At my wit's end with Y-layer shifting  

Page 1 / 2
  RSS
taylorstrait
(@taylorstrait)
Active Member

I've been debugging a y-layer shift problem and I've tried EVERYTHING. Here are the details:

  • Single objects print fine
  • Prints with multiple, separate objects, or multiple instances of the same objects, have at least 1-3 y-layer shifts 100% of the time
  • Single object prints with large surface areas are fine, and any multiple object prints even with a low XY size have shifts
  • The first shift commonly occurs with the first 2.5cm, but not always
  • All other elements of the print are excellent, including first layer

Things I have tried:

  • Making sure grub screws were aligned and tight
  • Tightening the belt (it makes a bass sound)
  • Belt tension reports 273
  • There are no physical obstructions impacting the bed
  • The bearings under the heat bed are not too tight on the rails
  • The pulley idler is not too tight and spins freely (but the belt does drift right)
  • Lowering the y Jerk to 6mm/s does not resolve the issue
  • Lowering the max y acceleration to 800mm/s does not resolve the issue

If anyone can suggest something to help with this I'd be grateful. I just can't come up with any other possible solutions. Maybe a bad motor?

This topic was modified 1 year ago 2 times by taylorstrait
Posted : 21/07/2019 2:32 pm
Tim
(@tim-m30)
Illustrious Member

Sounds like a speed problem.  Lower the front panel speed to 25% and try printing multiple items.  If this works then you have too much friction on the axis that is shifting.

Powered off, the bed should move forwards and backwards under gravity.  Same with the extruder in left-right motion (takes a bit more angle to move the extruder).  

Moving the axis by hand, there should never be any "sticky" spots along the travel. 

 

ps; Y shifts are often caused by the U-bolts being too tight, or wires catching on the sled.

Generally - if you did not grease the bearings during assembly, they are too dry and prone to damage.  A few drops of oil on the rods can help, but is only a temporary solution.

This post was modified 1 year ago by Tim
It is always wise to get more than one opinion......
Posted : 21/07/2019 6:58 pm
taylorstrait
(@taylorstrait)
Active Member

Thanks for chiming in. When checking the movement of my axes under gravity, my y-axis only moves at ~60' tilt and the x-axis/extruder doesn't move until 75-80' tilt. There doesn't seem to be a difference if the power is turned on or off. Is this too stiff? Should I loosen the extruder and U-bolts?

Posted : 22/07/2019 2:14 am
david.a66
(@david-a66)
Honorable Member

yes that's too tight

Posted : 22/07/2019 2:37 am
taylorstrait
(@taylorstrait)
Active Member

I tried to lower the tension on both sets of bearings, but it did not appear to help. For the x-axis/extruder, I unscrewed the 4 extruder bolts until the whole extruder assembly could wiggle freely, but the x-axis still did not move freely. For the y-axis, I pulled the heat bed and loosened the u-bolts until the heat bed chassis could wiggle freely, but still the axis did not move freely. I went a step further and unscrewed the chassis, and the y-axis bearings appear to move smoothly and freely on the rod. 

For both axes, it "feels" like all the resistance is coming from the motor/belt. On the y-axis specifically, I can feel the resistance as steps on the motor. These test were all conducted with the power unplugged. Any thoughts on what this means? Again, the x-axis only moves by gravity when tilted about 75' or more, and the y-axis moves earlier but still requires a steep 60' for it to move on its own. 

Posted : 22/07/2019 4:21 am
david.a66
(@david-a66)
Honorable Member

belts too tight?

 

Posted : 22/07/2019 4:24 am
david.a66
(@david-a66)
Honorable Member

also try running with crash detect off

Posted : 22/07/2019 4:24 am
Tim
(@tim-m30)
Illustrious Member

Hard to say about belt tension - belts have to be measured using deflection. The LCD is not useful since the reported numbers are just the motor currents fighting friction. More friction, lower numbers.  My numbers are 295 and 296 and nowhere near the Prusa figures, but I know my belts have correct tension because I have measured them directly.

Shifts can also be belt hop or a drive gear spinning on the shaft.  Fast moves as the only clue is tough.  That said, Y-shifts should be easy to find.

First, lets see if it is a crash or a shift:  what do the crash statistics report?  

 

 

 

It is always wise to get more than one opinion......
Posted : 22/07/2019 7:38 am
robin.a
(@robin-a)
Estimable Member

A few questions

What slicer are you using?

What settings are you using for Print speeds, Movement speeds, Infill type?

And as said are you getting crash detections showing?

And a photo of failed print on the bed

 

Posted : 22/07/2019 8:49 am
taylorstrait
(@taylorstrait)
Active Member

I do show crashes: 8 power failures, 5 x-axis crashes and 9 y-axis crashes.

I'm using Prusaslicer, with standard speeds for PLA. For infill I usually use 15% gyroid.

I don't have any pics on the bed, but here is an example when I tried to print 4 instances of a lightcycle. If I print only a single instance of the same STL, it works flawlessly. (the elephant foot in that pic has since been resolved)

Posted : 22/07/2019 1:32 pm
robin.a
(@robin-a)
Estimable Member

Is the power fail on the print to total?  You should not get any power fails unless you mains power has gone off.

with a power fail the recovery is not always where the print finished and can cause layer shifts.

without seeing the actual crash events only assumptions can be made. 

So here goes. X axis crashes can be from and not limited to.

Loose drive pulley on the servo.

Tight spot on the axis bearing, usually axis rods are not parallel on one end.

bad none drive end idler pulley, bearing or poor fitment.

infill patter crash with nozzle 

Y axis crashes. Rod bearing U bolts too tight.

bearings that have been too tight and damaged the rods, scrapped the rod surface.

loose servo drive gear grub screws.

idler pulley bearings bad or poor alignment.

infill crash with nozzle 

 

so you will need to watch the print or video it and work out what it may be.

Posted : 22/07/2019 9:59 pm
Tim
(@tim-m30)
Illustrious Member

The way the part shifts a distance, then later shifts again and comes back into alignment, points to a drive gear not properly attached to the motor shaft.  This is a very common build problem.

 

Flat First: Drive gears must be tightened flat-first.

1)   Start with both set screws fully loosened so the pulley is free to spin on the motor shaft.

2)   Align one set screw dead center with the flat on the motor shaft, slowly tighten the screw until it fully contacts the flat surface.

3)   Torque the flat set screw to spec.

4)   Now tighten the jam set screw, and torque it to spec.

Once tightened, never touch the flat set screw unless the jam screw is first fully loosened.

Why Flat First? Set screws have flat ends. If you tighten the jam screw first the set screw on the shaft flat doesn't fully contact the shaft, only one small edge of the screw surface is biting. Reversing torques can easily shift the shaft to a position the flat screw no longer contacts the shaft. This lets the jam screw wriggle loose. And after a while, vibration loosens the screws until the gear is free to rotate.

It is always wise to get more than one opinion......
Posted : 22/07/2019 10:09 pm
taylorstrait
(@taylorstrait)
Active Member

X axis crashes can be from and not limited to.

Loose drive pulley on the servo.

Grub screws have been 3x checked and tightened. Flat screw first, then screw on round end.

Tight spot on the axis bearing, usually axis rods are not parallel on one end.

Rods have been measured with digital calipers and are basically parallel. However, I think I initially overtightened the extruder and there is a small part of the axis where the bearing "crackles." So I assume this LM8UU will eventually need to be replaced. But could that cause y-axis shifts?

bad none drive end idler pulley, bearing or poor fitment.

As far as I can tell, the idler moves freely.

Y axis crashes. Rod bearing U bolts too tight.

I've loosened the bolts and even removed them. As far as I can tell, the only source of resistance on the y-axis is the belt. The bearings can glide freely on the rods without issue when I remove the heat bed chassis.

bearings that have been too tight and damaged the rods, scrapped the rod surface.

Though I suspect an x-axis bearing has been damaged, there are no rough spots, scratches, or noises from the y-axis bearings/rods.

loose servo drive gear grub screws.

These have been 3x checked for tightness.

idler pulley bearings bad or poor alignment.

The idler pulley turns freely and is loose. However, the belt inevitably drift right on the idler. It initially caused a squeaking noise but that has stopped. However, any attempts to reset the belt have not worked and the belt always works its way to the right side of the idler. If this is a problem, how can I fix it?

 

Posted : 22/07/2019 10:15 pm
stephen.h14
(@stephen-h14)
Estimable Member
Posted by: Tim

 

 

ps; Y shifts are often caused by the U-bolts being too tight, or wires catching on the sled.

 

I would have voted for this. I've seen the bed's power cables get bent down enough that they catch on the frame.

Posted : 23/07/2019 12:23 am
Tim
(@tim-m30)
Illustrious Member

loose servo drive gear grub screws.

These have been 3x checked for tightness.

Except TIGHT isn't the problem.  It is how they are touching the motor shaft that is important.  Tight set screws spin on shafts, that's why there is a FLAT on the shaft.  You tighten one set screw to the flat. It must be aligned - then tightened.  You then tighten the jam set screw.  This action multiplies the torque on the flat set screw so it can't move and wiggle loose. 

Any other procedure to tighten the set screws will fail (and why so many apply lok-tite, even though that isn't as good as doing it right).

This post was modified 1 year ago by Tim
It is always wise to get more than one opinion......
Posted : 23/07/2019 1:26 am
taylorstrait
(@taylorstrait)
Active Member

I have reset the grub screws (I paid close attention and saw it rotate into place) and loosened the bearings and am performing a test print. Thanks so much for your time on this.

Posted : 23/07/2019 1:48 am
Tim
(@tim-m30)
Illustrious Member

Just understand we are only trying to help; and we aren't there looking at the printer.  It could be something amazingly simple causing the problems, but just not something you see or feel is important to report on, so it's basically blindly working through a mental list of most likely causes.  Sometimes we hit the nail the first whack, sometimes we miss entirely.  

 

This post was modified 1 year ago 2 times by Tim
It is always wise to get more than one opinion......
Posted : 23/07/2019 3:37 am
robin.a liked
robin.a
(@robin-a)
Estimable Member

for a mod to the Y axis bearings i can recoment this type of mounting. One of my mk3 printers has them and i have never needed to touch them since installation

https://www.thingiverse.com/thing:2823171

Bear mk3 y axis bearing holder

Posted : 23/07/2019 9:19 am
taylorstrait
(@taylorstrait)
Active Member

I believe the issue has been resolved (fingers crossed!) -- my latest 9hr test print involving 8 separate parts was a success.

Based on the feedback in this thread, I think I have diagnosed what happened. The axis bearings were overtightened. So there was some shifting. I fix this, I tightened the belts, which then put a lot more strain on the grub screws and caused shifting. I loosened all bearing connection points until they were just tight enough to grip the bearings. Then I loosened the y-axis belt. Then I redid the grub screws. So things were too tight on my unit, and loosening them seems to have resolved the problem that had affected my prints ever since I put it together. The printer also runs extremely quiet now.

Thanks everyone for taking the time to help troubleshoot this!

Posted : 23/07/2019 1:40 pm
robin.a liked
robin.a
(@robin-a)
Estimable Member

If you start hearing a clunk or thud as the Y xais moves its the U bolts on the bearing coming slack. No major issue but annoying.

Thats the reason i fitted the new type bearing holder.

There is too much emphasis put on belt tightness and it causes more issues than it fixes. I pay no attention to the belt numbers and just make sure there not able to slip over the gear teeth.

Happy printing again

Posted : 23/07/2019 2:30 pm
Page 1 / 2
Share:

Please Login or Register