UM2 retracting back with retraction off, very fast, and only grinding, why, how to stop?

[Rm7 1]

Hello,

My UM2 retracts a lot and fast once every couple of hours, always grinding filament only, not retracting, while printing support, but retraction being off.

What could be the reason for it? Tried updating firmware, changing slicers between the old Cura and the most recent official one downloaded yesterday.

The G-code, retraction just happened at around 3mm height +-0,5mm:

https://drive.google.com/file/d/0B01Jt12VWzuPOTh2VnY0LVhqbzQ/view?usp=sharing

Thanks,

Regards

Paulius

[Dim3nsioneer 557]

Sounds to me as if you experience skipping of the extruder motor. As it happens at 3mm height, it's most probably because you try to print to fast and/or at too low temperature. Can you please give us those numbers (speed and temperature)?

[Rm7 1]

Speed around 20mm/s at 0,1 layer height, 212C.

The problem manifests itself in a different way to regular skipping:

It is printing slowly happily quietly and suddenly there's a noise like when changing filament, the extruder motor starts going back at rapid speed, of course grinding the filament. And then the printing resumes but without any returning of the retracted filament. It happens rarely but sometimes multiple times per print at the same spot. Sometimes the print manages to continue with ~2 layers hanging on air, mostly the print fails after this "stunt".

[Dim3nsioneer 557]

When the extruder motor skips it does indeed go back rapidly... like in

.

As temperature and speed correspond (we are talking about PLA, nothing else, are we?) you might have an issue which blocks the extrusion somehow. Could be something in the nozzle, or something in the Bowden tube. Try a cold pull if you don't have yet.

edit: What do you mean by

 

when changing filament

?

Edited August 1, 2015 by Guest

[Rm7 1]

edit: What do you mean by

 

when changing filament

?

 

Hello, I've meant that the motor starts rapidly moving backwards for a long time as if it would be trying to change the filament.

I'll try to find the defective line in the G-Code, hopefully it is at fault here. But I'm not sure if I can search for it effectively. Was hoping someone would have had this problem before. Will write if I get to find it.

[mostlyatomicbob 34]

I am having the exact same issue!

It usually happens to my printer just before a change in the Z layers.

It then looks like the plate has dropped by more than one layer.

Out of interest how new is your machine? My UM2 is pretty much one of the first made, but I have been having this issue since I replaced the main board a few weeks ago. When I contacted support they seemed to think it is something to do with the latest boards.

Bob

[Rm7 1]

Out of interest how new is your machine? My UM2 is pretty much one of the first made, but I have been having this issue since I replaced the main board a few weeks ago. When I contacted support they seemed to think it is something to do with the latest boards.

Bob

 

Sorry to hear you're at the same boat though strength is at the numbers so more chance solving it!

I have 2 UM2 from very early on but one of them has the motherboard changed this Spring.

Now I'm delegating our intern to write a Python parser of the GCode to extract E moves only and search for where E movement is going back instead of forwards.

Hopefully it's at the Cura end as if it is a Hardware issue, sad prints acoming.

[mostlyatomicbob 34]

Morning,

So it happened again this morning on a print I printed with no issues on Friday.

So it is not an issue with the g-code

[Rm7 1]

Happens to me at almost every print now.

I've tried slicing with Cura 14 and downgrading firmware to that of Cura 14. But it always decides to retract, skip a few layers and then continue printing without detracting back.

Attached is pictures of the result and G-Code which was being printed.

https://drive.google.com/file/d/0B01Jt12VWzuPYXd2WHZfQXhMV0U/view?usp=sharing

[mostlyatomicbob 34]

And again......

this time there is a long retraction (as if changing material)

So then there is no material in the hot end, and the printer air prints for a while until the material is fed back through!

Do none of our resident experts have a view on this?

[Rm7 1]

And I've began to notice a similar problem with another previously good printing UM2:

https://onedrive.live.com/redir?resid=F8D3D45A21801F82!158164&authkey=!AODkrHCBIdc5mSY&ithint=folder%2c

It has started to constantly skip layers by mischievously retracting. The end layers at the legs is this occuring again just at the end of printing. And the video in the same link shows how weirdly the printer is acting after finishing the print!

I am not sure though if this behaviour is connected to the original problem.

Sliced with Cura 15.06.03.

I was told at the Google Group that it has been occurring to others and easy to fix but the guy has ironically suggested me to post this in the forums

Lets hope someone with an answer hears us!

[gr5 2,168]

I was told at the Google Group that it has been occurring to others and easy to fix but the guy has ironically suggested me to post this in the forums

Lets hope someone with an answer hears us!

 

Okay, that was me. I didn't see this thread until Joris just pointed it out - sometimes I miss a thread. I'm following this one now.

I thought you were talking about normal extruder-suddenly-slipping motion - is that what you mean? Is it somewhere about a 1/4 turn on the extruder?

[gr5 2,168]

So there are 2 things that come to mind when you mention the Z is also moving.  One is that there may be a new issue that UM is dealing with were some of the new boards have too much current and sometimes the Z axis drops suddenly for apparently no reason.  It seemed to be related to too much current which seems strange but maybe it's an overheating thing either on the board or in the stepper motor (probably in the board).

One fix that worked for some people was lowering the current to 1200 ma.  I'm not sure if that's the Z axis only or all the axis.  You can do this on a single print with gcodes:

M907 Z1200     (1300 is default)

or

M907 X1200 Y1200 Z1200   (to do all 3 axes)

or you could just try removing the bottom cover and print with the printer on it's side (it prints fine upside down also) and with a fan blowing on the circuit board - just to see what happens (is the air around your um2 warmer now than it was in the winter?)

It might not be a heat issue and might just be something weird on the new boards - I know for sure if you use too much current the steppers do very strange erratic movements and make strange sounds.

The second thing that comes to mind is that there is still a bug in marlin where it sometimes moves some of the axes much faster than they can possibly move but only when moving at least 3 axes at once (and if the extruder is freaking out then all 4 axes).  Normally this doesn't happen and the Z moves alone but if there is a gcode that moves Z at the same time as x or y that could possibly explain one of your issues.  The bug is very consistent such that if you do the same set of gcodes 3 times it will consistently move the Z axis (and maybe E also) well above it's max speed.  This seems less likely the problem.

There may be a 3rd issue.  It would really help to see a video of the extruder when this happens.  With audio.

Edited August 15, 2015 by Guest

[mostlyatomicbob 34]

Thanks Gr5.

I will print some feet and see if I can find a fan to blow onto the board.

[mostlyatomicbob 34]

Unfortunately I cannot do videos as taking photos/videos on my current project is cause for instant dismissal.

[gr5 2,168]

I would try the lower current first because that helps both with over temperature issues (really rare unless it is very hot where your printer is - like 35C or 90F) and it also fixes some issues some people saw recently with some defective boards or parts or something.

[mostlyatomicbob 34]

I have dropped the current on both the Z & E and so far it seems to be working out.

I only dropped both by 50 as a test but all good so far.

[gr5 2,168]

E also? Well E is less current by default then the other 3 axes. I think maybe 1200 by default.

If you connect to the printer using pronterface -- when it first connects it should tell you the currents on all the axes I think. You can change them and then do an M500 and I assume it will save the new default currents but I'm not certain. It might not be possible to save current as one of the savable settings.

[mostlyatomicbob 34]

So is this a known fault with the Board??

I am encountering his again(despite lowering the current)

If it is a fault with a batch of boards surely I should be able to get a replacement?

[Rm7 1]

Hello, lowering the current has made a huge improvement on one of my UM2s:

https://onedrive.live.com/redir?resid=F8D3D45A21801F82!165597&authkey=!AAlLutGmiMcBEu0&v=3&ithint=photo%2cjpg

But the other had a much weirder problem when it would not as much as would skip steps in Z, but it would overshoot Z to start printing on air.

I plan to film the whole process with a couple phones in a couple of weeks to illustrate when we start needing that printer working again.

Thanks for the lowering current suggestion!

 

So there are 2 things that come to mind when you mention the Z is also moving.  One is that there may be a new issue that UM is dealing with were some of the new boards have too much current and sometimes the Z axis drops suddenly for apparently no reason.  It seemed to be related to too much current which seems strange but maybe it's an overheating thing either on the board or in the stepper motor (probably in the board).

One fix that worked for some people was lowering the current to 1200 ma.  I'm not sure if that's the Z axis only or all the axis.  You can do this on a single print with gcodes:

M907 Z1200     (1300 is default)

or

M907 X1200 Y1200 Z1200   (to do all 3 axes)

or you could just try removing the bottom cover and print with the printer on it's side (it prints fine upside down also) and with a fan blowing on the circuit board - just to see what happens (is the air around your um2 warmer now than it was in the winter?)

It might not be a heat issue and might just be something weird on the new boards - I know for sure if you use too much current the steppers do very strange erratic movements and make strange sounds.

The second thing that comes to mind is that there is still a bug in marlin where it sometimes moves some of the axes much faster than they can possibly move but only when moving at least 3 axes at once (and if the extruder is freaking out then all 4 axes).  Normally this doesn't happen and the Z moves alone but if there is a gcode that moves Z at the same time as x or y that could possibly explain one of your issues.  The bug is very consistent such that if you do the same set of gcodes 3 times it will consistently move the Z axis (and maybe E also) well above it's max speed.  This seems less likely the problem.

There may be a 3rd issue.  It would really help to see a video of the extruder when this happens.  With audio.

 

[gr5 2,168]

How hot is it in Lithuania? What is the air temp? I recommend removing the bottom cover and adding a fan blowing under the printer.

[SandervG 1,521]

So is this a known fault with the Board??

I am encountering his again(despite lowering the current)

If it is a fault with a batch of boards surely I should be able to get a replacement?

 

Hi Bob, If we can establish a default in the board you are entitled to a replacement.

Lets see if we can isolate the problem and figure out the solution!

[mostlyatomicbob 34]

Thanks Sander,

I am in contact with Ultimaker GB and hopefully this will be sorted soon.

[zxen 4]

Happening to me too all of a sudden with my Ultimaker 2 with 2+ upgrade kit and replacement heater bed. I am trying a different print object to see if it happens again. The symptom is that the filament suddenly retracts about 15cm, then pushes straight back to where it was and continues printing a couple of times, and then the table drops about a centimeter and nothing continues printing in mid-air.

[gr5 2,168]

This sounds like a bug with either cura or firmware. Could you post the gcode file somewhere? Maybe dropbox. Thanks.