X-axis layer shifting  

RE: X-axis layer shifting

I don’t have an answer though I’ll offer some brainstorming for free🙂

First off, is crash detection not working as in “reports a crash when none is apparent” or as in “does not report a crash when one is seen”? I ask to determine if we should follow this clue further.

Second, please confirm that, with power off, that when manually moving the X-axis slowly and fully in each direction,  you feel NO snagging, just a uniform pressure along the entire axis.

Third, please confirm that, while moving the axis as above, the belt is not riding up the wall of the idler pulley (right side of printer).

Also, please post a picture of the problem part, as many folks often see what is not obvious to the rest of us.

RE: X-axis layer shifting

Crash detection does not report any crashes when the layer shifting occurs.  Crash detection is on, and silent mode is off.  I moved the X-axis over its entire range of motion when powered off, and the bearings felt very smooth with no snagging.  The belt remains on both pulleys, and does not ride up the walls.

Here is one example of a failed print due to layer shifting.  It occurred about 1.6mm up from the bottom, and is shifted in the X direction by about 0.5mm.

Here is another very interesting example.  An X-axis layer shift occurred as usual, but this time I noticed before the part was finished.  Knowing the print was already ruined, I deliberately stalled the print head to see what would happen.  This time the printer detected the stall, re-homed, and finished the print properly.

Any thoughts as to what might be happening, and how to fix it?

Thanks!

RE: X-axis layer shifting

Now it's happening in the Y-axis as well.  This is becoming very frustrating...

RE: X-axis layer shifting

Rachel,

Since you are having this issue on both the x and y-axis have you checked that the pulleys on the stepper motors are tight? If you find one (or both) loose be sure to follow the correct tightening sequence for the screws.  First tighten one screw fully, ensuring that it is contacting the flat on the stepper motor shaft.  Second tighten the remaining screw.  Failure to follow this sequence will result in the pulleys loosening on the shaft again.

Regards,

Steve

RE: X-axis layer shifting

Yep, the pulley set screws are firmly tightened on the motor shaft flats.

RE: X-axis layer shifting

If you really slow down the printing speed, does it still happens?

RE: X-axis layer shifting

Just a few more 🙂 thoughts to help us decide between a SW and a HW issue.

You originally mentioned that it did not happen 100% of the time.  This would seem to rule out some kind of slicer or FW bug (Gcode should not change after slicing onto SD card), none the less, what version of slicer are you using? What version of FW is your printer running?

Are you printing from SD card (vs Octoprint, etc)?

Have you seen this when printing one of the Prusa Gcode suppled models?

Does this happen if you slice your model again?

Are you slicing with plain vanilla slicer settings?

When did this first start happening?

 It is curious that this is not always seen, and that it occurs at about the same height. Its not entirely clear if this is due to some Gcode instruction not getting to the Einsy controller properly, or if the hardware is occasionally incapable of following instructions. Have you seen this effect at higher layers? When this does not happen (the other 75% of the time), does the object print as expected?

RE: X-axis layer shifting

I've been using the Prusa slicer, both version 2.2.0 and one previous. Very plain settings; nothing fancy, I just want my part.  The printer is running the latest firmware 3.8.1-2869.  I print from both octopi and the SD card, and both have the same problem.

The layer shifting does not always occur at the same height.  The exact same G-code printed multiple times might fail near the bottom, near the top, or not at all.

When the layers don't shift, the printed object seems fine.  I have not tried Prusa G-code yet, but given how this problem is not fully repeatable, G-code errors seem unlikely.  This problem started immediately after purchase.

I have not tried changing the printing speed.

RE: X-axis layer shifting

@rachel

if everything is ok on the hardware side, i recommend to reduce the printing speed. Try with 40mm/s. I also would print one of Prusas templates with unchanged settings. If this does not work you should contact the Prusachat.

RE: X-axis layer shifting

I talked to tech support again, and another hypothesis for the problem was presented: the filament was catching on the notched edges of the spool.  The suggested solution was to reposition the spool to be centered over the printer.  I also taped over the spool notches.  So far it seems to be working  fine, but I will continue to watch it.

If this really was the problem, it raises a few questions:

1. Why do the assembly instructions say to position the spool off-centered?
2. Why do the spools have notches in them which so easily snag filament?
3. Why didn't the stall detection catch this problem?

I do hope these issues are addressed, and will be fixed for future users.

RE: X-axis layer shifting

have you considered printing a filament guide? 

the filament guide can stop the filament catching on the sides of the reel, it can also help prevent de spooling, on new rolls of filament

regards Joan

RE: X-axis layer shifting

@rachel

Glad to hear you’ve identified the issue.  As Joan suggested, you may want to try printing a filament guide. There are many out there and I’ve been using the angled racetrack version of this one successfully. Happy printing.

RE: X-axis layer shifting

Ive had this exact same issue, can you confirm it has fixed it? ive rebuilt my poor x-axis 3 times trying to nail down this issue of .5mm shifts in prints and been tearing my hair out (i didnt have a ton left to start with) trying to figure this out.  

if this solves my issue i just want you to know you are my obi-wan kenobi.

RE: X-axis layer shifting

I mounted the spool significantly above the printer, and always tape over the spool notches, which seems to help.  The layer shifting is greatly reduced, if not outright eliminated.  I do wish Prusa would fix this design flaw though.

RE: X-axis layer shifting

is up to you entirely how and where you mount your spool, how you feed it to the print head, how you control backlash etc.  It's not a toaster.

RE: X-axis layer shifting

Hmmm well I placed my spool as shown so it's off center however the prusamint spool itself doesn't seem to have any ridges the filiment could be catching on.  

I'm going to center it and try again anyway honestly I'm at a total loss as to what to do otherwise. 

RE: X-axis layer shifting

@dan-rogers could you please elaborate about this not being a toaster?

RE: X-axis layer shifting

Why go against the flow?

RE: X-axis layer shifting

As for those unexplained - non-crash shifts - a motor drive gear not properly affixed to the motor shaft can explain it. If set screws are not tightened in order (flat first) they will loosen and slip. They'll even feel tight, but still be prone to slip. Start with both set screws loose so the gear can spin. Then set the screw on the flat, ensure the face of the screw is flat against the shaft flat. Rocking the gear as you tighten the screw will help to ensure this. Once the screw has fully engaged the shaft: tighten to spec.  Then, and only then, tighten the set screw on the round of the shaft. Then you are done: do not go back and retighten the set screw on the flat: you'll strip the head. 

The remove the set screws without stripping, loosen the set screw on the round first; that will free up the set screw on the flat.