Ghosting or layer shift issue.

Hi everyone,
I'm having an issue I can't track down. Not sure is it's a rippling/ghosting issue or layer shift. Printing a 20x20x20mm calibration cube. Once it starts printing the X right after it I get this effect. Doesn't look like the typical rippling/ghosting as it makes those layers stick out further. I attached a video I slowed down to show movement. I checked my belts/ tightened, pulleys ok. I'm have a prusa clone with a proximity sensor running skynet firmware. Any suggestions would be great.

[youtu.be]

Thanks,
Mark

Also I bumped down my acceleration to 800 from 1000 and jerk to 10 from 20 with no improvement.

Those sorts of artifacts seem to be quite common, and I'm not sure how to get rid of them. I usually see them on a corner near a hole. I guess the obvious thing to try is slowing the print speed way down, but I have a feeling that speed isn't the root cause in this case.

Thanks for the reply James!

I'm printing at 30/mms. This is driving me crazy. Maybe I'll try dropping it down to 15/mms to see if I can spot it.

Yes, it's pretty frustrating. I've been putting off trying to fix it as I have a feeling it could be quite the rabbit hole. My best guess at the moment is that the change in tool path pushes excess plastic in different directions, so you get a bulge in a different place depending on where the extruder came from. If this is the case, then reducing the extrusion (flow rate multiplier) might have a good effect - the problem would be worse if there was any over extrusion, and should be less if it was slightly under extruded. That might be no help at all if fixing the problem means having to significantly under extrude, but it would be nice to know...

I think that our slicing gurus need to think about final layer paths. Noticed that Slic3r seams to jump around instead of the best path that goes from one corner to the opposite corner.

Mark,
I looked closely at the photos.

One way to test this is to rotate the part 45 degrees, print and observe. If it was only an extrusion issue, then the two prints would be identical.

I suggest that this is caused by backlash, because it appears there is a slight shift when changing direction. The direction it is shifting will give you a clue about which axis to work on. This appears to be identical to another reprapper's issue - and they solved it finally by repeatedly tightening the belt, but there are several things that could contribute.

I suggest looking at the following:
- Make sure the axis moves very freely with no hangups, or resistance from wires or filament in one direction.
- Make sure there is no play in the axis at all
- Make sure that everything is tight, including the extruder. Grab it and try to wiggle it. Does something get loose when hot?
- Are you using a belt known to have low backlash - like GT2?
- Are all the grub screws tight, and a flat ground on the stepper shaft? Do not miss this.
- Are the belts quite tight? One way would be to pluck them - do they make a medium tone?
- Do the belt mounts flex at all, or anything else flex for that matter?

Additional possibilities on the Shifted Layers wiki page.

Hope that helps.

Edited 2 time(s). Last edit at 12/31/2016 05:57PM by Paul Wanamaker.

---

My printer: Raptosaur - Large Format Delta - [www.paulwanamaker.wordpress.com]
Can you answer questions about Calibration, Printing issues, Mechanics? Write it up and improve the Wiki!

Thanks for the replies,
I did as Paul suggested and printed the cube 45 degrees. The issue still showed up. I'm seeing a difference when manually adjusting layer thickness instead of letting slic3r calculate it. Still fine tuning to see if I can get rid of it completely.

Quote
MarkM
I did as Paul suggested and printed the cube 45 degrees. The issue still showed up. I'm seeing a difference when manually adjusting layer thickness instead of letting slic3r calculate it.

Hmm, the two prints were identical? I actually didn't expect them to be exactly the same. Tricky. Perhaps the problem is with variations in extrusion rate.

Quote
MarkM
I'm seeing a difference when manually adjusting layer thickness instead of letting slic3r calculate it.

That's interesting. I think you are actually manually adjusting extrusion width?

If there is an issue with extrusion (due to varying pressure in the extruder) then here is a good test:
On the Advanced tab:
Set all of your extrusion widths to the same value: your nozzle diameter * 1.2 (this is a safe value).
It is best to start with all these values set to a specific value (not auto) so you can control what happens.

On the Speed tab:
Set all the speeds to the same value, say 25 mms.
It is best to always set speeds to a set value instead of auto so you can control what happens.

Then when you print the cube, all the features will print at the same extrusion rate, and that will at least rule out some possible issues, including a known bug in Slic3r.

Thanks Paul!
Issue looks like it's gone. After adjusting the extrusions width manually it was still there slightly. Once I brought down the overlap percentage (set it to 12% instead of default 15%) did the trick.

Thanks for your help.

Excellent. Glad to help.

---

My printer: Raptosaur - Large Format Delta - [www.paulwanamaker.wordpress.com]
Can you answer questions about Calibration, Printing issues, Mechanics? Write it up and improve the Wiki!