Ultimaker S5 - Brutal shift on some prints

[Vprods 0]

Hello everyone,

 

I've been using the UMS5 for about 4 months now and in random cases it does a brutal shift on the x axis and a bit on the y axis before continuing normally without ever going back in position.

It seems to be consistent if I use the same printing file and if I make and use a new ufp file the problem seems to go away but I have not made extensive testing on the issue.

Any help on the matter is appreciated.

 

[Slashee_the_Cow 438]

When you make a new file (which then apparently prints fine) are you using the exact same settings as the original file you made? And if you could post an original (which fails) and a new one (which works) so we can see if there's any important differences, as well as a Cura project file for what you're trying to print (.3mf, get it set up then go to File > Save Project) that would really help.

 

Also something that came to mind: does it do it at the exact same layer if you try to print a "bad" one multiple times? And what sort of filament are you using, just PLA?

 

The only time I've experienced layer shift that bad myself was with a tangled spool of filament, the extruder had to pull so hard to get past a tangle it actually moved the print head with it (direct drive).

(Since I can't really afford to not use filament which is otherwise good, for about the second half of that spool I had it sit next to me fed into my printer's tent through some spare Bowden tubing and I'd manually pull some off the spool every so often while it was printing).

 

The main cause of layer shift is sudden changes in speed - usually the jerk. I have no idea about UltiMaker printers (unfortunately a printer isn't part of the signup bonus for the Assistant Moderator position) but other printers I am familiar with (like my Ender-3 V3 SE... or most of the Ender family... or most of the Chinese ones in roughly the same price range, they're all pretty similar), if the belts aren't at the correct tension, then a sudden change in speed/direction can cause the belt to slip on the motor, causing it to miss a few steps and no longer be in line with the layers below. Do it hard enough (I can do it deliberately if I want to, though it's not exactly a party trick) and even a belt at the correct tension can slip, or the motor can be shaken out of position so loses steps itself.

 

There's also really "out there" cases, like if your printer is a bed slinger (moves the bed instead of the print head for the X and/or Y axis) and you get poor adhesion at the start a jolt can move the bottom of the object, on the plate.

 

That long explanation aside, my untrained eye says that's probably not it. It almost looks like there's basically no adhesion between the two sections (something it's easy enough to do deliberately, but hard to do accidentally). Are they stuck firmly together (just not right on top of each other) or is there a bit of movement?

 

About the only setting I can think of testing would be if you had Special Modes > Print Sequence set to One at a Time that could maybe cause it to get knocked but if it was going to be a problem it probably wouldn't slice successfully.

(Worth trying changing it to All at Once if it's not already on it).

 

Also I'm going to throw @gr5 under the bus here because he knows about UltiMaker printers 

[Vprods 0]

When I make a new file I do indeed keep the same settings, usually I just move the 3d model or models around on the plate to distance them a bit more and that does the trick.
I'll if I can find back the original file, if not I'll keep my eyes peeled for the next failing file. I'll also include the mf3 project of course.

The issue has shown itself with PLA as well as Though PLA and across files it showed different heights as point of failures as shown in the following picture.

With the black figurine the cut/shift has been at a consisten heigh across 3 different prints while using the same file (in the last print I changed the material to another PLA I had but that didn't change the outcome).

The belt tension seems to be good but I'll check it again tomorrow as it's possible I missed it if it's not an obvious change since I've received the printer, the belts themselves are in good condition with no missing parts.

As for the "bed slinger" it is a possibility as the printer is set on a table with wheels.

 

That seems improbable though as it is a rare outcome for the print to shift like shown in the initial post. Also the base of the print always remained solidly attached to the plate. 
The two parts of the cube are firmly stuck together and it doesn't move at all. I have to use lot of force to separate the parts and stick them in the correct position with glue.

I will try out the special setting to see if it makes a difference of failing printing files.

[gr5 2,235]

The steppers don't know where they are.  There is no feedback.  The printer tells the stepper to move one step at a time and assumes everything worked but sometimes the resistance to moving the head is too high for the head to move and you lose a step.  Or more.  You lost hundreds of steps in one moment.

 

The most common causes:

motion control settings (mostly acceleration) set to high - very unlikely

increased resistance - very common as a printer ages it needs lubrication and alignment

weak stepper driver - (pretty much only happens if your printer is too hot)

 

What is the room temp that the S5 is in?  If it's over 30C then that is likely the problem.

 

Push the head around (with power on or off - it doesn't matter) and feel the resistance.  The resistance is likely higher on the failing axis.  Also the resistance might have some bad spots where your part was located.  You can feel the steps by the way in the stepper - that's normal - but no other repeating resistance.

 

Look down on the printer from above - are the 2 rods through the head perpendicular?  It's common for one of the 2 long belts on an axis to skip a tooth.  This makes it visibly crooked looking from above.  This is easily fixed.  It also causes excessive binding (friction).

 

Feel free to add a drop of oil to all 6 rods (not 4, not 2) for both axes.  No more than one drop each.  The center rods need the least oil as that has ball bearings internally and oil can bring in dirt into the bearings.

 

A pulley may be loose.  Almost always the pulley on the stepper motor.  This was common on the UMO but rare on the S5.  Still, maybe use a sharpie to mark the pulley on the stepper for the worse axis - mark the shaft and the pulley.  It's hard to get a sharpie down in there though but worth it if you want a definite diagnosis.  Might as well mark all the other pulleys as well but 90% of the time it's the one on the stepper.  The set screws on those pulleys need to be tight as hell - so much that your hex driver feels like it's going to break.