. The only thing: I removed them the day before taking the picture... But I indeed had less of them. $\endgroup$
    – Emile D.
    Commented Jan 24, 2019 at 16:51
  • $\begingroup$ Alright, I managed to fry the power supply. I just replaced it. The travel speed for non print movement is 100 mm/s. The print speed for perimeters is 40 mm/s. I increased the retraction from 2 to 5 mm and its speed to 80 mm/s. It did improve the result. But I tested it back with 0.3 mm layer. I will still try to lower the layer height, which was the core of the question, to see if I have some improvement. $\endgroup$
    – Emile D.
    Commented Jan 29, 2019 at 2:10
  • $\begingroup$ Ahh, and so far, I tried to add a cooling fan, but I doesn't run. I will investigate this problem as I believe it can really improve the quality of my prints. $\endgroup$
    – Emile D.
    Commented Jan 29, 2019 at 2:13

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .