Thread Rating:
  • 0 Vote(s) - 0 Average
An artifact: a thin vertical slice with layers printed too late
#1
Photo 
Hi, my new Saturn printer produces sometimes strange artifacts. This image illustrates it:
[Image: G0TI1.png]

I contacted the support, and they answered that they suppose that it is a mechanical problem caused by the support. I do not think so, because the artifact looks "digital". I wonder if it is a bug in Chitubox or in the firmware. The bug is rare and happens only if there are several such cones accompanied by a complex model. It has shown only in about 10% of the cones. When I slightly modified the complex model in the latest print, all cones printed out correctly.

1. The density of the support structure is typical for default values in Chitubox 1.9.1. More than this and resin sticks between the support, makes it difficult for the air to flow, suction appears and makes holes in the model.

2. Once the bottom of the cone is printed, it also becomes a support of the higher layer next to be printed. So the discrepancy would self-correct itself and disappear in higher parts of the model. But it goes up the cone without any change.

3. The raised part has very sharp borders which form a perfect, precise vertical plane. It looks nothing like a mechanical problem with the support.

4. The raised part is very thin, about 0.5 width. A discrepancy between two parts of the model caused by a weak support would create a single, possibly wide deformation and not two close, razor-sharp discontinuities which precisely continue forever and have a perfect model surface between them.

5. The bottom of the cone does not show the arching characteristic for a weak support and for a reason: I made the bottom surface slightly diagonal to have the layer self-support effect.

6. The problem shows rarely, in about 10% of the cones and if they are accompanied by a large, complex model placed nearby.

7. If you look at the close-up, the print layers are perfectly horizontal to the built plate and thus do not show any mechanical problem with the support. There is a normal layering visible even on the vertical shear surface.

What do you think about these artifacts?
Reply
#2
I looked more closely at it. The number of layers in the "shear planes" is larger if closer to the cone's top, smaller if closer to the build plate. So it is not just a simple shift of a vertical slice. Possibly an effect of the new "blur" algorithms? An omitted one-pixel plane between tiles?
Reply
#3
Just an update on the problem.

The Z-shifted YZ vertical plane goes exactly through the central vertical Y line of the LCD (vertical if the LCD has a landscape orientation). In Chitubox, the line is exactly an extension of the triangular mark on the side of the horizontal plane.

I am almost sure that it is either a Chitubox or firmware bug, because the LCD does not have any kind of multi-frame buffer which could desynchronize the sliced contents like that.

I browsed via Google and it seems that I am not a single person whose printer has this artifact. See the last image here: https://www.reddit.com/r/ElegooSaturn/co...ing_error/. There is also a layer shift, very thin and inexplicable by support instability.

I do not know if the Z-misplaced Y line shows in all models or only in more complex ones. It seems rare but it happened already two or three times. In any case, I have a Chitubox project which shows this artifacts, if anyone is interested.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)