A Tale of Two Printers

[mendells 55]

We also have 2x UM2s and one of ours started having the exact same problem with stringing a few weeks ago. So far I've replaced the PTFE tube, nozzle, thermal sensor and heater with no change in results. I'll look at the set screw but I have my doubts that that is it. It feels tight from the outside. With "Move Material" on the motor "brake" is on and I can't feel any play in the drive mechanism. Both printers are running the exact same material, the exact same settings and the exact same program with different results.

[gr5 2,167]

Hi guys - I'm late to this thread. Reading all the posts from the start I've slowly picked a theory that explains everything and is trivial to test for. But a little tricky to explain. Plus the theory could be wrong. Anyway the super quick resolution would be to increase retraction from 4.5 to maybe 6.5. But there is a better solution...

the problem definitely seems to be related to retraction. It's doing *some* retraction but not enough. I also think it could be partly temperature - the temp sensors can vary by +/- 10C but you lowered temp by 20C and although that helped it was not enough. So here's my theory.

If you make sure there is no pressure on the filament (standard situation when printer finishes a print and is on or off). Then lift up on the bowden at both the print head and at the feeder and see if there is some play. It's easy to put the bowden tube back in such that there is play - even with that little horse shoe clip installed properly. To remove all play you have to know how the bowden holder thing works - but first realize that for every mm of play you have at either end of the bowden you need to increase retraction by the same amount.

To reduce play (only if it moves up and down by at least a half mm) realize first that there are 4 metal blades in the bowden holder. Loosen the 4 long thumb screws 3 times (about 1.5 full rotations) so that there is about .5 to 1mm play in the entire print head. Then remove the horse shoe clip, then push down on the bowden holder ring. Only move the bowden in and out of the head while that is being pressed down by the other hand. Now push down on the bowden firmly and also lift the bowden holder ring. Then tighten the 4 thumb screws so that the head is tight. Now the bowden should be tight enough that it doesn't move even without the horse shoe clip. Reattach the horse shoe clip anyway (you don't want to lose it I suppose).

If I'm wrong about play in the bowden at the head or at the feeder then alternatively simply increase your retraction distance by 1mm at a time and test again at 5.5mm, 6.5mm and 7.5mm.

[gr5 2,167]

foehnstrum's theory has merit also. Something wrong with retraction.

[taikonaught 2]

Thanks for the suggestions, I will try them out and let you know the results.

I know that retraction is working to some extent, as I can see the material shoot back up the tube, and the feeder at the back grips the filament very tightly and there doesn't seem to be any play in it at all.

One more piece of information that might or might not be useful is that the printer head is making a lot more noise than the first one. It is hard to explain the sound, but it is kind or a vibrating sound mixed with a hum, that sounds a bit like chewbacca... I have tightened every screw in the head, but the sound remains. If I pull the feeder tube towards myself while it is printing, the sound stops, but starts again as soon as I release the pressure. The sound only starts several minutes into a print.

I don't think that this is connected to the retraction issue, because that problem has been there from the beginning, but the sound is relatively new, but it is yet another issue plaguing this machine...

Thanks again for all the help.

[neotko 1,417]

Have you try to check the machine screws and everypart with the assembly manual of um2 in hand?(https://github.com/Ultimaker/Ultimaker2/blob/master/um2%20assembly%20manual%20V1.1%20_english.pdf)

Maybe it's just something displaced,etcetc

[IRobertI 520]

The vibrating sound could very well be the side fans. Try turning them off when you start hearing the noise (via the tune menu on the printer) and see if it changes/goes away.

The other noise is trickier since I'm not sure I know what a printer shaped Chewie sounds like Could possibly be a bad bearing in the head but I'm just guessing wildly here.

[SandervG 1,521]

A wire messing with the third/back fan?

But I don't see how pulling the bowden tube would stop that..

Don't overtighten the thumbscrews on the printhead, it will lead to other problems.

Make sure they are secured, but don't overdo it.

Well hey, at least you really get to know your machine and mechanics. You will know it inside out in no time! Knowledge is power!

[LePaul 321]

My printer would howl when the rods needed some oil. (In my case the printer sat idle for a few months and using the sewing machine oil on the rods fixed that sound)

[gr5 2,167]

I let @SandervG have my printer for a few days when he was in boston 2 weeks ago and when I got it back it was making some weird noises (but printing fine). One of the steppers had skipped a tread on one belt and the two small rods that go through the head were no longer perpendicular. Fixed that and it's as good as new again. Didn't affect the prints as I did a long print before I fixed the noise.

If you video the noise with your phone and post it we might be able to diagnose. But your retraction issue is more serious. did you increase or decrease the retraction? Also with some new printers people have been having troubles with E and Z axis which are fixed by *lowering* the current of all things! Very strange.

M907 X1300 Y1300 Z1300 E1250

Does nothing because those are the default currents. You could lower extruder to 1200ma instead of 1250:

M907 E1200

Just "hand" edit the gcode file and add those commands nearish to the top of your gcode file - the settings should get lost when you power cycle the machine. Not sure how to save those permanently without modifying Marlin (which isn't too bad but still...)

But I think you should increase your retraction distance by 2mm from 4.5mm to 6.5mm to see what happens. If it gets *worse* then try 4.0mm, lol.

[SandervG 1,521]

'I let @SandervG have my printer for a few days when he was in boston 2 weeks ago and when I got it back it was making some weird noises.. '

Lol that doesn't sound very good if you put it like that Sorry to hear it skipped a thread, someone must have lifted it wrong. I am happy to hear you were able to fix it. (@GR5 as always has been a huge help for us during Fab11!)

Looking forward hearing how these tips will affect your print!

Edited August 21, 2015 by Guest

[gr5 2,167]

No problem - the printer Joris messed with was a bigger deal: extruder plugged into extruder#2 so it wasn't turning, Z axis was at 2000 steps/mm instead of 200 (takes 5 minutes to spin the knob around 2000 times), heated bed power adapter was loose. But I had all 4 printers going at once this week printing up scans of teenagers at a 3D workshop a friend was running.

[taikonaught 2]

The vibrating sound could very well be the side fans...

 

Yes! It seems that the noise was indeed caused by the side fans. While it was printing, I did as you said and slowed the fans down, and the noise stops when the fans are at 50%. Progress! Should I just do that from now on, or is there a more permanent fix? Thankyouthankyouthankyou.

[taikonaught 2]

Hi guys - I'm late to this thread.  Reading all the posts from the start I've slowly picked a theory that explains everything and is trivial to test for.  But a little tricky to explain.  Plus the theory could be wrong.  Anyway the super quick resolution would be to increase retraction from 4.5 to maybe 6.5.  But there is a better solution...

the problem definitely seems to be related to retraction.  It's doing *some* retraction but not enough.  I also think it could be partly temperature - the temp sensors can vary by +/- 10C but you lowered temp by 20C and although that helped it was not enough.  So here's my theory.

If you make sure there is no pressure on the filament (standard situation when printer finishes a print and is on or off).  Then lift up on the bowden at both the print head and at the feeder and see if there is some play.  It's easy to put the bowden tube back in such that there is play - even with that little horse shoe clip installed properly.  To remove all play you have to know how the bowden holder thing works - but first realize that for every mm of play you have at either end of the bowden you need to increase retraction by the same amount.

To reduce play (only if it moves up and down by at least a half mm) realize first that there are 4 metal blades in the bowden holder.  Loosen the 4 long thumb screws 3 times (about 1.5 full rotations) so that there is about .5 to 1mm play in the entire print head.  Then remove the horse shoe clip, then push down on the bowden holder ring.  Only move the bowden in and out of the head while that is being pressed down by the other hand.  Now push down on the bowden firmly and also lift the bowden holder ring.  Then tighten the 4 thumb screws so that the head is tight.  Now the bowden should be tight enough that it doesn't move even without the horse shoe clip.  Reattach the horse shoe clip anyway (you don't want to lose it I suppose).

If I'm wrong about play in the bowden at the head or at the feeder then alternatively simply increase your retraction distance by 1mm at a time and test again at 5.5mm, 6.5mm and 7.5mm.

 

I did the trick with the bowden cable, and it did seem to help a bit. I am now experimenting with extending the retraction distance... Thanks a lot!

[taikonaught 2]

Just a clarifying point regarding the stringing problem, the lines of material that extend out from the body of the model are made up of many individual blobs, added one at a time further out from the model with every layer, not a string of hot plastic stretched out by the extruder, as it might seem. It is kind of hard to explain, so I have drawn over a photo to help illustrate the issue:

[gr5 2,167]

yes - that's very common. I've seen that many times. When it is thin enough it's very easy to remove with a candle or lighter or just pull them off with your fingers.

Different filaments string more than others. I have found white filaments tend to be particularly nasty on this type of stringing.

[taikonaught 2]

Here is a video of the machine running:

https://www.dropbox.com/s/s3rbcty5h1qksfu/20150823_185831.mp4?dl=0

[Labern 774]

Is the printed material smooth or is it gritty or rough?

[neotko 1,417]

I know nothing but that filament seems to be really really marked by the feeder.

[gr5 2,167]

I listened to the video twice. None of the sounds seemed serious. Did you check that the two rods going through the head appear to be perpendicular?

There are so many sounds - the quick vip when it does a travel move, the crunchy grindy sound when the Z moves to the next level, the rear fan, the side fans, the vibrations when it is doing infill, the hum of the steppers, it's hard to find crunchy bearing sounds among all that, sorry.

[gr5 2,167]

I know nothing but that filament seems to be really really marked by the feeder.

 

Yes, that's good. That looks perfect - you want that feeder to have one hell of a good grip on the filament.

[jonnybischof 60]

Does the UM2 feeder have a an adjustable tension spring for the filament drive gear? Maybe there's just not enough tension on your new machine so the filament slips instead of retracting?

Usually, that would rather lead to under-extrusion, not "too much material" like in your case.

As for temperature sensor inaccuracy: The UM2 uses platinum RTDs which are in general very accurate, even without calibration. Whether or not there are large differences in temp reading between different machines depends mostly on the circuitry that reads the sensor.

Just had a quick look at the Ultimainboard 2 schematics, and they actually use a pretty good circuitry with an instrumentation amplifier, precision resistors and even filtering capacitors. I think that is not where the problem comes from.

/edit:

missed the second thread page o.O

Just a thought: The Ultimaker Original frequently had problems with the temperature sensor getting electrical noise induced by the fan PWM signal. Messing a little bit with the wiring could solve the problem. The UM2's temperature sensor is much more rugged against noise induction, but maybe - especially if there is something wrong with the fan which might lead to higher than usual current in the wiring - your problem lies there.

Edited August 25, 2015 by Guest

[Labern 774]

I just got a printer that had similar issues. I couldn't get a decent print with no stringing unless I lowered the temp to 175deg and 30mm/s print speed.

I put an external temp probe into the nozzle and when the machine was set to 100deg I got a reading of 122deg.on my meter. I changed the temp sensor and now its printing fine with no stringing.

[mendells 55]

I just got a printer that had similar issues. I couldn't get a decent print with no stringing unless I lowered the temp to 175deg and 30mm/s print speed.

I put an external temp probe into the nozzle and when the machine was set to 100deg I got a reading of 122deg.on my meter. I changed the temp sensor and now its printing fine with no stringing.

 

I've got an extra thermal sensor, I'll have to give that a try.

[mendells 55]

I just got a printer that had similar issues. I couldn't get a decent print with no stringing unless I lowered the temp to 175deg and 30mm/s print speed.

I put an external temp probe into the nozzle and when the machine was set to 100deg I got a reading of 122deg.on my meter. I changed the temp sensor and now its printing fine with no stringing.

 

That seems to have been the problem! I replaced the thermal sensor and started a new print. So far no stringing and better looking walls.

[IRobertI 520]

Regarding those noises. Rattling fans are fairly common. You can try tightening the screws, but be careful, it's very very easy to strip the "threads" (there are no actual threads) which makes the screws just spin freely. There is no easy fix for this unfortunately. The way the fans are attached is simply not good. One alternative might be to remove the screws and use some very thick/viscous glue to attach the fans. Or you could try one of the alternative fan shrouds that people have designed.

The other noise I'm betting is a bad bearing in the head. I doubt it will affect print quality in any way but it can be annoying (though I bet it's mostly covered up by the fan noise once they're at full speed). If it bothers you a lot I would contact the place you bought it from and ask for replacement bearings.