Chaotic mess... is this underextrusion?

[joatrash 1]

After a promising start with my new UM2, I have started running into problems. Two major prints have failed spectacularly:

http://umforum.ultimaker.com/index.php?/gallery/image/4090-chaos1/

http://umforum.ultimaker.com/index.php?/gallery/image/4091-chaos2/

Here are my settings:

http://umforum.ultimaker.com/index.php?/gallery/image/4095-settings-p1/

(How do I get these into a text file?)

Using standard Cura settings besides lowering the print speed slightly and the brim to 5 instead of 20 since the models have a lot of flat areas. The failure occurs at random intervals somewhere mid-print. Since the print is so long it's hard to catch the machine "in the act". These should be simple prints to do.

Is this a heat issue? An issue with filament not being pulled through?

 

[geeks 1]

Bonjour,

 

C'est un problème connu sur cette machine.

 

Vous pouvez lancer un test d'extrusion comme sur cette série de post (http://umforum.ultimaker.com/index.php?/topic/4586-can-your-um2-printer-achieve-10mm3s-test-it-here/?p=40984). Vous verrez alors que l'extrusion est réussi pour des vitesses basses.

 

Soit alors vous diminuez votre vitesse, soit il faudra en discuter avec ceux qui ont eu ce genre de problème.

 

I have the wrong post when writing my answer. It was meant for someone else!

 

Tennais do not count!

 

Sorry.

 

[gr5 2,172]

This is partly why we suggested one print at a time instead of all at once.

 

(How do I get these into a text file?)

 

After you post to the gallery, start a new post and click "my media" next to the smile face.

So what is causing this? I'm not sure. It looks like some parts are printing perfectly and other's aren't? Is that correct? Or do you have prints from 2 different runs placed on the same platform?

1) It sure looks like much more glue than you should need. You can't really see the glue on my machine - it's very subtle. Too much glue can clog the nozzle on the first layer. I doubt this is your issue but still...

2) What's your print temp and bed temp? Bed temp off (0C) might explain some of this - I recommend 60C.

3) It looks like some parts are perfect and other's are printing over thin air - is it possible that some of the parts don't start until maybe the 5th layer or so? Because it looks like it is printing them in air or that they never stuck to the glass on the first layer. Maybe your levelling is off towards the front of the machine so that it is not printing on the glass on the parts near the front but printing above the glass.

Or maybe you combined different prints in the same photo and that is confusing me.

This does *not* appear to be typical underextrusion despite what geeks says. Although it is possible.

 

[gr5 2,172]

Looking more carefully it would be nice to see a larger picture. This picture is so small it's like on a camera from 20 years ago that uses a floppy to store the image. I'd like to see more detail on the first layer which means I need to see the brim better as that's the only thing left visible from the first layer.

Or better, when you start the print, take a picture a few minutes in before it starts the second layer. It really looks like first-layer issues. I think you need to level more carefully.

 

[joatrash 1]

This is partly why we suggested one print at a time instead of all at once.

 

Well, I did lessen the number of objects considerably (and cut and repositioned them to print better). These are super-simple pieces that the UM2 should handle effortlessly even if there are several of them on the bed. (It would be rather useless if I had to print them all separately.)

 

After you post to the gallery, start a new post and click "my media" next to the smile face.

 

Ah, thanks. This forum is rather different than most others in some respects. (I keep going crazy because the "follow" button isn't automatic on new topics!)

 

So what is causing this? I'm not sure. It looks like some parts are printing perfectly and other's aren't? Is that correct? Or do you have prints from 2 different runs placed on the same platform?

1) It sure looks like much more glue than you should need. You can't really see the glue on my machine - it's very subtle. Too much glue can clog the nozzle on the first layer. I doubt this is your issue but still...

2) What's your print temp and bed temp? Bed temp off (0C) might explain some of this - I recommend 60C.

3) It looks like some parts are perfect and other's are printing over thin air - is it possible that some of the parts don't start until maybe the 5th layer or so? Because it looks like it is printing them in air or that they never stuck to the glass on the first layer. Maybe your levelling is off towards the front of the machine so that it is not printing on the glass on the parts near the front but printing above the glass.

Or maybe you combined different prints in the same photo and that is confusing me.

This does *not* appear to be typical underextrusion despite what geeks says. Although it is possible.

 

It's actually photos from two different print runs. Two completely different files. It doesn't appear to be a first-layer issue because it happened halfway through the print the first time around. It's like all of a sudden the filament stopped sticking to the previous layer. The ambient temp in the room didn't change... the only thing I can come up with is if the filament had trouble unspooling. I'm using Faberdashery stuff and it didn't want to properlyspool around the stock UM reel so I've had it laying on the side. (I'm going to try a lazy Susan setup instead)

1. I added lots of glue because I had initial problems with models coming loose. Also... I'm finding it extremely hard to get models off the build plate without lots of glue on there. (The glue makes it easier to get a spatula under the model.)

2.Whatever is standard for the UM2. (I don't even know where to check this yet... I've been afraid to tamper with those settings.)

3. I think, after som further testing today, in one instance it may have been a 1st layer problem but the second photo (the one BIG piece I am holding in my hand) is the bigger mystery. The print was more than halfway through when the problem started but it does seem to have begun when some details were being "capped".)

I did level the bed rather carefully... I may have been TOO nipicky.

I've got the spool set up a little diffently now... going to start another long print as it seems to only happen on longer ones- something else that makes me suspect the spool.

 

[IRobertI 520]

1. Are you trying to remove the prints before the bed has cooled down? If you try while the bed is still warm it will be nigh impossible. When the glass is cold prints almost pop off by themselves. You do not want lots of glue, at all. Put on a thin layer of glue then wet a piece of cloth, use the cloth to dissolve and spread the glue. When the water dries you'll be left with a very thin and even layer that you can barely see.

2. You can check that in the material settings menu on the printer.

3. Just a thought. Is your third fan running properly? The one on the back of the print head. It should turn on the second power is switched on, and stay on.

You can never level too perfectly

Also a clarification. When we're talking about printing sequentially or "one at a time" we mean to use cura's ability to put several pieces on the build platform but print them one at a time. It'll print one piece, raise the bed back up and start the next one and so on. This way if the third piece fails you'll still have the two previous pieces and don't have to re-print those.

 

[joatrash 1]

1. Ok I'll give it a try. But I ran cold water on the prints and it was still extremely hard to get them off... I had to really bang on a few of them!

2. Ah.. gotcha. Should it really be necessary to mess with temp settings when doing simple files on a stock machine?

3. Will double check but I think it was running.

About the sequential thing... if I print that way, chances are the first prints will be knocked off the bed, no?

Anyway... I've got MUCH bigger problems now. I've have to send UM a support message because the X axis (I assume... the one that runs left-right) doesn stop! It causes the print head to HIT THE LEFT WALL and KEEP GOING, causing the entire thing to sound and shake like a minor earthquake as the belt "skips". (I'll start a thread about this too, shortly...)

 

[gr5 2,172]

(I keep going crazy because the "follow" button isn't automatic on new topics!)

 

Oh - it's an option in your settings. I have it turned on so if I post anything in a thread I automatically follow it and get emails of every posting.

After reading your responses I don't think your leveling is necessarily related. I was just confused by you having multiple objects from 2 different print attempts.

DON'T be afraid to mess with default temps. The bed defaults to 75C which is much too hot and will give you distortions on the bottom 5mm or so of your part. It should default to 70C or colder. Most people seem to like 60C. Anything from 60C to 70C should be fine but I recommend 60C as so many people like that temp and it doesn't take as long to heat up to that temp or cool back down again.

 

The glue makes it easier to get a spatula under the model.)

 

Wow! I never heard that before. Some people may think that's a stupid idea but I like it - it's "thinking outside of the box". I might try that some day. For now my spatula has been sharpened as much as a razor blade using files and sand paper and it slips under brim okay.

So it sounds like printing suddenly stops for no apparent reason. I know, I know - that was your original claim but I am starting to believe you. The fan comes on around the same layer - around 1 or 2mm up - there is a small chance it comes on too fast or too strong and the PID controller that regulates the nozzle can't handle it. Raising the nozzle temp to 240C will GREATLY reduce this kind of problem.

In fact raising temp to 240C can help all kinds of issues so I would raise it to at least 230C. When you start a print, go to TUNE menu before the printer starts moving (while heating the bed). At that point adjust bed temp to 60C and nozzle temp to 230C. Don't go over 240C unless you are really printing incredibly fast.

 

[gr5 2,172]

Anyway... I've got MUCH bigger problems now. I've have to send UM a support message because the X axis (I assume... the one that runs left-right) doesn stop! It causes the print head to HIT THE LEFT WALL and KEEP GOING, causing the entire thing to sound and shake like a minor earthquake as the belt "skips". (I'll start a thread about this too, shortly...)

 

This can be caused by a few things. Cut power and locate the X axis limit switch hidden under the top there. Push the print head to the left and see why it isn't hitting that switch it is most likely either:

1) The linear X bearing on the print head has slipped out and is sticking out of the head on the left side and hits the left block when you push it left.

2) The Y rod that is supposed to touch the limit switch is either popped out of the back block or isn't sticking far enough past it to reach the limit switch. This will be extremely obvious if you watch what is supposed to hit that switch.

3) The X limit switch is loose or has moved to the left - to fix, loosen the 2 screws, slide it as far right as it goes in those slots and re-tighten. Don't tighten too much as the threads strip in the limit switch easily - just tighten enough so it won't slip again. This is the least likely problem.

 

[gr5 2,172]

2. Ah.. gotcha. Should it really be necessary to mess with temp settings when doing simple files on a stock machine?

 

No but unfortunately, the default is 75C for PLA and it should be lower. The nozzle temp should normally be fine at the default but different brands and colors of PLA have different melting temps.

 

[illuminarti 18]

Turn the power off, and slide the head slowly to the left. The back sliding block should trigger the x-endstop, which is mounted inside the top panel of the printer. You should see the block catch on the lever of the switch, and hear it click.

If the head (or the fans) hit the side of the printer before the switch clicks, then that will cause the problems you see. In that case, just take a pair of long nosed pliers, and gently bend the lever a little bit in the middle, so the front-most part sticks out just a fraction more, and the block then triggers the switch sooner. You could also possibly bend the left side fan in and down a bit, if that's what's catching before the switch triggers.

If the switch is clicking, before the head hits, then the switch may be defective, or perhaps unplugged at the other end.

 

[joatrash 1]

I will check this righ after dinner. Darn you guys are fast... you answered while I was writing a new post! THANKS!

 

[vincentp 0]

hey there joatrash ! glad to see you here.

mmm... so... more LG parts in the burner ? :mrgreen:

good luck with your machine tuning, I do have a UM 1 and got a lot of configuration issues but it's not working allright !

 

[joatrash 1]

I am bringing this thread back, because I am seeing this problem more and more.

After printing about 1cm up (more o less) the print starts failing and "skips" a few layers and starts pulling strands of filament all over the place, only to resume again as if nothing was wrong. I'm having difficulty finding any consistent pattern besides the height, which makes me wonder if there is a hardware error or damage to the mechanisms that maybe stops the print bed from moving properly. It seems to ALWAYS happen at the same height.

Some things I can more or less conclude:

- bed or print head temp has little effect (tried different settings)

- material feed possibly catching on something does NOT appear to be the issue because I've tried many different variants and I know that at least one of them is guaranteed NOT to "catch" anywhere. If there's an issue it could be within the feeder... but why would it be so consistent?

-bed has been leveled very carefully.

-doesn't matter if I print ONE big item, one smaller item or lots of items... it happens anyway.

-it doesn't ALWAYS happen... but maybe 20-25% of the time.

I'm out of material at the moment but when I get new stuff I will TRY to catch it "in the act".

 

[IRobertI 520]

If you're suspecting something mechanical is happening at the 1cm height you could test this by manually controlling the bed and measuring its travel. Download Pronterface to control the printer manually, it's fairly self explanatory. You can find the com port to use in "enhetshanteraren" (sorry English speakers the english name is escaping me at the moment) under "Ports". And the speed should be set to 250000.

Then move the bed down in steps and measure that it is moving the same amount each time. I've done this by simply holding a digital calliper against the side of the machine and letting the bed push/pull it closed/open.

 

[joatrash 1]

Well, the hardware thing is the only thing I can come up with unless there is something wrong with the software. Wil measuring with a caliper be reliable though? I mean, the build plate is a bit "springy"- so much so that measuring 100 microns can be difficult.

Here are some more photos. I also noticed that the "split" seems to be preceded by a couple layers not aligning properly (see the red arrows). It's not in the EXACT same spot each time, but pretty close... and the differences COULD be due to the position of the bed leveling screws in comparison to the screw that holds/moves the platform. I'm going to send this to UM support as well.

Note: I did some cleanup on the blue parts... they looked just like the grey one at first with strands of plastic everywhere.

 

[joatrash 1]

My headache grows.

I just got home and did a test.

First, is the ultimaker 2 supposed to make a NOISE when the build plate is lowered during printing? Whenever my buildplate is lowered as part of a print, I hear a short "grinding" noise for a fraction of a second. Kind of like a "Ksssht". I felt the plate screw with my finger as this happened and I felt it turn slightly so I know that it occurs when the plate is lowered.

Second, I watched the print going and it looks like at one point, the build plate fails to lower for maybe a level or so, causing the filament to "smush" as it is printed. This creates the misaligned layer that shows in the prints.

Third, a few mm later I watched as the build plate DROPPED 1-2mm in one go with a small "kadunk"! The print immediately failed as the new layer had no old layer to grab onto and I stopped the machine.

So, the question is- what is causing this? A loose gear or something somewhere? Any ideas?

//Joe

 

[Dim3nsioneer 557]

Sounds to me like the coupler of the z axis and the z motor has come loose...

 

[joatrash 1]

Yet another discovery. My build plate is LOOSE. I can pull it up or down maybe 2-3mm. The screw doesn't seem to move, so there is something loose inside where the plate attaches to the screw...

Would the coupler be inside the housing just behind the buildplate?

 

[gr5 2,172]

You are doing great! You have definitely proved the problem is in the Z bed movement. If power is off you should be able to raise the bed by pushing up and down on it firmly but don't do this on the front edge - do this near the back of the bed near the screws and such.

It is true that the Z motor is very noisy compared to everything else on the printer so it's hard to say if what you heard was normal or not. Maybe 10X louder.

Is your Z screw coated in green grease? It should be. You should have gotten a tiny packet of green grease with your UM2. But I don't think that is the problem. I think something else is loose/broken. It's okay to *oil* (not grease) the 2 vertical metal cylinders with a light petroleum oil. Any petroleum oil will do preferably with as few as possible additives.

 

[Dim3nsioneer 557]

Oops... I missed the detail that you have an UM2. There is no coupler between motor and z screw. The z screw is part of the motor itself. Maybe the motor itself has come loose?

 

[joatrash 1]

Yeah, the Z-screw is lubricated enough- something inside is probably not screwed down correctly or has vibrated loose. I can easily move the plate up and down.

Here's a video of it:

 

 

[IRobertI 520]

Yeah.. that's definitely not how it's supposed to be, there shouldn't be any play at all. Unfortunately I haven't personally torn the bed assembly apart so I don't really know where to start on that.

 

[joatrash 1]

Well, I removed the two screws holding the cover and looked under it. There's a brass washer/ring held in place by four bolts to the build plate. I assume that the brass ring is threaded inside so that the screw can move it. This brass piece was NOT loose, so there is either a washer or something underneath that is faulty or the brass piece itself was machined defective. There's not much more I can do until UM get back to me with instructions. Removing two screws to look under the cover is one thing but I'm not going to start disassembling the build plate without the go-ahead from UM. (I can't fix anything until I get replacement parts anyway.) Sigh... I'm in full production of a prop kit at the moment too so I can't really afford to be without the printer for long.