Inconsistent extrusion, periodic pullback (ultimaker 2)

[jdudeo 0]

I'm having a strange problem where my print starts out fine, but mid-way through the extruder periodically pulls back on the filament like a retraction but doesn't push the filament back through the nozzle, it just continues extruding at the same rate, creating really stringy prints.

It would appear that this is a problem with the G-code but I don't know why such retractions would end up in there, it happens even if I'm printing a model that doesn't require any retraction at any point.

I'll post a video and pictures in a few hours but for now here are my Cura settings...

 

[SandervG 1,519]

Are you saying your print still retracts even when you do not have retraction enabled?

On the screenshots however I do see retraction enabled..

At what speed and distance are you retracting?

Could you check if the grubscrew on the teethed gear is still secured and super tight?

You could test this by manually trying to hold back your filament?

Do you see the teethed gear slipping on the Extruder motor?

You could also put a mark on the extruder motor and the gear an see if this mark shows any off set after a print.

At what temperature are you printing?

80mm/s is maybe a bit fast for default temperature (depending on what filament you have).

 

[IRobertI 516]

I'm willing to bet good money that you're experiencing "skip back" due to the printer not being able to give you the volume/sec that you're asking from it. I'm assuming you're hearing a "tock" sound and the filament jumps back in a split second rather than hearing the controlled and a fair bit slower normal retraction?

If you're not sure which one it is, it's easy to verify by simply printing slower and see if it still happens. Try 50mm/s or even 40mm/s and see how that goes. If the print works fine at that speed you'll know that it's the former I described above (and I'm 99.9% sure that's the case).

You can try the test in the following thread to see what your limit is at the moment. Lots and lots of experimentation is being done to increase this limit.

http://umforum.ultimaker.com/index.php?/topic/4586-can-your-um2-printer-achieve-10mm3s-test-it-here/

 

[jdudeo 0]

Robert it sounds very likely that that's what I was experiencing, I've uploaded a video:

 

What seems strange to me though is that it starts off fine, if it can extrude properly at that speed for almost an hour then why does it suddenly start doing that part of the way through? And even if the speed and volume is the problem what is the rationale for that movement? What code is causing this?

Also I can't confirm that it's the same problem but I printed a much larger object, I was there for the first 2-3 hours of printing then I went home and left the printer to do its thing. When I came back in the morning it had improperly extruded layers and clean properly extruded layers both at once (as in some clean layers managed to happen after dirty layers) I thought it might be a problem with the filament spool not rolling properly and the filament getting stuck (I'm not too fond of the Ultimaker 2 spool holder because of the distance it leaves between the machine and spool causing the extruder to pull the filament sideways) but after what happened today I thought maybe it might be this same problem.

Nevertheless I'll try printing slower and see if the problem persists.

 

[woofy 12]

Yes, that's exactly the skip-back Robert is referring to, and lots of us have it to some degree.

Print the test cylinder from the link in Roberts post and tell us how you get on.

(Print it at 230C).

 

[IRobertI 516]

When you say it starts an hour or two into the print, do you mean it always does that or was it just in this particular print? If it was just this one it could be any number of things. The printer is probably just on the edge and any extra tension/friction could cause it to fail. So it could've been a coil of filament getting slightly stuck under another, a bit of extra friction as the spool rotated etc etc. It's difficult to say as none of us have figured out a definitive cause yet (more than the fact that printing slower usually "fixes" the problem). There's a gigantic thread by Nicolinux on the forum with tons and tons of theories and tests if you really want to get your hands dirty

The rationale behind the behaviour is to stop the feeder from grinding the filament do dust or possibly causing other issues. On the UM1 for example some people had problems where the bowden tube would pop out due to the extreme pressure the old feeder could put on the filament. I guess you could say that the code that is causing it is the current setting for the motor that limits its power. The actual skip back is a mechanical event rather than something caused by code. Once the holding power of the motor is overcome it will skip back which relieves the pressure in the system. In my case increasing this current setting above the default meant that I could complete the cylinder test. For others it had little to no effect.

 

[pikey 0]

Since Robert is too modest to advertise his own designs, I'd suggest checking out the https://www.youmagine.com/designs/low-friction-um2-spoolholder

I also installed a https://www.youmagine.com/designs/filament-guide.

If tension on the filament or friction from the spool was causing your issues, then these two items should help. I'm still printing the spool holder (and need to wait for my bearing to show up from eBay) but already have the filament guide installed. I noticed an improvement after installing the filament guide, but I also slowed down my general printing speed just to be safe.

 

[jdudeo 0]

Thanks for the explanations I think I've got a much better grasp on this problem now.

I printed the extrusion test, the problem appeared at around the 8mm/s mark but by increasing the temperature to eventually about 255 I was able to print the whole thing successfully.

I'm trying my print again now at 230 and it seems fine so far, I guess I should have done more troubleshooting before posting here but it seemed so odd and I had no idea what to search for.

About the spool holder design I would like to try that out, currently I'm printing my own free standing spool holder design but it would probably be very efficient to incorporate that low friction design into it.