Ultimaker Material Station Errors [Maybe solved]

[UlrichC-DE 109]

Hello everybody,

 

Sporadically, the MS does not rewind the material.

 

It does not matter whether I print generic or UM filament.

There is no error on screen. He just doesn't do it.

How can this be prevented?

Edited July 21, 2020 by UlrichC-DE

[UlrichC-DE 109]

I've had a lot of problems with the material station in the past two weeks. I had printed with Ultimaker machines for many years before and had no reason to go through the forum.

 

ER998, ER63, ER52, ER56, ER65, ER57
With Material Station, I had a new surprise after or before each print.

Also sporadic effects such as empty spool turning and spool not turning.

I have probably finally found the main cause of the errors. I had previously checked ER57 and therefore all Bowden.

Maybe interesting for others:
The Bowden from the Material Station Material Sensor to the printer feeder was not installed properly. The Bowden hadn't been inserted deep enough in the material sensor or had come loose.
They were pre-assembled on delivery, so I didn't have them on my screen at first.
The first Bowden was long because of that and not fastened enough. This sometimes confused the material station when unloading.

The profiles for unloading could not be carried out properly.

Had the problem on print head 1. But it also transferred to print head 2. Probably because the feeder is not started up like the first one at the start (time is money).
Printhead 1 was almost not extruded at start up. On the other hand, the filament ran at full throttle in extruder 2.
- This often caused the PVA material in the bowden to break.
- Other materials reported clogged printhead even though it was unaffected.
- By extending the bowden on extruder 1, I had the usual "Can't unload material" problems at the end of printing.
  For TPU95A filament, this was probably expressed in ER998. With other filaments ER63. When unloading, the station moves back a certain length of the filament.
  First warm pull then cold pull in millimeter work. If the bowden is too long, only the tension from the filament is relieved. But the filament in the extruder doesn't move a bit.
(I don't have any technical documentation. Therefore, I can only observe and draw conclusions.)

 

I couldn't see the problem from the outside. It looked like the first day. I loosened the clamp and brought the Bowden into position.

So far, the printer has been running without errors for three days, one of them with the new software.

 

Hope the mistakes don't come back.

 

Thanks for watching

[UlrichC-DE 109]

addendum. At the moment after the printer starts, it looks like in the picture on the bottom of the printer. Before that, he had barely extruded when starting. Most of the time, the filament did not come at all or just before printing.

 

[TrueNorth 3]

15 hours ago, UlrichC-DE said:

ER998, ER63, ER52, ER56, ER65, ER57
With Material Station, I had a new surprise after or before each print.

 Amen. I'm new to UM, having been most familiar with Lulzbot previously. To this point have not been impressed with UM. It seems there is forevermore something wrong, aborted prints, etc. Perhaps it is due to the Material Station. 

 

15 hours ago, UlrichC-DE said:

Also sporadic effects such as empty spool turning and spool not turning.

I have probably finally found the main cause of the errors. I had previously checked ER57 and therefore all Bowden.

Maybe interesting for others:
The Bowden from the Material Station Material Sensor to the printer feeder was not installed properly. The Bowden hadn't been inserted deep enough in the material sensor or had come loose.
They were pre-assembled on delivery, so I didn't have them on my screen at first.
The first Bowden was long because of that and not fastened enough. This sometimes confused the material station when unloading.

The profiles for unloading could not be carried out properly.

Had the problem on print head 1. But it also transferred to print head 2. Probably because the feeder is not started up like the first one at the start (time is money).
Printhead 1 was almost not extruded at start up. On the other hand, the filament ran at full throttle in extruder 2.
- This often caused the PVA material in the bowden to break.
- Other materials reported clogged printhead even though it was unaffected.
- By extending the bowden on extruder 1, I had the usual "Can't unload material" problems at the end of printing.
  For TPU95A filament, this was probably expressed in ER998. With other filaments ER63. When unloading, the station moves back a certain length of the filament.
  First warm pull then cold pull in millimeter work. If the bowden is too long, only the tension from the filament is relieved. But the filament in the extruder doesn't move a bit.
(I don't have any technical documentation. Therefore, I can only observe and draw conclusions.)

 

I'm not following your thoughts on the Bowden length, but I have had repetitive problems with the collets slipping in the decoupler, and have learned not to trust them to stay put. 

 

I'd be interested in hearing others' tips on how to smooth out the myriad of errors that seem inherit to the S5 Pro Bundle. The design is still quite unstable in my limited experience (6 months). 

[UlrichC-DE 109]

Hello, I have already read posts from you here. So there are others who don't want to wait for tickets.

 

I've only had the S5 for half a year. Airmanager 4 months. The material station has been in operation for about three weeks.
The S5 and Airmanager run stable for me. The MS is a curse.
Unfortunately I need them because of PA and PVA (it actually works with it)

 

I also have other Ultimakers. (They are really good)

 

Wrote everything under reservation. Unfortunately it was my problem. The support team had questions and options.

Developing machines is usually my job. Finding out why they don't work is sometimes fun.

This is difficult with the printer because there is no deep information. But there is something worth knowing in every contribution to the UM5. The sum of the information currently has an answer. The MS is beta in my opinion.

 

I noticed one thing: the flow sensor creates most problems. This can only be fixed by firmware. The part is unfortunately mechanical and probably analog. More sensitive than any feeder but unfortunately not configurable. For this, someone should think about a calibration.  

If there are, I unfortunately don't know about it 🙁

 

My private list (V4.7.2):
- Do not touch the material station while printing
- Load only the required filament. In the order (Ultimaker-Original, Generic, Ultimaker-Original, etc.)
- Switch off the digital factory (if not required)
- Switch off the WLAN (if not required)
- Switch off home calls (This hasn't really helped so far, but it felt better)
- Do not change material profiles in Cura. (According to support, changes in material flow in MS can maybe cause an error)
- No firmware update if not absolutely necessary.
- Update only via USB. Archive the old versions.

I am currently testing (because of new problems with V4.7.3):

- If possible print 0.8 nozzles.
- To clean the nozzle after 1 week or when changing filament types

 

I'm writing that now. Perhaps my opinion will change in 24 hours. Trial and error / read repead

 

I hope it helps you a little.

Edited July 22, 2020 by UlrichC-DE

[UlrichC-DE 109]

Short update:

I just cleaned the printheads from last night. (I was running version 5.7.3 and during the first third of a 15 hour print the printhead was clogged.)
- They were clean but clogged. Fascinating. I had to wait a long time and then press firmly to get the plug out.

- He also says the material spool is empty. But it's new.
Something wrong.

I'll go back to version 5.7.2 and switch the printer back to stealth mode.

Maybe someone should test the firmware. Just set up a pair of printers and print different parts like in a production environment. In IT we call this integration environment. Developers and basement kids don't have much fun with it, but the quality of the software is definitely improving.😈

 

See you next time. I haven't given up hope yet.

Edited July 21, 2020 by UlrichC-DE

[SandervG 1,521]

Hi @UlrichC-DE, thank you for your detailed reports. I understand you are already in touch with support so they are also receiving this feedback? 

 

I am not entirely sure if the first report was saying that the bowden tube was not inserted far enough, but I also read you didn't trust the couplers; have you ever checked if they are both in spec? You should press down on the collet to be able to pull out the bowden tube. You should then be able to get the collet out of its holder. It should have 4 little metal blades on the inside which are holding onto the bowden tube. If they are not all equal size or if there are not 4, you should get in touch with your reseller and get them replaced. 

 

Your private list of workarounds to make it work sounds bizar. Do you have an idea what switching of Digital Factory does for example? (and by switching off, you mean printing via USB?)

[TrueNorth 3]

14 hours ago, UlrichC-DE said:

I just cleaned the printheads from last night. (I was running version 5.7.3 and during the first third of a 15 hour print the printhead was clogged.)
- They were clean but clogged. Fascinating. I had to wait a long time and then press firmly to get the plug out.

 

 

Aha. I've been fighting the same issues (Error 65, error 58), and like you have not been able to find any causes. Cleaning of the cores did not show any reason for the plug. No job has reached completion (including several failed prints) since updating the firmware. I came into work this morning purposing in my heart to remove the MS and just print. We need to get some jobs done with a minimal amount of fanfare, and the pressure is building. Thanks for sharing your observations, @UlrichC-DE. Maybe I'll try to retrograde to a stable firmware version instead. 

[UlrichC-DE 109]

Hello SandervG

oh yes, the first report is written a little cryptically.That's what you get when you try to squeeze three weeks into one article.

In short:

The Bowden's were not deep enough in the material sensor. Had the errors. Now i don't have them.

I checked the collets. They have four identical jaws with blades. The printer has PVC Bowden, the material station has PTFE Bowden.

My list shouldn't be a guide for others. i do not want to print anything else. Or rather i have to print because i have deadlines.

Yes i am printing again via USB like in the old days. i switched off everything that is not absolutely necessary.I do this from experience and purely technical considerations always so with machines that do not work properly. Usually i have a debugger running and know the system. In this case it is of course different.
Therefore i can't say exactly what this does. Actually, i don't want to find out what it does either. That's not what i bought it for.

But here are the considerations.

The shutdowns...
- Switching off the network allows the printer to be controlled/interrupted via network.
- Switching off the control and protocol functions (aka. Digital Factory) ensures that the printer is not disturbed by internal processes.

 

All the things you think about when the system components work on their own but show errors when used together. Or even if the system proves to be sensitive and influenceable. Then you lose some of your trust in the software and start taking measures.

 This happened to me when i put a roll of BreakAway next to a PVA roll and the printer did the Jesus number.He turned PVA into BreakAway (My first thought was. The machine works like a smartphone APP. But not like a machine for automation)
Or when the printer tried a cold pull on the TPU retraction (A new thought. Has anyone ever tested this?)

 

A few worst case examples from other projects:
- A system sends messages. These messages contain strings that are interpreted as commands by another machine.
- A process gets very busy and pulls the whole system down.
- A function is not memory or thread safe.
- A sensor is analog and time critical.
- A function does not work and creates dead loops
- A system is tested in the laboratory. The laboratory conditions do not correspond to the real conditions.

 

 i think you're gonna have somebody come around the corner with an idea soon. The printer alone is working. Only the material station is not completely integrated. The challenge is who controls when when. Which sensor can i trust when. i know these problems and can therefore not write anything bad.

 

This is only my opinion. i just want to print. 😔

Edited July 22, 2020 by UlrichC-DE

[UlrichC-DE 109]

Hello TrueNorth,

 

thanks for the info!

 

Good idea this will also be my next step if the other one does not work.

 

Unfortunately the MS cannot be switched off.
The serial connectors are very sensitive. i get sweaty if i have to change them.

 

Good luck

Edited July 22, 2020 by UlrichC-DE

[SandervG 1,521]

Hi @TrueNorth

 

I see you mention ER65 and inspecting the print core, but there are other reasons why this error can occur. Have you read this article?  If those tips didn't help, did you create a support ticket?

 

Same goes for this article for ER58.

 

Thank you for your detailed feedback @UlrichC-DE and your understanding towards the complexities when it comes to such a machine and its integration. Obviously we don't depend or request you to do any testing and you should be able to dedicate all of your time to 3D printing. But in some occasions it can help us identify the issues better and faster, which would benefit your desire to just 3D print. But I suspect and trust our product experts helping you via support are supporting you with this. 

[UlrichC-DE 109]

Hello SandervG

 

No problem, I have a hammer here if I need it.

 

I have a question about the Feeder.
Is there a description. Best way to adjust the feeder tension for a bundle with MS. It would be a pity if I had a different feeder setting than the Ultimaker in your house. This would result in the material station not working properly with the printer.

The errors would be different depending on the firmware version only the cause remained the same. 🤨

[SandervG 1,521]

Hi Ulrich, as far as I know the feeder tension should be at default tension with or without a MS. Which would be in the middle setting. Hope this helps. 

[UlrichC-DE 109]

Hello SandervG,

 

i take it from the information that i will return to the middle position in the future unless i am printing with TPU95 or other special ER65 sensitive material.

 

I will wait until the next firmware is released and continue printing until then, so far without problems.

 

Then I test the new firmware again.  By test I mean that I try to use it first.

 

Thanks for your patience.