MMU2S wont load Filament at all  

  RSS
Stonegate
(@stonegate)
Active Member

Hi,

i built the MMU2S and controlled all possible steps where i could have made an error. But i am still not getting my MMU2S to load any of the 5 Filament slots.

When i look inside with a small flashlight, i can see, that the pulley gears are rotating. They are all perfectly alligned to the middle of the filament flow. I did this with an allen key as well with a straight part of 1,75mm filament. They are correct. And the small screws are secured tight enough.

But i can see the pulleys rotating and rotating and rotating but the filament is not travelling a micron.. the bearing above (which sits in the idler) is pushing against the filament from above. However the pulley obviously does not get a grip on it or does not have the strength to pull it in.

I tried it with loosing all kinds of screws (the 2 with the springs closing the MMU chassis, the 4 fitting the small PTFE tubes in the MMU, even all screws in the spool buffer where the PTFE tubes are secured). Nothing helped. The filament can glide through all PTFE tubes. There is some force needed to push it but its minimal and its the same amount of force you need when you push the filament trough a simple PTFE tube that is just lying around loose.

So i have absolutely no clue why its not getting grip on the filament.

Does any of you have an idea ?

KR

Stonegate

This topic was modified 1 year ago by Stonegate
Posted : 30/08/2019 9:42 pm
gnat
 gnat
(@gnat)
Prominent Member

Either the idler door is not tight enough to cause the bearing to compress the filament to the drive gear or the idler is not being moved into the correct position. Those are really the only to reasons for it to fail to load as described.

Test setup:

  1. Remove the long PTFE tubes if they are installed in the back of the MMU and install the short tubes (the inlets for the buffer) instead.
  2. Prep 2 filaments for loading.

Test:

  1. Start the loading process (tool doesn't matter).
  2. Insert filament into the correct tube and push it in until it reaches the gears.
  3. If it does not grab the filament from you, push down on the idler door to add tension.

If it took the filament from you at step 2, there is too much friction in your path to the MMU. Make the filament move more smoothly.

If it takes the filament after pressing on the idler door, tighten the screws for the door and repeat the process until it takes the filament without help.

Be aware that it could be a combination of both of those issues.

If friction and tension aren't the issue then more likely than not the idler isn't moving to the correct position, check it's installation and position.

MMU tips and troubleshooting...
Posted : 30/08/2019 10:15 pm
Stonegate
(@stonegate)
Active Member

Hi there,

thanks for your time and hints.

When you talk about the "idler door", you refer to the idler body that sits over the pulley body i think.

Its the part with the two M3x40 (?) screws with springs that close the whole MMU and keep it together. The screws with the springs shouldnt be to tight. The manual says you should screw them in only a little below the surface of the screw holes. However that seems absolutely tooooo slight. With that force on the screws the pulleys were not pulling any filament at all.

 

Then i followed your advice and tightened the screws. It helped a little. Filament 1 was loading now. I also removed the whole spool buffer part as it takes too much space in my little setup anyway and seems to be not giving more benefit than it uses space. I will probably print self retracting spool holders ( i saw some that seem to do this job well enough ).

But my problem is still there.

1) With maximum tightened screws the idler could not rotate into every position well and was blocking.

2) With looser screews filament 1-2 loaded. 3-5 did not.

3) Sometimes filament loads in some slot and it does not find its way into the correct hole to the extruder. So filament runs out of the upper side of mmu infinite until you stop it, or it tangles around the idler.

4) Sometimes filament loads in some slot but when the idler moves to the other slots, the blade pulls the filament string all the way to the right side causing a nice jam. I found no way to prevent that. (As mentioned, the blade has no strenght to cut the thick filament at all).

 

By experimenting i got 2 new problems now:

A) The blade cut into the plastik part "front PTFE holder" and got stuck in there, tearing the whole blade out of its blade holder part. I had to remove the stuck blade with needle nose pliers. I dont know for what the blade is really needed. It has absolutely no strength or sharpness to cut 1,75 mm thick filament strings and it does not sit tight enough for that purpose. Probably its just there to cut the fine strings from retracting the filament ? The manual does not mention for what it is. At least i could not find any word on it.

B) After screwing and unscrewing for 2 hours, the nuts in the idler body, that are locking the M3x40 (?) spring screws, became loose inside the plastik part of the idler body. They rotate with the screws and i found no way to hold them so i can unscrew and open my MMU again. I tried with some sizes of the allen keys, i tried it with a flat small screwdriver. The holes are too small and the motors are around them so i cannot try with needle nose pliers.

 

So i got the problem now that i cannot adjust the MMU anymore because of the locked screws. Also my idler is overtightened and cannot rotate in positions 2-5. I´m lucky.. after 1 Hour i managed to load filament into Slot 1 and i have original prusament jet black PETG here. I made calibrations and after another hour i somehow managed to get a constant filament feed in slot 1. Since the idler is stuck in that position and the blade is also missing, i consider my MMU to be non-functional. I managed to start printing somehow with that corrupted setup and i´m printing a brand new idler body and PTFE and blade holder now. The print looks very good so far and after 10 hours (2.5 more to go).

I still dont know why my idler can grap filament 1-2 but not 3-5 (even when the screws were optimal tightened - remember.. i tried everything, every force and every screw position for like 3 hours).

I will probably also print the idler and the pulley body again. Maybe original parts have some problem.

What i find curious is that the original part of the idler body has a small "M1" on it. The downloaded and new file from the prusa side prints a part that says "R1" on it. However its the exact same and correct file i picked "Original Prusa i3 MK3/S or MK2.5/S  MMU2S upgrade:"

Why do they differ? i dont know.

Also a few thoughts of mine needs to be dropped here: 

- I think the quality of the MMU2 differs from the design and quality of the Printer itself. It seems as it would have been designed by other people

- The handbook seems to be not as informative as the printeres assembly handbook. In the german version there is a whole chapter of assembly missing, the english handbook does not tell you what the blades are for, or i could not find something about it (maybe my fault - then please take my apologize), i couldnt find what the 3 Buttons on the MMU do, also the pictures and "loading filament" section is veeeery thin. For me there are too few information on how to correctly load the filamanet into the spool holder. I had to guess a lot and watch videos on youtube.

 

- Loading the filament into the spool buffer feels so wrong at all. You have this little tiny corner where you can TRY to grab the tangling filament and push it into the MMU-feeding side PTFE tube. And then you need to push it all the way down. Loading a filament takes 1-5 Minutes (considering that you have to move everything and find an angle where you can grab everything well enough. As i said, i have limited space and this construction is using way too much space. I cannot place 5 spools and the buffer next or behind the printer. Why cant you create a spool buffer that can be placed vertically behind or left or right sided to the printer? The spools should be under the table of the printer if you ask me, or in a shelf above it.

I know thats alot of text. Sorry for that. But alot of thoughts and questions i couldnt find an answer for by myself.

Have a great weekend.

Stonegate

This post was modified 1 year ago 2 times by Stonegate
Posted : 31/08/2019 12:31 pm
gnat
 gnat
(@gnat)
Prominent Member
Posted by: Stonegate

Hi there,

thanks for your time and hints.

When you talk about the "idler door", you refer to the idler body that sits over the pulley body i think.

Yes

Its the part with the two M3x40 (?) screws with springs that close the whole MMU and keep it together. The screws with the springs shouldnt be to tight. The manual says you should screw them in only a little below the surface of the screw holes. However that seems absolutely tooooo slight. With that force on the screws the pulleys were not pulling any filament at all.

Much like the door on the extruder, it's a judgement call and you have to adjust to suit your system. The recommendation in the manual is just a starting point.

But my problem is still there.

1) With maximum tightened screws the idler could not rotate into every position well and was blocking.

As you've found, that was way too tight. There are clearly other issues going on. I think you focused too much on the tightness of the door and not enough on my other suggestions.

2) With looser screews filament 1-2 loaded. 3-5 did not.

This is not uncommon. I find I have to set the screw near the end tighter for 4 and 5 to work well while 1 and 2 need the screw on that end to be a bit looser. 3 is always the first to give me trouble and I think we could stand another screw to better balance the tension.

3) Sometimes filament loads in some slot and it does not find its way into the correct hole to the extruder. So filament runs out of the upper side of mmu infinite until you stop it, or it tangles around the idler.

Thats a pretty good sign that your tension is too high.

4) Sometimes filament loads in some slot but when the idler moves to the other slots, the blade pulls the filament string all the way to the right side causing a nice jam. I found no way to prevent that. (As mentioned, the blade has no strenght to cut the thick filament at all).

This is really sounding like your idler is not moving to the correct position.

By experimenting i got 2 new problems now:

A) The blade cut into the plastik part "front PTFE holder" and got stuck in there, tearing the whole blade out of its blade holder part. I had to remove the stuck blade with needle nose pliers. I dont know for what the blade is really needed. It has absolutely no strength or sharpness to cut 1,75 mm thick filament strings and it does not sit tight enough for that purpose. Probably its just there to cut the fine strings from retracting the filament ? The manual does not mention for what it is. At least i could not find any word on it.

The blade is there to cut the tips when it thinks there is a problem, but that feature is currently disabled in the firmware as it doesn't work well. Usual advise to those having issues with the blade is to just remove it.

l dont know why my idler can grap filament 1-2 but not 3-5 (even when the screws were optimal tightened - remember.. i tried everything, every force and every screw position for like 3 hours).

Based on everything you've said I suspect your idler is not functioning as designed. Pay close attention to it when you reassemble the MMU.

I will probably also print the idler and the pulley body again. Maybe original parts have some problem.

What i find curious is that the original part of the idler body has a small "M1" on it. The downloaded and new file from the prusa side prints a part that says "R1" on it. However its the exact same and correct file i picked "Original Prusa i3 MK3/S or MK2.5/S  MMU2S upgrade:"

Why do they differ? i dont know.

M (and C on the printer) indicates that Prusa printed the parts and R indicates they were user printed. Otherwise there is supposed to be no difference.

- I think the quality of the MMU2 differs from the design and quality of the Printer itself. It seems as it would have been designed by other people

Development of the printer is over a decade ahead of the MMU. The early printers were not as good as the MK3 either.

- Loading the filament into the spool buffer feels so wrong at all. You have this little tiny corner where you can TRY to grab the tangling filament and push it into the MMU-feeding side PTFE tube. And then you need to push it all the way down. Loading a filament takes 1-5 Minutes (considering that you have to move everything and find an angle where you can grab everything well enough. As i said, i have limited space and this construction is using way too much space. I cannot place 5 spools and the buffer next or behind the printer. Why cant you create a spool buffer that can be placed vertically behind or left or right sided to the printer? The spools should be under the table of the printer if you ask me, or in a shelf above it.

One of the mods I mentioned greatly improves loading the buffer. Not only does it allow the filament to move easier it also allows you to move the inlets and outlets all to the same side and provides more space to work. 

As far as placement, that was a problem for me as well so I started with rewinders. Now I have a 2020 frame holding the spools above my printer and the buffer hangs vertically off the back. All in all it adds 250mm in height, 50mm in depth, and 100mm in width from what the printer used already.

I still think rewind spools are the best way to get started until you have everything working well. Once you are confident that you've gotten everything sorted out, then look to what you want your permanent install should be and decide what management option is best for you.

MMU tips and troubleshooting...
Posted : 31/08/2019 4:11 pm
Stonegate liked
Stonegate
(@stonegate)
Active Member

Hi Gnat,

thank you so much for all the time you invested into helping me with your posts. I will try all you said. And since i print all important MMU parts again, i hope i can find the issue. I will leave the blade ouf of my setup for now.

Stoney

Posted : 31/08/2019 8:38 pm
Stonegate
(@stonegate)
Active Member

Hi there,

Update for you (and others that might read this):

I´ve re-printed the idler body and pulley body and finda selector and blade holder parts with my normal PEI coated Prusa Steelplate.

The results were more than good. In fact the finda selector turned out to be much more precise due to its very flat surface. Prusa is printing its parts with a PEI coated and slightly rough steel plate. This makes the parts looking more rough and give them a shiny funny look on the side that lies on the bed.

HOWEVER this seems to be a problem for the finda selector as the area where you have to slide the small blade in is not turning out precise enough to hold the blade well. At least this was the problem with my original prusa part. With my self printed part the blade sits in the small hollow much better. 

Talking about the other parts, i could not see a big difference. However the parts fitted together so much better than the original Prusa printed parts.

The result is 100 % better. Now my MMU works "almost" like a charm. I could easily load all 5 filaments.

Since i ruined the idler door screws AND the springs inside, i could not use 2 springs for the new printed parts. I found ONE new spring in one of my replacement bags from PRUSA. I ordered new springs already. Could the one (right side of the MMU) missing spring cause the problems mentioned below? 

I still have some issues left:

1) When turning on the printer the Finda Selector unit goes to the right and the idler is giving a "rattling" (?) sound like it cannot move to a position but wants to. However i built everything as the manual said. This sound sometimes also happens when filament changes are happening during a print.

2) Talking about prints: I could print a five material test gcode. It has like 100 filament changes and its duration is 3,5 hours with 5 filaments.

I was mixing PETG and PLA (1x Prusament PETG Jet Black and 4x Amazon Basics PLA) together.

The print seemed to work well until after round about 15 Minutes. After like 10 filament changes and a few layers on the bed, it unloaded the black PETG in Slot 1 and was trying to load the white PLA in Slot 2. However this failed and the MMU2 LED flashed red after a while and the LCD said "MMU requires user interaction" (or something like that)

I noticed that the tip of the white filament was a little bulgy. Cannot say if the others looked the same. However it was a little thick and probably could not run through the opening holes of the mmu2 fine. I guess this was the reason for the failing of the print.

Additional to the above problems, i have some questions:

A) When it says "mmu requires user interaction (?)", it also says "fix the problem and press the button".. WHAT button is the printer talking about ? Pressing the round Printer LCD attached Button does not do anything. And the MMU2 has 3 Buttons. Which one should i press ? i tried all and nothing happened at all. I had to turn off the printer. The print was lost of course.

 

B) The purge or nozzle cleaning process when the MMU is emptying the nozzle to the purge block before it unloads the filament, seems to be veeeery slow. Is this normal ? Why does it take so long ?

 

C) What is this cluttering sound in my idler ? is this going to damage something inside ?

D) Why is the filament retraction working fine for one filament and not for another ?

KR

Stoney

Posted : 02/09/2019 2:52 pm
Dave Avery
(@dave-avery)
Honorable Member

the idler is homing by driving to the stop, the rattling is normal

the slow retraction is the end forming process it moves the hot end of the filament back and forth in the heatsink to form a better end before unloading

 

Posted : 02/09/2019 3:09 pm
gnat
 gnat
(@gnat)
Prominent Member
Posted by: Stonegate

Talking about the other parts, i could not see a big difference. However the parts fitted together so much better than the original Prusa printed parts.

I haven't had issues with the MMU parts, but I was not happy with the quality of the MK3S parts. Seems to me that Prusa would want to show off the quality potential of these printers with the parts that they print, but they seem to be more about whacking them out fast rather than good 🙁 

Since i ruined the idler door screws AND the springs inside, i could not use 2 springs for the new printed parts. I found ONE new spring in one of my replacement bags from PRUSA. I ordered new springs already. Could the one (right side of the MMU) missing spring cause the problems mentioned below? 

Yes you will have filament change issues with tools closer to the missing bolt/spring. 

1) When turning on the printer the Finda Selector unit goes to the right and the idler is giving a "rattling" (?) sound like it cannot move to a position but wants to. However i built everything as the manual said. This sound sometimes also happens when filament changes are happening during a print.

I'm surprised that the selector moves at all, but the rest of your description is normal.

I was mixing PETG and PLA (1x Prusament PETG Jet Black and 4x Amazon Basics PLA) together.

Mixing materials has it's own issues. I would stick to just a single material for all tools while you are getting things sorted out.

I noticed that the tip of the white filament was a little bulgy. Cannot say if the others looked the same. However it was a little thick and probably could not run through the opening holes of the mmu2 fine. I guess this was the reason for the failing of the print.

So the issue was that it couldn't load the white? Had it already used the white previously? If it had then the filament tip was not too big, otherwise it would not have retracted. There could be various issues for this, but I wouldn't bother trying to troubleshoot until you get the other tensioner bolt installed. 

A) When it says "mmu requires user interaction (?)", it also says "fix the problem and press the button".. WHAT button is the printer talking about ? Pressing the round Printer LCD attached Button does not do anything. And the MMU2 has 3 Buttons. Which one should i press ? i tried all and nothing happened at all. I had to turn off the printer. The print was lost of course.

It is referring to the buttons on the MMU, but yes the messing is terrible.

Press the middle button to have it attempt to continue the load/unload process. Once it is cooperating, press the right button to continue with the print.

B) The purge or nozzle cleaning process when the MMU is emptying the nozzle to the purge block before it unloads the filament, seems to be veeeery slow. Is this normal ? Why does it take so long ?

Yeap. Not only is the tower a giant waste of plastic, it's a giant time sink too.

Yes being slow is normal, it is just their method for purging and tip forming.

C) What is this cluttering sound in my idler ? is this going to damage something inside ?

You mean the noise you described at start up? Perfectly normal as it homes during startup.

D) Why is the filament retraction working fine for one filament and not for another ?

Lack of tension (missing bolt) is the place to start.

MMU tips and troubleshooting...
Posted : 02/09/2019 4:14 pm
Stonegate
(@stonegate)
Active Member

Hi there,

i bought new springs and everything works fine now for like a day. Suddenly the MMU2 stopped after like 16 hours of printing and could not load the filament anymore. The finda selector did not move anymore and it was just trying to grab the filament for 1 second and then stopped rotating the pulleys.

I tried cleaning the pulleys, checked for stuck filament (there was no dirt or stuck filament at all. No grind. Nothing), i tried to help it load the filament by pushing it further - i needed no strong force and the filament went just fine directly into the selector and into the PTFE tube to the nozzle.

I tried all kind of stuff for like 30 minutes and then gave up and turned off the whole printer. When i turned it on again it could load all filaments just like a charm without a problem.

Thats nasty 🙁 I lost a 16 hours print and dont know what went wrong.

The MMU2S was not reacting to the first and third button. The middle button just made it try to load filament for 1 second. I could not move the selector to any other position (F2,F3,F4,F5) with the buttons.

I did various testprints on friday and all of them worked well. I have no idea why it stopped working magically after 16 hours when it worked well after a reboot. Pulleys were rotating, idler was in the right place, but filament was not grabbed. I even cut 15 cm away and loaded it completely new. It did not feed.

Maybe an MMU2S firmware bug ? I run the latest (think its 1.0.6)

Is there a way to reboot the MMU2S without powering off the printer ? 

If something like this happens again, what can i do to save my print? How can i resume it if the MMU2S is not reacting?

Thanks so much

Stoney

Posted : 07/09/2019 7:44 am
Dave Avery
(@dave-avery)
Honorable Member

you can reset the mmu with the reset button in the hole next to the micro usb port.

 

when it was having problems could you see the dim light in the finda probe changing state on the load attempt?

one common mmu mod is a modified selector that adds a magnet below the filament path to pull down on the steel ball

i'm using one like https://www.thingiverse.com/thing:3199061

Posted : 07/09/2019 8:10 am
Dutchy
(@dutchy)
Eminent Member

Hi,

My MMU2S starts to have problems now and then loading filament.

When pushing the filament into the right tube the drum just doesn't grab the filament and it doesn't come up on the front where the 2 spring bolts are like it used to. It looks like its not in the right position and the filament just passes between the gear and bearing?

Any ideas? Spring bolts already tightend.

Regards,

Posted : 30/10/2020 12:04 pm
andreas
(@andreas-t10)
Eminent Member

@jasper

I actually had a similar problem a couple of weeks ago. The ball bearings didn't grab the filament to pass it through. I had to remove and reinstall the drum. It seems that at some point was misaligned.

Posted : 30/10/2020 6:53 pm
pcweber
(@pcweber)
Estimable Member

What comes to mind with me right now are two things that I had wrestled with.

1. The idler stepper was off 90º, I could tighten the screws to where the idler did pick the filament up. Eventually I found an idler on thingiverse which uses a coupler as opposed to the screws on the idler. No more problems.

2. Cabling, I just recently had reset problems and thought that maybe there was a fractured solder pad. I had blown the connectors with canned air, but this time I used Lubrilimp, a spray cleaner and mild lubricant for electronic connectors, relays, and variable potentiameters. I then paid particular attention to place the zip-ties exactly like the installation manual. No more resets, continuous printing with tool changes that do not fail.

One note, when Prusa updated the selector knife blade in March I printed it. It works really well especially if you have a poorly formed filament end. The printer attempts to load the filament three times. If the filament doesn't load after that the MMU cuts the filament and keeps on running.

Stay safe and healthy, Phil...
Posted : 01/11/2020 5:52 pm
Share:

Please Login or Register