Prints Shifting layers - Could this be an OctoPrint on a Pi Zero 2W issue?

[ImaginaryTango]

What is the problem?

Prints shift at some layers

What did you already try to solve it?

Resliced prints, chagned other settings.

I'm printing something at a higher density because I need a stronger part. Here's what the test print of this item, done at 15% infill, looks like:

For scale, the four slots hold 1" aluminum tubes.

I've tried printing this part at a higher density, at 50% and 40%. When I do, layers shift. Usually just once, like this:

But here's a test print that shifted twice, once far to one side, the other time it shifted back to almost on top of the original location.

This print takes a long time, even just to get to where it usually fails the first time, so I'm working on toubleshooting it. I will be trying safe mode from OctoPrint after the current test print is running.

I've noticed that the shifting only happens while it's printing this big part of the print at a high density. I'm also running OctoPrint on a Pi Zero 2W. Is it possible the failure is related to running OctoPrint on something less powerful, like the Pi Zero? With it at a higher density, that means a lot more gcode and more commands. I don't see indications any commands are being dropped.

As part of the process, I'm wondering how likely the lower power Pi Zero could be part of the problem here. Other than trying "safe mode," is CPU power and speed a possible issue here?

[Ewald_Ikemann]

Layer shifts are usually a issue with the printer.

---

I'm a bit confused:
Are you using a Pi Zero (as in the topic noted) or a Pi Zero 2 W (as mentioned in the text)? These are two different SBC boards. You may correct your topics (also the other thread)

[ImaginaryTango]

@Ewald_Ikemann

Are you using a Pi Zero (as in the topic noted) or a Pi Zero 2 W (as mentioned in the text)? These are two different SBC boards. You may correct your topics (also the other thread)

Sorry. I got sloppy there. Pi Zero 2W.

@Ewald_Ikemann

Layer shifts are usually a issue with the printer.

I know this forum is about OctoPrint and not about printers, but can you add any more than that? I'm using a Prusa Mk3S+ upgraded to a 3.5. I've been wondering if this could be an issue with buffering, since the denser prints require a lot more gcode (especially if I use gyroid infill) and I'm wondering if there could be an issue with commands being dropped from the buffer. I'm perfectly willing to accept it is likely with the printer, so I'm not trying to argue. I'm just wondering if you might have more information so I get a better idea of what to look into.

[CmdrCody]

Buffer sizes can cause stuttering. Layer shifting is usually over extrusion. At 15% infill there is plenty of room for the over extrusion to be pushed into. Also, there is more time for the plastic set and cool so it won't push over. Since it only happens once or twice, I would try a higher temp for a part to see if it helps, like 5-10 degrees.

[ImaginaryTango]

@CmdrCody

Layer shifting is usually over extrusion.

I wasn't going to get too far into this if it's a printer issue - there are other forums for that, but at this point, I'm not yet sure where to look for the main cause of this issue. I had several prints shift about the same amount at about the same level, like this:

That was at 40% infill, using a gyroid pattern. I had 2 prints fail in the same way with PETG and a couple tries later, another fail almost identically in PLA.

Then I had this one today:

If you look at the layers, the printing is rough. I've tried this with dried filaments. (I'm in a high humidity area, so I use a filament drier, sometimes running filament through multiple 4 hour drying cycles before using it.) This has 2 shifts, the first one goes far to the left, but doesn't seem to shift forward or backward. The other shift was far back to to the right, almost back to the original position, but a little over it.

I definitely am having an extrusion issue (and, as part of that issue, have a new extruder due tomorrow), but I don't think the distance things shifted in the 2nd photo, and the shift in the first, where the entire print and all subsequent layers shifted, would be due to extrusion. (When you say "extrusion," I'm thinking you mean that a glob came out that was big enough that, on the next layer, the blob would have blocked normal printhead motion. If so, it's worth noting that the 1st photo was a shift backwards, on the Y axis, where the print bed, not the print head, is moving.)

I'm thinking these kinds of issues don't seem like they'd be from extrusion. Honestly, I hope they are and that, since I had dried the filament to avoid humidity issues, that my issue is likely related to the print head.

Any thoughts on this from the photos?

[CmdrCody]

The shift will occur on a rapid non-extrusion move to the next layer/type of perimeter/type of infill etc. You can reduce it by always doing a z-lift but then you run into retract/restore nozzle plugging, which can fixed with hotter temps.

[jneilliii]

you could put the file on a sd card and print directly from printer. that could potentially eliminate OctoPrint as a root cause.

[ImaginaryTango]

That's almost my next step. First, I need to do some cleaning and deal with the extrusion issues. Once I do that, I'm going to use SneakerNet and the SD card to do a test print.

[ImaginaryTango]

Update on this:

I asked about this here and on the Prusa forum. It was not a simple issue to troubleshoot! While it came down to the printer, there were things I tested (and was asked or suggested to test). It turned out to be one source issue, but that caused a 2nd issue, so when the 1st was cleared up, things still weren't working. In the end, even though it started with 1 issue, there were 2 secondary issues that were basically created by the original.

So nothing to do with OctoPrint and everything to do with some print issues.