S5 Wall/Infill Extrusion Issues

[wburnham 0]

Hi All. I am fortunate enough to have two ultimaker s5's at work with air managers. For a time, they both printed great, now, one of the two exhibits severe under extrusion on walls and infill (ie not on roof, floor, or brim). I have ruled out any mechanical issues, swapped both print heads, verified the filament drivers are functioning properly, etc. Both printers are running the latest stable firmware, and I have confirmed they have the exact same configuration in cura. I've reinstalled cura several times. My colleague doesn't have the same issue, I'd like to think I have considerable experience with printers, but this is testing me a little bit/ I don't have the time to fuss around with it. I've played around with the slicing settings but nothing seems to fix it. We have had an intermittent issue with this printer maintaining temp during endurance prints, but it is not consistent enough to diagnose or say it is related to this issue.

Any suggestion are welcome.

 

-WB

[wburnham 0]

Okay, I have just done a fresh install of Cura and reset both printers. Both are running the exact same nozzle and filament setup, regular ultimaker pla, and AA 0.4 nozzles. The problem persists on one, and is absent on the other. I monitored the printer in web interface during the print. Everything was correct as far as I could tell. I chose to DOUBLE the flow, and half the print speed. It is extruding closer to the correct amount of material. What would cause such gross under extrusion on factory settings? So weird to me. No extruder slip or any mechanical issues.

[gr5 2,172]

First and foremost - check that the front fan of the 3 print head fans is spinning when you get a printcore above 60C.  So you can just go to the middle menu, click on the filament, click "..." in corner and click move.  Then don't move, just wait for temp to reach 60C.  Then open front door and make sure the fan is spinning.  When it isn't, this can cause underextrusion.

 

Then try swapping out the printcore.  Even before I read that you have had temp errors in the past.  There are many possible failure modes including issues with teflon, nozzle, and temp sensor.  The most common error with the temp sensor is a wire/connection that has some extra resistance.  This causes the temp to read high which means the printer heats to a lower temp.  This causes lots of underextrusion.  print cores are consider consumables like filament.  You should be spending much less on printcores than filament but still, they need replacing but just do a test by swapping left/right if they are both AA 0.4, or if the right core is BB then every printer comes with 2 AA 0.4 and a BB 0.4 so find one of those extras.

 

Thirdly, test the feeder.  There are parts in the feeder that can cause issues.  Or you could just clean it.  It's pretty easy to disasemble.  There are videos and pictures at ultimaker support.  Anyway - how to test the feeder - go back to that same MOVE option and this time move the filament a little and fight that movement by grabbing the filament just below the feeder and fight it hard.  The feeders can pull about 15 pounds before slipping (about 7kg).  Anything under 9 pounds I consider a fail although it will still print okay down to about 5 pounds.  10 pounds is good enough.  Hopefully you don't have the material station as then this test is much more difficult.

[wburnham 0]

2 hours ago, gr5 said:

First and foremost - check that the front fan of the 3 print head fans is spinning when you get a printcore above 60C.  So you can just go to the middle menu, click on the filament, click "..." in corner and click move.  Then don't move, just wait for temp to reach 60C.  Then open front door and make sure the fan is spinning.  When it isn't, this can cause underextrusion.

 

Then try swapping out the printcore.  Even before I read that you have had temp errors in the past.  There are many possible failure modes including issues with teflon, nozzle, and temp sensor.  The most common error with the temp sensor is a wire/connection that has some extra resistance.  This causes the temp to read high which means the printer heats to a lower temp.  This causes lots of underextrusion.  print cores are consider consumables like filament.  You should be spending much less on printcores than filament but still, they need replacing but just do a test by swapping left/right if they are both AA 0.4, or if the right core is BB then every printer comes with 2 AA 0.4 and a BB 0.4 so find one of those extras.

 

Thirdly, test the feeder.  There are parts in the feeder that can cause issues.  Or you could just clean it.  It's pretty easy to disasemble.  There are videos and pictures at ultimaker support.  Anyway - how to test the feeder - go back to that same MOVE option and this time move the filament a little and fight that movement by grabbing the filament just below the feeder and fight it hard.  The feeders can pull about 15 pounds before slipping (about 7kg).  Anything under 9 pounds I consider a fail although it will still print okay down to about 5 pounds.  10 pounds is good enough.  Hopefully you don't have the material station as then this test is much more difficult.

 

GR5, thank you for the thorough response.

 

I'm pretty confident it is a software or slicing issues as my colleague isn't having issues printing from his computer. That being said, I have done the following to address what you've described above:

 

I did confirm the fan is spinning using the method described.

 

I tried two brand new print cores, and cleaned the contacts on the carriage and the print core.

 

The feeders are pulling strong, after taking them apart they had very little debris (we clean them regularly).

 

Fortunately, no filament manager, so I don't have to worry about that.

 

After checking printer logging, I am pretty much certain the temperatures are good to go, unless I have used several cores in a row which have bad thermocouples.

 

I am going to run both printers with the exact same setup and view the web interface data logging. My suspicion now is that the feeders are not rotating as fast as they should. I'll let you know what I find.

[gr5 2,172]

On 2/28/2023 at 4:37 PM, wburnham said:

several cores in a row which have bad thermocouples.

agreed.  Extremely unlikely that 2 cores in a row would have bad connections in their sensors.  Probably less than 1% of cores have this issue.  Quite a bit less.  They are all tested at the factory and quality is reasonable (still they are difficult to make as you can't use solder which would melt).

 

Well you can share project files.  Or you can check the gcode with a gcode analyzer.  The file created for the S5 is typically a "ufp" file.  You can rename it to zip, open it, and locate the gcode file, then extract it from the archive.  Then there are some nice analyzers that check "flow".  they are visual and show flow by color.  Or maybe they call it "line width".  Let me check...

 

 

[gr5 2,172]

This one seems pretty good.  Make sure you have it set to 2.85mm filament - just drag and drop you gcode file onto this page where it indicates:

https://www.gcodeanalyser.com/

 

The thickness of the lines seems to properly represent the amount extruded.  It seems a bit off but I think it's correct.  If you lower the filament diameter it shows thinner lines for sure but it seems to me the linex are a bit thicker than reality?  Maybe it's taking into account the previous lines are "in the way" and it's actually correct?

 

Anyway this could be helpful.

[gr5 2,172]

This ones great as well: https://gcode.ws/

 

Similarly, just drag and drop your gcode file onto there.  Make sure you go to 2D options and enable "emulcate extrusion width" and then go into printer settings and you can adjust filament diameter but I found it most useful to purposely lie and say it was 1.3mm diameter filament which showd me the gaps in the infill and you can then see if gaps are bigger in certain areas that you are concerned with.  The update on the dispaly is instant.  Very nice tool.

[wburnham 0]

Really cool tool -Thanks for sharing. I didn't see any extrusion issues in the gcode, but it must be there somewhere. I am printing right now with 0.8mm nozzles flawlessly, so something about the 0.4mm nozzle config/profile must be screwed up. Are there and global core/nozzle settings somewhere in cura?

[gr5 2,172]

Just because the 0.8 prints fine doesn't mean the feeder isn't weak.  It could be the 0.8 prints require less pressure in the nozzle (I mean it *is* a bigger hole in the nozzle) so less force from the feeder.  It all depends on layer height.  If you aren't doubling the layer height for the 0.8 nozzle then the feeder isn't working as hard.  It's spinning faster for those 0.8mm wide lines but less force needed.

 

My favorite test of the feeder is to use the MOVE command (middle menu on printer, click on filament at top, click on "..." in top right corner, click "move") and make sure the filament is well above the print core and you can see it in the bowden.  That move the filament a bit and fight it hard with the other hand by pulling down on the filament.  Hard.  Like 10 to 20 pounds hard.  You could compare it to another printer.  Typicall S3/S5/S7/UM3 pull with about 15 pounds force but 10 pounds is definitely a "pass".  5 pounds is a fail.  not sure where to put the line - maybe around 9 pounds - if it can't pull around 8 pounds then it's a bad feeder.  There are parts that can wear out or it can have so much junk in the feeder that it barely functions and needs cleaning.  The splines on the wheels that pull/push the filament can wear down (unlikely as they are hardened steel - but it can happen).