Very weird model shift issue

Talk about anything.
Post Reply
Posts: 18
Joined: Wed Jan 15, 2014 12:07 pm

Very weird model shift issue

Post by Helicopter » Tue Jul 29, 2014 9:04 pm

Hi all,

after quite a while of happy printing and producing parts I have encountered a weird issue with model shift.

An STL file, that I have printed many many times before, suddenly exhibited a shift in the Y axis of about 8mm at a rather high layer (about 96mm height), after almost 9 hours of printing.

Now, when that happened once, I didn't put too much thought into it, thought it might be a jumped belt tooth, or even just the cat pushing it a little.

A few other parts, and a few copies of this one printed flawlessly then, but then it happened again. Now I was curious, but not yet really worried.
I updated the software now to 2.12.

Now it happened a third time right after the second time and I'm starting to be seriously concerned - wasting so much time for an error in the last few minutes isn't
too enjoyable ;) .

So I examined the parts and found that it always happened at the exact same layer. That alone, (and also together with looking at the y axis drive belt and finding it to be perfectly tight) eliminates any chances of a randomly slipping belt or any of the like for me. Then however, the model prints fine usually, so I doubt it is a problem with the STL either. Also, I can't relate it to a sudden change in the Software - it happened with 2.0 and 2.12. Someone mentioned errors like this happening with inadvertent rotation during calculation for printing - but wouldn't it be a 1 in 676 x 676, so 1 in 456976 chance of that happening at the very same layer 3 times? I think I've had this issue with another STL as well, totally different model and happened at a different height - since I wasn't aware of this behaviour though I have no real recollection wether it was a shift in the Y axis or something else, as I said, I only think it was.

But what is it then? I'm at a loss right now, has anyone else experienced an issue like this?

Some more info:
  • The model is always the same, and has been an untouched STL file since months, printing fine most of the time
  • The model is always printed in the exact same orientation, by importing the STL into UP and then placing it with place and rotate commands. It is always printed by itself, no other models in the print, centered on the build area.
  • My Software3 was 2.0 and is 2.12 on Windows. I have updated to 2.12 before the last (3rd) occurrance error, before that I ran 2.0 since I got my UP in January. The Update didn't seem to change anything
  • Printer is an UP Plus 2
Right now I'm trying to print it rotated 90° in the Z axis, curious about the results.

Thanks a lot for your valued input!

User avatar
Posts: 25
Joined: Tue Jul 29, 2014 3:42 pm
Location: Kansas City, MO USA

Re: Very weird model shift issue

Post by KCSteve » Tue Jul 29, 2014 10:30 pm

I just got my UP! Mini last week so I'm very new at all of this, but I had a similar problem with one set of models.
They would print fine up to a point, then basically snap off a section of layers, and maybe print the rest fine, maybe wind up 'printing' it as a bunch of unconnected strings.

I had been trying to print the part in the most solid fill mode and had arranged all four parts of the model into one print.

I went back to just one part - same failure.

Switched back to the normal fill option and it printed ok.

During this print I noticed two things that are probably related to the problem I was having.
1) The size and orientation of the part meant that when it was doing the fill it was really vibrating the platform. Quick, short, sharp snaps back and forth across the part were just shaking the heck out of things. I think that the finer pitch fill was making this even worse.

2) When the part finished I noticed that one end had lifted up off of the raft. I think this kind of warping, combined with the vibration, was causing the part to hit the nozzle and get knocked off the lower layers.

Opening the doors cut the heat buildup and the later parts for that model didn't warp.

I'm printing another model where out of the first four parts one failed. As I print the other parts I set to 'Fine' mode to make the printer run slower to see if that works.

I'll be watching for the input from the folks who know what they're doing.

KCSteve :->

User avatar
Posts: 57
Joined: Thu May 09, 2013 7:24 pm
Location: I'm a Colney Heathen, UK

Re: Very weird model shift issue

Post by Rootz » Thu Jul 31, 2014 11:01 pm

I count myself as by no means the most experienced Up! user on this forum but I have experienced this issue once or twice on my Up! Mini and it came down to something very simple: One of the bulldog clips that holds the perf board onto the build plate had fallen into the workings of the front-to-back (Y?) carriage and must have locked the belt for a second or so. This wasn't enough to cause a tooth jump but it would have been enough to stall the stepper motor for a few steps. Thereafter there was a permanent translation in the ( Y? ) movement.

Totally my own fault. When a print finishes, I hardly wait for the build plate to slop moving before I slide the perf board off it. Of course, I can reach the front clip but not the back one and the perf board clicks out, leaving the clip on the heated plate. The next time I slide a new perf board onto the build plate, the rear edge of the board forces the clip off the plate and it drops, un-noticed into the Y rail.

Could that be what's happening?



Posts: 10
Joined: Tue Aug 12, 2014 8:11 pm

Re: Very weird model shift issue

Post by middelweerd » Wed Aug 20, 2014 5:21 am

Hi All,

No solution yet?
Apart from having serious warping issues on these models I have the same problem. It has happened to me twice now. Two of the same models shifted on a certain level. A few sides are square but some curve to one side. It seems to be shifting in the direction of Y-.
The prints are set to Fine and Solid print.

Posts: 94
Joined: Wed Oct 13, 2010 10:27 pm
Location: Denver, CO

Re: Very weird model shift issue

Post by FallGuy » Wed Aug 20, 2014 5:25 pm

If you upload the model you were printing when this shift occurred I could see if it happens on my machine. Because it happens at the same level this makes it sound like a bad STL file somehow, or something that is causing the slicing program to have a problem. Do you hear any "clicking" noises when it skips? I would also upload a picture, maybe there is something we could see.


User avatar
Posts: 286
Joined: Mon May 05, 2014 4:31 am

Re: Very weird model shift issue

Post by Tiertime-Jason » Fri Aug 22, 2014 7:05 am

Your may share your file so that we could have a look.

Posts: 5
Joined: Thu Aug 21, 2014 5:35 am

Re: Very weird model shift issue

Post by Kingsten » Fri Aug 22, 2014 7:15 am

I think shifted printing is mainly due to print head being blocked. There are some possible causes base on my experience:
1. Warping of model from the perf board cause the print head to hit the model somewhere.
2. Warping of a particular structure in the model that block the print head.
3. Acumulation of dirt on the print head causing unwanted collision between model and nozzle.
4. Loosening of the belts.

it is unlikely that defective STL or slicing error causing the problem but your sharing of the file may shed some light on the casue.


Posts: 327
Joined: Tue Feb 14, 2012 11:47 am

Re: Very weird model shift issue

Post by mr6k » Mon Aug 25, 2014 5:42 pm

also check for dirt, dust build up in the belt drive area especially the top belt, i have found abuild up here fouls the bearings.


Posts: 10
Joined: Wed Mar 14, 2012 1:04 am

Re: Very weird model shift issue

Post by merengueman » Mon Aug 25, 2014 11:22 pm

Also try checking for slippage of the drive gear in the stepper motor shaft. The glue used from factory seems to be quite brittle and I found that it cracked and eventually my prints started to lean to one side - it happened to me in the X-axis so the parts started leaning forward.
After trying many(!) types of glue, I finally fixed it with a type of glue used for attaching bearings to shafts. Touch wood, but no slippage for over 10 prints now.

User avatar
Posts: 581
Joined: Sat Feb 12, 2011 7:39 am
Location: Silicon Valley California

Re: Very weird model shift issue

Post by DrewPetitclerc » Tue Aug 26, 2014 12:25 am

I never had this problem on my "UP Plus Pro", all the steppers have a flat on the output shaft and all the drive gears have a "D" shaped hole, so they are naturally "keyed".
They must have saved the cost on the die cut gears and the flat spotted stepper shafts in later models.

Drew Petitclerc
Petitclerc Designs
Owner/Senior Principal Designer, prototype, tooling and test equipment design and 3D printing

Post Reply