Single Drastic Layer Shift

[Herrminator117]

So I am currently working on a quad copter frame, because its finals week and modeling/printing is my stress reliever, and my printer has been making it not as fun recently!

Basically the print will be just fine for the first 15-30 layers then all of a sudden the effector will start printing anywhere from .5mm to 5mm to the right (negative X direction) of where it should be. Ive had my printer for over a year now and have seen almost 5km of filament through it. The only time ive seen this issue is when I still had the original arms and the hotend would hit the print and a belt would slip. Now the interesting thing is I have since upgraded to magnetic CF arms, so theres no way the hotend could hit the part with enough force to cause a belt to slip, because if it did the effector would just pop off, since its magnets (trust me its happened). This issue happens mostly with ABS. If I use PLA its like the filament jams for a single layer and then unjams for the next which causes the print to also mess up, but stays in the correct place.

I am really at a loss and would love absolutely any suggestions, I have 6 or 7 error prints and the layer it messes up at isnt consistent, or even how badly it messes up. It might even be specific to this part because I was able to print something for a client that was only 15 layers high with absolutely zero issues.

Possible Issues I think it could be:
-Damaged belt (I dont have the time or patience to rip them all apart at the moment)
-Cheapskates too tight?
-Loose gear on stepper?

Ive included one of the parts from the quad frame that are giving me issues. If someone has some extra filament and time and want to try, Id be in your debt. (I'll send you the rest of the frame when Im done finalizing it!) Also there is a picture of a less drastic shift, and one of the worse ones, I can just peel off the shifted layer on the bad ones

[Nylocke]

I'd guess that either you're getting cross talk (interference from motors or heaters) on an endstop or two at some point and it is causing it to trigger and skew, or your motors are skipping a few steps, getting caught on something. Make sure the bowden tube and wiring aren't getting caught on anything.

To fix the endstop problem, grab the firmware you last uploaded to your machine, go to the configuration.h file, do ctl-f (or cmd-f if on OS X), and search for "always". It should come up with ALWAYS_CHECK_ENDSTOPS, and if it is 0 set it to 1 and reupload your firmware. If it is 0 already then you have another problem. Be careful with this set like this though, you can easily crash your machine since it will only check if the endstops are triggered on a home all move.

[Herrminator117]

Would the endstop problem cause an issue midprint? I didnt think the endstops had any effect after the initial homing of the cheapskates and then it begins printing. Thats why if the hotend gets caught while printing itll shift, because the printer still believes its in the correct place from the initial homing, correct?

Also, another question, I havent had any reason to reupload the firmware to my printer since I built it last March, and I already feel dumb, because it seems pretty obvious I should be keeping the firmware up to date, yes?

[Nylocke]

You don't have to update the firmware, the older versions work just fine. I just like some of the newer features.

That is true, the steppers assume they haven't skipped steps. I said to disable always check endstops because sometimes EMF from the motors/heaters can cause false a false trigger of the endstop during a print. This will cause the motor to stop moving in the upwards direction, which will skew the print. This can also happen with failing endstop switches, but I don't think you have that problem. Steve of SeeMe had this problem on his printer at a makerfaire a few years back.

[Herrminator117]

Okay! Now that makes sense. And that also may explain why it happens more with ABS. More current through the heating wires could cause differences in the electric fields and magnetic fields around them.

Ill give that a shot! Im going to have to go back through the manual to remember how to even update the firmware, haha. Its been a minute!

[Nylocke]

It could just be it snagging on something though. Have you tried running the print again and staying around for it to see what happens? Or maybe recording it with a phone or webcam if you're busy.

[Herrminator117]

Ive tried, I will set up a camera on the next iteration to see. Im always watching when everything is fine, and then I look away and everything goes wrong! But I'll update this once I have tried both!
Which will be after my final today.

[Nylocke]

Good luck on those, I've gone mine next week.

[geneb]

If you routed the end stop wires by themselves in a tower, the chances of EMF being your problem are pretty small. I'd check to make sure you don't have a lose gear on a stepper.

g.

[Nylocke]

The possibilities of EMF are slim, but if the pairs aren't twisted and the wiring is a bit messy then it is certainly possible. Its almost impossible to keep the wiring in the base clean unless you wrap it all up with something (and not a lot of people think to do this during assembly). Thats the only reason my printer looks nice in the base.

But yeah you should also check the pulleys.

[Herrminator117]

Finally got it fixed! Im still not quite sure what caused the layer shift, but this last time I printed there was a small layer shift, and then 3 layers later the head rammed into the part.

I think the issue was that the slicer was having trouble with the amount of moves that were required for the piece and the quick changes in direction were causing the belts to slip (or something to slip atleast)
Because after I watched the video there were about 3 seconds worth of jumpy back and forth moves without any printing before it went straight down into the print.
I switched to the CuraSlicer and I printed both halves without any issue!

I have one question, has anyone gotten Slic3r to work on Mattercontrol? Cura is fine and dandy but it doesnt have nearly the same customize options MatterSlicer or Slic3r has.
When I attempted to use Slic3r I couldnt get the print to center. The part was fine in the mock up on MatterControl but when I would generate the slice it would be half off the build plate. I think it might be putting (0,0) <- The center at the bottom left most corner of the part, which would cause it to be shifted, but thats just a guess.

[barry99705]

Slic3r in Mattercontrol is broken, and has been for a few versions now. I don't think they want to fix it.

[Herrminator117]

Slic3r in Mattercontrol is broken, and has been for a few versions now. I don't think they want to fix it.

Thats honestly really unfortunate. Although this is only the first time in an entire year Ive had issues with MatterSlice, its nice knowing there are options

[PartDaddy]

:/ I had to run my big PartDaddy printer with Repetier to have the option of slicing with Slic3r. I hope they fix it soon!

We almost exclusively use Matter Control. I only imagine the massive programming task to keep things perfect while adding features and things. The Matter Control folks are active and always working toward the overall user experience. I'd think they'd get this fixed soon. Having slicing options within Matter Control sure is a good plan on their part.

Always double check your STL model too. Modeling errors can cause things to go wrong too.

[IMBoring25]

Won't MC also take raw gcode?

[PartDaddy]

Yes, you can open a g-code file. But you can't edit it within MC.