Failed prints. Electronics error?

[Sealcat 0]

Hello!

I'm fairly new to the magical world of 3D-printing and recently bought a used Ultimaker 2 with the intent of upgrading it to a 2+ some day.

From the start I had problems with feeder skipping, I managed to get at couple decent prints but the support structures (auto generated from cura) was always stringy and had small holes in it, and larger prints (10+ hours) failed as the filament stopped extruding. I tried incresing temperature, increasing the feeder tension slightly, lowering speed and cleaning both the feeder and the nozzle with no noticable difference.

So, I switched the PTFE-coupler and the filament to a brand new roll of colorfabb white PLA, but the problems actually seem to have gotten worse. The feeder doesn't tick quite as often, but it stops extruding despite a seemingly clean nozzle.

I've managed a 6 hour print at most. Could the something else be worn out, or is there a clog somewhere despite doing the athomic method? Does the problem simply lie with the settings? In what end should I begin?

3 hours in on a 10 hours print, standard Ultimaker 2 PLA-settings. 210 C, 60 C bed, 60 mm/s, 100% fan.

What I'm most concerned about though is something weird the printer does at intial startup, the screen flickers in blue for a few seconds and a high-pitched screeching is heard. Could there be a problem with the electronics, and is something like that even possible to fix?

A video of how it looks: http://tinypic.com/r/no659i/9

I'm at a loss how to tackle these problems.

Any help would be much appreciated!

Edited November 21, 2016 by Guest

[Torgeir 242]

Hi Sealcat,

Welcome to the community.

You have some known problems for sure, but first, you'll need to upload your picture using the second icon from right. This will give us a better quality of the print. Also the video have a very lo volume, so it is difficult to analyze the "kind of" sound that's occur when switching on the printer.

For sure you have some under extrusion, can you see any wear/chafing on the filament at the point it stopped feeding?

Thanks.

Torgeir.

[Sealcat 0]

Thank you! It's just unfortunate that I had to join because of problems instead of pure printing joy.

I uploaded the picture the way you said in the first post, and added another one that hopefully gives some more insight into the problem. It has small holes, zits and stringing in the support structure and infill, but none or very little on the objects (if they get finished).

I've realized that the tension between the knurled wheen and the filament is a bit too high, the teeth marks in the filament seems too deep. I've mucked around with the tension quite a bit before, both lowered and hightened it, but it didn't make much difference either way. The white marker on the side of the motor is at the top though, so I thought it was fine. Should I lower the tension?

This is the piece of filament removed after the print in the pictures failed, the knurled wheel has ground in the filament quite a bit. I brushed out the filament dust from the feeder motor as well as it was possible without disassembly.

The sound at startup seems to be too faint for my camera to pick it up, I made another recording of just the sound. It's quite hard to catch, you'll probably need to turn the volume up a bit. The first click is the printer being switched on, then you hear the little screaming sound for a few seconds (at the same time as the screen flickers).

http://vocaroo.com/i/s1AcOGXlGU12

I hope it could just be a loose cable somewhere?

Thank you for the help!

Edited November 21, 2016 by Guest

[Torgeir 242]

Hi Sealcat,

To answer your last question first, I'll guess this is a sound from the small piezo speaker on the board? If there is no stepper movement.. hmm..

My printer have a little "flashing" during start up, but I'll guess the software people never tried to "dress" this away.. Sometimes you'll need to do some delay routines in the software to avoid such thing to occur. Well.. Power supply issues, -but I would not worry about this.

--

This looks like the very much known issue with the UM2.

Your main problem is the feeder arrangement I'm pretty sure.

I'll think you should join in here:

https://ultimaker.com/en/community/33678-is-this-an-under-extrusion-issue?page=1&sort=#reply-163776

Cause, I'll want you to try the same thing.

My best advice.

Thanks

Torgeir.

[fbrc8-erin 272]

Hi, Sealcat,

I agree with Torgeir about running the extrusion test. It was developed to run at 230 C, and will do that automatically from the provided gcode.

What's the volumetric extrusion rate you're printing at? That's your linear print speed (mm/s) X layer height (mm) X nozzle width (mm).

I'd also look at are your retraction settings, on the printer, and in Cura. It looks like there was a lot of flattening going on in the filament before it ground out, which could indicate too much retraction going on.

The last thing I'd check out is the screw tension on the 4 screws holding your feeder on. The UM2 feeders are a little more sensitive than the UM2+/UM3 feeders. A quarter turn of the screw can make a lot of difference. I usually recommend making all the screws hand tight with a screwdriver and then slacking off the screws by 1/4 of a turn. If the 4 screws are too tight, it can lead to extrusion issues.

[Sealcat 0]

Thank you very much for the answers!

So I probably don't need to worry about the flashing screen and sound? It does concern me, but I haven't had any other problems with the software.

I finally had som time to try out these methods at least.

So, I changed the temperature to 230 C, did some atomic nozzle cleaning and ran the flow test you provided, Torgeir. I didn't understand what cover to remove to observe the stepper motor though, but I didn't notice any immediate strange movement duing the print.

At 9mm/s the plastic starts to get very thin, a bit transparent, and at 10 mm/s it's held together by hopes and dreams in some places. It's better than I expected though, and I didn't hear any grinding/skipping back from the feeder duing the entire print. It did grind when I switched the material after the atomic method though.

What should I take from the result?

Just for the sake of trying I made a small, more advanced print with support structure and retractions. The quality is actually worse than before, I assume because of the higher temperature, but there was no feeder skipping at least.

It seems to be very inconsistent in what amount of material it extrudes, it's either too much and gets zits/stringing or these thin sections with small holes, mostly on the support structure.

For the print(s) that failed earlier the print speed was 60 mm/s, layer height 0.1 and nozzle 0.4, so the volumetric would be 2,4 mm^3/s? It seems quite low.

I didn't change the retraction for this print, it's set at 4.5 mm in length and 25 mm/s in speed on the printer itself, in Cura it's enabled but the boxes for retraction length/distance are greyed out for some reason? Should I try to go even lower than the 4.5 mm in retraction?

I forgot to loosen the screws on the feeder before the flow test print, but did it afterwards and a couple of them we're really tight so hopefully it did something good.

Thanks again for the help, it's tricky to find the problem with all these factors playing in! You think you're starting to understand something, then you find ten more things to consider.

[Torgeir 242]

Hi Sealcat,

Great print and it worked out better than I expected..

So, did you check the filament when finished? It's a good thing to inspect "the look of the feeder track" on the filament around the point the feeder wheel was when your print finished.  Any sign of grinding here,. etc.?

Well, I'll think you have a feeding issue. So, not to change to many thing at a time, we'll do another test printing the same object, but this time you'll be stand by holding your hand lightly on the filament's input to the feeder. This so you can feel the feeding all the time from about 8 mm³/sec., until finish.  

If you feel that the feeding stops, you have to help with your hand to push the filament in order to have proper feeding -all the way until finish.

So, next question, is the print good now, after this additional help??

We might do some hardware "adjustment", if more force is needed.

The cover I was thinking about is (this one) now removed:

Here is the feeder stepper motor:

As "fbrc-eirin" pointed out; the pressure on the filament is a little bit to high, cause it is some flattened. Adjust it back until you see "two teeth" track and leave it there. The idea is not to flatten it, but have a clear track without flattening the filament.

It might be a good idea to print a test object that's reflect the problem your printer might have, before going into a more time consuming print. This as lot's people here have done that, and can help in advising better if you are using a known printing object.

Happy printing.

Thanks.

Torgeir.

[Sealcat 0]

Hi again!

So, I did the test again, now with the cover removed. I also lightened the pressure on the filament, the teeth marks look a lot better now, two clear lines of dots.

The underextrusion started earlier this time, around 7 mm/s, I assume because of the lighter pressure on the filament. Around then I started to push it in manually, and the print was a lot better from there on to the end.

There still was some holey parts in the 10 mm/s area, but considerably less than in the first print.

I didn't see any strange movement on the stepper motor (it's very possible I missed something though) or grinding on the filament.

The print to the right is the new one, compared to the old one to the left.

Do you have any suggestions for more well known objects to print? I thought about the benchy boat, but it seems a bit big for the purpose of testing extrusion.

Thank you!

[Torgeir 242]

Hi Selcat,

Great info from you.

This is just caused by friction between the filament and the bowden tube. In addition your feeder stepper motor may be quite warm during print. Can you touch it without "burning" your finger?  This motor tend to go very hot during a long print. As you might know about the + upgrade, that's add a few new thing to the UM2 printer. Among those items a new feeder stepper motor unit, geared to be stronger and isolate the knurled feeder wheel from the stepper shaft. This will avoid heating up the knurled feeder wheel ("softening" the filament).

There is an easy way to improve your printer for this matter.

You can do it in two step, the first part is to reduce the "friction coefficient", this will reduce the force to overcome the "friction force".  As this reduce the energy used for filament feeding, we can reduce the current to the feeder as well in order to reduce the heat produced inside the stepper motor.

(Step two is to separate the feeder (knurled) wheel from the stepper motor shaft.

Will do this later to improve further if needed.)

What to do is to mirror the feeder unit, right, -turn it around.  Now the printer will grab the inner side of the loop hence less friction force to overcome.

Here's a picture of my printer from the back with two feeders installed, this is the way to installe the feeder units for minimal friction with this (standard) filament roll setup.

This picture show the new positin of the feeder unit when mirrored and the outer half is removed.

PS. Those four screws are also holding the feeder stepper motor on the inside of the printer, so leave the last screw (upper left) and prepare to hold by hand the stepper when removing the last screw. (Remove the bowden tube before start). It is smart to tape the feeder unit on the sides, this in order to prevent a sudden surprise, in here there is a strong spring and some parts that suddenly can go into different direction.. If you like to clean and check this unit, open it inside a transparent plastic bag.

Edit: There is no need to remove any PCB for this mod. Only to open two covers!

As this operation make the feeder wheel to turn in the wrong direction, we'll need to change the motor direction only.

There is two ways to change the stepper motor direction, an easy hardware change and an as easy firmware change. However, I’ll do recommend -only to fiddle with the hardware not the firmware!

Here is a picture of where the conector is installed at the main PCB (located on the underside of the printer).

To change the feeder stepper rotation direction you can swap two of the four wires on the feeder connector E1. This will be pin1 and pin two, -or pin 3 and pin 4.

Here is a picture I've borrowed from "3DVerkstan", a detail of how to release a pin from this types of connector, here a a two pin connector but same method is user for the four pin connector .

Here's another picture of the type of pin used in this connector types, also borrowed from "3DVerkstan".

All you need for this improvement/modification is a hex driver, a pliers or wrench (for the nut) and a needle.   And I’ll guess this will take about 45 minute..

This is actually all to do for this improvement.

Have a look into this, and let me know if there is any thing that's not easy to understand.

Anything, -just ask.

This modification is easy to reverse, but I'll think you never will go back after trying this "adjustment".

Stage two, will be to install two new pulleys and a GT2 belt. The pulleys are the same size, so no gearing here as this is done to have better isolation between knurled feeder wheel and the feeder stepper shaft. There will be some printing for this mode, but will not ruin you budget in any way..

Thanks.

Torgeir.

Edited December 6, 2016 by Guest