I truly hate the MMU2S: "MMULoad failure"  

  RSS
SeattleDavid
(@seattledavid)
Estimable Member

I truly hate the MMU2S, as it is constantly failing to load.

The filament loads to the nozzle, then it unloads, and then it tries again, and again, and again.

Load...unload...load...load...unload...MMU Load Failure.

There is no useful message to indicate what the problem is. And, I have not been able to find any successful advice to prevent this from happening. Is it a spring adjustment? A broken sensor? A defective piece of software? Something else?

This happens over and over again.

Why does it do this? And, what is the problem, and most importantly, how do you fix it?

 

Posted : 08/10/2020 10:49 am
SeattleDavid
(@seattledavid)
Estimable Member

Anybody?

Posted : 12/10/2020 12:19 am
gnat
 gnat
(@gnat)
Prominent Member

You haven't provided enough information about the issue for anyone to provide useful input.

Where is it failing? Does filament make it to the extruder? Does it enter the gears? Does the IR sensor recognize it? Does it extrude at all? Is the MMU idler chewing a notch in the filament?

Try removing the PTFE from the extruder before it starts to load as this will help answer many of those questions. Pull out extra length of filament so you have slack to work with. Get a good look at the tip to make sure it is formed correctly. Manually feed the filament into the extruder and see how it acts.

MMU tips and troubleshooting...
Posted : 13/10/2020 4:36 pm
SeattleDavid
(@seattledavid)
Estimable Member

The filament goes all the way down into the heater assembly, then it retracts, and then it does this repeatedly.

Answers to your questions:

  • Where is it failing? If I knew the answer to this, then, of course, I wouldn't need to post here.
  • Does the filament make it to the extruder? Well, it seems to, As I did mention in my posting, the filament goes up and down and up and down, loading and unloading. I have no idea what it is doing in the non-visible areas./
  • Does it enter the gears? I have no idea of even how to determine the answer to this question. And, there are many gears...clearly it makes it through he gears f the MMU. As for the hot end, I don't know...it is not visible and Prusa hasn't seen fit to provide any indicator or LCD message.
  • Does the IR sensor recognize it? I have no idea of what the IR sensor can see. Had Prusa provided an LED indicator, or a message on the LCD screen I could then know, but who knows what the IR sensor is able to see?
  • Does it extrude at all? This is a vague question. If you are asking if the printer can print, why, yes, it generally does a good job of it. But when changing filaments it does what I am complaining about, with the new filament end repeatedly traveling up and down the Bowden-like tube between the hot end and the MMU2s.
  • Is the MMU idler chewing a notch in the filament? No. If it had done that the end of the filament wouldn't be able to travel repeatedly up and down the bowden-likle tube, right?
  • Try removing the PTFE from the extruder before it starts to load as this will help answer many of those questions. Which of the many PFTE tubes are you referring, and how would removing it answer any of your questions? If you are referring to the Bowden-likle tube that the filament end is repeatedly traveling up and down, how would this answer questions like whether the filament is in the gears or what the IR sensor is detecting?
  • Pull out extra length of filament so you have slack to work with. What do you mean by this? Pull it out of where? What slack would I work with? I really haven't the lightest idea of what you are suggesting here. Once again, the issue is that the nice, cleanly, properly cut end of the filament is traveling up and down the Bowden-like tube between the MMU2 and the hot end. For some reason it just cycles doing this until it gives up.
  • Get a good look at the tip to make sure it is formed correctly. I have on countless times removed the filament and re-formed the top to be the proper angle cut, etc. I have tried square, angled, and even chamfer cuts.

Let me again repeat what the symptom is:

When changing filaments, the new filament passes down through the Bowden-like tube and to the extruder. Then, it reverses and passes all there way back to the MMU2, and then this will repeat, with it making a number of down and up cycles. The MMU then ends by fast flashing an orange LED and the LCD display providing the detailed, helpful, informative, detailed description of the problem as "MMU needs user attention.":

Posted : 13/10/2020 8:38 pm
gnat
 gnat
(@gnat)
Prominent Member
Posted by: @seattledavid
  • Where is it failing? If I knew the answer to this, then, of course, I wouldn't need to post here.

We'd like to help, but you have to help us. A few of your responses here are rather snarky and while I totally get being frustrated with the MMU (I completely removed mine this weekend after a month of erratic behavior after about 8 months of flawless operation), you need to calm down a bit and work with us.

In this case (as expressed more specifically in my further questions) what I meant is where in the process it is failing. Clearly the source of issues are different if (for example) the filament never makes it to the extruder than if it does make it to the extruder.

  • Does the filament make it to the extruder? Well, it seems to, As I did mention in my posting, the filament goes up and down and up and down, loading and unloading. I have no idea what it is doing in the non-visible areas./

"Seems to" isn't good enough when we are trying to determine what the issue is so we can try to figure out it's root cause. This is where removing the PTFE from the extruder helps. During the load process, if the filament does not come out the bottom of the tube, then the issue is clearly upstream. If it does come out, then we can most likely rule out an issue with the MMU itself.

  • Does it enter the gears? I have no idea of even how to determine the answer to this question. And, there are many gears...clearly it makes it through he gears f the MMU. As for the hot end, I don't know...it is not visible and Prusa hasn't seen fit to provide any indicator or LCD message.

There is only one place in the filament path that matches the plural "gears". These are the Bondtech gears in the extruder. The parts you are thinking in the MMU itself are dubbed as pulleys in the assembly instructions.

Knowing if the filament is making it into the gears is important as that it is the first step in triggering the IR sensor. You can tell if the filament has entered the gears by watching the idler door. It will move a bit once filament is inserted. 

  • Does the IR sensor recognize it? I have no idea of what the IR sensor can see. Had Prusa provided an LED indicator, or a message on the LCD screen I could then know, but who knows what the IR sensor is able to see?

Yes this was a terrible oversight by Prusa in my opinion. You can go into the Support menu and scroll down to "Sensor info" to see the status of all 3 sensors (PINDA, FINDA, and IR), but there are two major flaws. The first is that you can't get to that menu once you have the "MMU needs attention" message and secondly the refresh is too slow to show fluctuations that may be causing you grief.

There is a mod to add an LED to the IR sensor. It is probably the most useful mod until Prusa adds it themselves.

  • Does it extrude at all? This is a vague question. If you are asking if the printer can print, why, yes, it generally does a good job of it. But when changing filaments it does what I am complaining about, with the new filament end repeatedly traveling up and down the Bowden-like tube between the hot end and the MMU2s.

Does it extrude plastic during the tool change. A common incarnation of IR issues is that it will just keep loading filament and it starts coming out the nozzle. This most commonly happens at the beginning of the print at the initial load before starting the purge line. In that case it will just sit there and create a blob under the nozzle.

  • Is the MMU idler chewing a notch in the filament? No. If it had done that the end of the filament wouldn't be able to travel repeatedly up and down the bowden-likle tube, right?

Generally speaking yes, but I have seen cases where the notch got chewed during the prior load of the filament just as the extruder caught it. Filament was used during the print so the distance between the notch and filament tip no longer matched the distance from MMU pulley to extruder gears. The MMU managed to unload the filament OK, but on the next usage it would only load up to the notch. Typically it would then get stuck and make the issue obvious, but I have seen cases where it couldn't load notched filament, but it did manage to unload it.

  • Try removing the PTFE from the extruder before it starts to load as this will help answer many of those questions. Which of the many PFTE tubes are you referring, and how would removing it answer any of your questions? If you are referring to the Bowden-likle tube that the filament end is repeatedly traveling up and down, how would this answer questions like whether the filament is in the gears or what the IR sensor is detecting?

The tube from the MMU to the extruder. Only remove the end at the extruder.

If filament comes out of the tube during the load, we know the MMU is doing it's job.

By pulling enough extra filament through you then have room to manually insert the filament into the extruder so you can see how it is behaving better.

If you insert the filament into the gears, but the MMU continues to push filament out of the tube that is a good indicator that something is amiss with the IR sensor.

How the extruder behaves is important as well (are the gears running backwards and thus trying to reject the filament? Does the extruder take the filament, but then it stops moving as soon as it's past the gears?). If the extruder does take the filament, an idea of how far the filament is taken before being rejected is important too (e.g. it tells us where in the hotend it reached).

MMU tips and troubleshooting...
Posted : 13/10/2020 9:37 pm
SeattleDavid
(@seattledavid)
Estimable Member

@gnat

A few of your responses here are rather snarky

Many of your questions are vague or require a level of ESP beyond my ability.

Your comment "Seems to" isn't good enough when we are trying to determine what the issue is so we can try to figure out it's root cause." is about as precise of an answer that I can give. It is akin to "I don't precisely know, but probably, but I can't know for certain what you are asking because your question is vague and imprecise.

You ask if the filament is making it to the extruder. I answered "it seems so". Now we get into lawyer-land on exactly what you mean by the extruder. If you are asking if it gets to the extruder assembly then it appears to. If you mean through the feeder gears I can't see and don't know. If you mean to the hot end, I also can't see. If you mean to the nozzle then possibly, but again I cannot see. However, the nozzle isn't blocked, it works just fine when it does feed. So somewhere beyond the Bowden-like tube the the end of the nozzle is as far as it gets. Hence, my vague but accurate to your equally vague question.

Let me clarify what I mean bu the filament going down and up repeatedly. The end of the filament travels visibly from the top of the Bowden-like tube to the bottom of the Bowden-like tube. Repeatedly. Over and over again. From top to bottom and then back to the top. I can see through the tube and that is why I describe it this way.

What goes on within the hidden mechanisms is unknown as there are no inspection ports, there is no visibility, there are no indicator lights, there are no specific messages on the LCD screen.

There is only one place in the filament path that matches the plural "gears". These are the Bondtech gears in the extruder. The parts you are thinking in the MMU itself are dubbed as pulleys in the assembly instructions.

A lawyerly reply. Forgive me for not being a lawyer and for not knowing the precise but arbitrary language that is buried in the manual.

Yes this was a terrible oversight by Prusa in my opinion. You can go into the Support menu and scroll down to "Sensor info" to see the status of all 3 sensors (PINDA, FINDA, and IR), but there are two major flaws. The first is that you can't get to that menu once you have the "MMU needs attention" message and secondly the refresh is too slow to show fluctuations that may be causing you grief.

Thank you for explaining the feature that would answer the question about the IR sensor if only it could be used while the problem was happening. It is good to know that the answer to your question is not obtainable at the time that it is needed.

Does it extrude plastic during the tool change. A common incarnation of IR issues is that it will just keep loading filament and it starts coming out the nozzle. This most commonly happens at the beginning of the print at the initial load before starting the purge line. In that case it will just sit there and create a blob under the nozzle.

That does not seem to be happening. The filament just goes down and up the Bowden-like tube repeatedly, perhaps 4 or 5 times before getting the "there is a problem" error message. Nothing is coming out of the nozzle. I have every reason to believe that the nozzle is not blocked because I can abort everything and start over and the nozzle works perfectly.

Generally speaking yes, but I have seen cases where the notch got chewed during the prior load of the filament just as the extruder caught it. Filament was used during the print so the distance between the notch and filament tip no longer matched the distance from MMU pulley to extruder gears. The MMU managed to unload the filament OK, but on the next usage it would only load up to the notch. Typically it would then get stuck and make the issue obvious, but I have seen cases where it couldn't load notched filament, but it did manage to unload it.

This would be the general case, then.

The tube from the MMU to the extruder. Only remove the end at the extruder. If filament comes out of the tube during the load, we know the MMU is doing it's job. By pulling enough extra filament through you then have room to manually insert the filament into the extruder so you can see how it is behaving better.

Your question implies a static situation. There is no opportunity to pull any extra length because as soon as the filament is all the way down the tube it then starts reeling it back up. The filament will travel up and down and up and down without pausing, without disengaging, without giving me any opportunity to do what you suggest regarding "extra" filament.

Perhaps I am not clear when I say that the filament travels down the Bowden-like tube, then it travels back up, and then it travels down and then back up and it does this perhaps 4 or many more times.

I am not sure what you mean by the "MMU doing its job". But if you define its job as to select the correct filament and send it down the tube then it is doing that...and then is pulling it back again. I wouldn't expect that pulling it back again is "doing its job", but you could argue that when it once again pushes the filament back down the time it is again "doing its job."

What is happening is that the filament is traveling the complete, full length of the Bowden-like tube, over and over, in succession.

How the extruder behaves is important as well (are the gears running backwards and thus trying to reject the filament? Does the extruder take the filament, but then it stops moving as soon as it's past the gears?). If the extruder does take the filament, an idea of how far the filament is taken before being rejected is important too (e.g. it tells us where in the hotend it reached).

Yes, when the MMU starts reeling the filament black the extruder gears run backwards. Again, I have no visibility of the filament within the extruder because well, Prusa didn't feel it necessary to provide any inspection ports.

It does seem that the i3 is deciding to "reject" the filament (to use your term.) I cannot determine how far into the extruder the filament travels before the rejection happens.

 

Posted : 14/10/2020 1:50 am
gnat
 gnat
(@gnat)
Prominent Member
Posted by: @seattledavid

@gnat

A few of your responses here are rather snarky

Many of your questions are vague or require a level of ESP beyond my ability.

They are basic questions about how the MMU functions and is currently operating for you at the present time.

Your comment "Seems to" isn't good enough when we are trying to determine what the issue is so we can try to figure out it's root cause." is about as precise of an answer that I can give. It is akin to "I don't precisely know, but probably, but I can't know for certain what you are asking because your question is vague and imprecise.

Which is why I gave you all the details on how to more specifically determine what it is doing so that we can figure out what the issue is.

A lawyerly reply. Forgive me for not being a lawyer and for not knowing the precise but arbitrary language that is buried in the manual.

It is a basic understanding of the machine. Having the knowledge of what parts are what and how the system functions is critical to being able to use the MMU effectively. Unfortunately it is not at the plug and play stage and requires an ability to tinker and triage.

Thank you for explaining the feature that would answer the question about the IR sensor if only it could be used while the problem was happening. It is good to know that the answer to your question is not obtainable at the time that it is needed.

There are other ways to get at least some of the information, which I described.

That does not seem to be happening. The filament just goes down and up the Bowden-like tube repeatedly, perhaps 4 or 5 times before getting the "there is a problem" error message. Nothing is coming out of the nozzle. I have every reason to believe that the nozzle is not blocked because I can abort everything and start over and the nozzle works perfectly.

This is useful information, but we still need more information.

Your question implies a static situation. There is no opportunity to pull any extra length because as soon as the filament is all the way down the tube it then starts reeling it back up. The filament will travel up and down and up and down without pausing, without disengaging, without giving me any opportunity to do what you suggest regarding "extra" filament.

You have removed the PTFE from the extruder and witnessed this? Or you are making an assumption about it's actions? 

If it really is not coming out of the PTFE or only a little bit is before retracting, that is important information. Under normal operation it (assuming your PTFE is the proper length) should eject quite a bit of filament out of the end before retracting. Its been a long time since I looked at the source code, but it should continue 100-200mm (maybe even more) beyond the distance it expects to find the IR sensor. So if your PTFE is the normal length, then you should get more than enough extra filament out the end of the PTFE. Additionally by the time it gets to the end of the tube it should be doing "short" moves and in between moves the MMU idler should be releasing the filament (it is expecting the extruder to grab it). During that time you should be able to pull the filament through.

Unfortunately at this point I do not think I can help you further. I wish you luck and hope you find a solution.

 

 

MMU tips and troubleshooting...
Posted : 14/10/2020 2:29 am
mm-1
 mm-1
(@mm-1)
New Member

@seattledavid

I have the exact same problem as you. I have read your response to @gnat, i cannot for the life of me believe how rude and ungrateful you are. He is going out of his way to help you find a solution to YOUR problem, and you are messing around on semantics. Grow up, buddy, and recognize when people are actually trying to help you. I found @gnat posting to be of lots of help and while i respect it did not, for whatever reason, help you, you can at at least acknowledge the fact people here are trying to assist you, free of charge. 

Posted : 18/10/2020 12:58 pm
gnat liked
snaslund
(@snaslund)
Active Member

First of all no one on this forum owes you answers.  We didnt sell you anything.  If you have a problem, contact Prusa support.  They are the ones that owe you answers.  Secondly I will help you out and tell you that it sounds like your IR sensor at the extruder is not detecting filament presence.  I would recommend doing the IR sensor calibration as discussed in the MMU handbook.

Posted : 19/10/2020 6:28 am
gnat liked
SeattleDavid
(@seattledavid)
Estimable Member
  1. Nobody owes me any answers, you are correct. We are all free to "avert our eyes".
  2. I will try the IR calibration process. This seems like a reasonable step.
  3. When I was young, I made the decision to stop being bullied. Bullying happens in many ways, most typically by badgering or asking questions which are seemingly helpful but really are a form of attack because the questions are not formulated to actually be useful. I did feel this was happening, which is why I responded point by point. When I was called "snarky" I thought it was appropriate to point out that I was responding in kind. I tend to be quite polite, but when pushed I do push back.

Thank you fore the IR calibration suggestion.

 

Posted : 19/10/2020 10:16 am
Share:

Please Login or Register