MMU2 idler is not turning properly

I’m trying to troubleshoot a MMU2 I’ve just installed on my MK3 printer (with the B7 extruder upgrade installed).

I can only occasionally get the first filament to load. I have checked the proper spacing of the Finda sensor and ball. The alignment of all 5 pulleys appears to be correct. I’ve adjusted the tensioning system.

In examining the behavior of the system during filament loads, I’m seeing a couple of potential problems. The most significant one is that the idler is not turning beyond the point of the bearing for the first filament being properly positioned. I don’t believe that the Extruder motor of the MMU is even attempting to turn the idler to follow the selector lead screw as it moves across the 5 filaments. Though I can easily move the idler by hand when the idler unit is opened up, the Extruder motor is not resulting in the idler moving properly. It appears that the Extruder motor is not moving the idler to line up the bearings of the 2-5 filaments.

The other issue I’m having is that even the first filament doesn’t often load. It appears that the filament being loaded is bending inward and thus missing the entrance to the acceptance hole of the pulley body.

I’m at this point at a loss to further troubleshoot either of these issues.

Thanks for any help.

  • Bob
Answer this question I have this problem too

Is this a good question?

Score 9

Comments:

I've uploaded 2 150MB videos that show characterizations of both MMU2 issues referred to in this post. The first video shows how the idler spindle never rotates toward F5 further than about halfway between F1 and F2. The second video shows how filaments are getting stuck either as they enter the holes of the pulley unit or just after they've entered the hole.

https://gofile.me/6xxw4/NNwXfxLV5

https://gofile.me/6xxw4/pV13QJjui

But the real question that I have… Does the MMU2 that’s shipped 2 weeks ago ACTUALLY work reliably? Based on the posts and YouTube videos that I’m seeing, it doesn’t. Even those who participated in the beta are saying that many of them have yet to complete even a single multi-material job that is 100% error free.

Is there anybody who can speak up to say that they are completing at least 50% of their MMU jobs successfully?

Thanks for any help.

- Bob

by

I have the same exact problem. Looking forward to finding help along with you.

by

Add a comment

4 Answers

Most Helpful Answer

I had the same problem and finally figured it out yesterday. For me it turned out that the idler barrel itself was rubbing on the side of the pulley body (away from the motor that drives the idler). You can sort of tell this is happening if you try to turn the idler by hand (turn off the printer first!) and if it’s “grabby” or hard to turn then you might have this issue.

NOTE: For the process described below, I didn’t bother removing the MMU2 unit from the printer, but it may be useful to carefully cut off the zip ties holding the wires in place. This allows the top of the MMU2 to open a little further without stressing the wires or their connections. If you do this, follow the instructions in the assembly manual to replace the zip ties correctly.

The solution for me:

  1. Open the MMU2 unit (loosen the 2 screws on top with the springs).
  2. Rotate the Idler until the 2 screws that hold it to the motor are accessible. (They are on the left side by the motor.)
  3. Loosen those 2 screws almost all the way (easiest to leave them in their holes at least a little).
  4. Push on the Idler from the right side toward the motor where you just loosened the screws.
  5. Mine shifted slightly to the left (it took a bit of pressure before this happened).
  6. Re-tighten the screws that mount the Idler to the motor. (Be very careful here and do NOT over tighten these! I cracked the plastic of my Idler where it is thin above the square nuts and had to print a new idler.)
  7. Make sure the Idler can rotate easily by hand, and that the motion is smooth. (It should be much better than before thanks to no longer rubbing on the right side of the pulley body.)
  8. Close the MMU2 unit and re-tighten the screws with the springs.

This repair actually also fixed my issue where filament wasn’t making it all the way to the E-Axis. (The idler was almost never in the right place, and it wasn’t providing the needed grip on the filament to move it all the way down the orange PTFE tube.)

———

Upon further consideration, it's actually extremely likely that this happens to people due to the insertion of the Idler’s motor shaft pushing the barrel in that direction as you insert it. It seems to me that an extra step in the assembly guide that tells people to push the Idler back to the left toward the motor prior to tightening the screws that hold it in place would be an excellent addition that would almost entirely prevent this problem. The addition of a thin metal bushing of some sort on the right side of the Idler would probably also work to keep the two plastic parts from rubbing on each other.

Was this answer helpful?

Score 5

Comments:

I was having a similar issue, and this seems to have helped for me. Filament finally loaded! Trying my first print now... fingers crossed.

by

This seems to have fixed my problem as well. I used the middle of a old zip tie for a spacer between the idler bearing and the idler body. The idler can now be easily turned, and I'm now trying my first print.

by

Thanks a ton for this! I could tell it wasn't turning properly but couldn't figure out why. This resolved it. Mine would not slide by hand, so I took a bar clamp with the squeeze handle and used that to slide it toward the motor. Now it moves much more smoothly. Really appreciate the good explanation!

by

Add a comment

Hi Bob,


Can you first check if the motor is working properly by connecting it to one of the headers on the RAMBo board?

For an example, connect it to the X motor slot, then move X axis from the LCD.

Can you also check what happens if you loosen the tension on the springs? Perhaps it’s overtightened and doesn’t let the idler to rotate properly.

I was trying to download the videos, but I’m unable to. Please consider contacting us via email: info@prusa3d.com - you will be able to send the videos there if they’re not too big.

Was this answer helpful?

Score 1

Comments:

Filip, I have the exact same problem as Bob described above. The idler motor is definitely getting power as when it wants to rotate it moves like it wants to rotate but it doesn't. While the selector moves to F2, the idler bearing is still hovering over F1. I can continue to cycle through the filaments but the idler bearing stays over F1, while the selector is on F5. I have tried the factory reset but have only been plagued with the same results.

I'll be following this post closely as I have a few projects stacking up that were intended to use this "upgrade".

Thank you

by

Ryan, have you tried reflashing the firmware on both MK3 and the MMU?

by

Filip, I've flash firmware twice on both units with the same results. The top extruder motor that rotates with the bearings is warm to the touch and makes a noise like it wants to rotate with the selector but does not rotate. While the selector moves to F2-F5, the extruder motor with bearings stays at F1. I've also taken the MMU apart to expose this part and completely rebuilt it with the same plagued results.

by

My idler motor also gets warm. I checked, and the idler bearing assembly spins very freely for me, when the device is open, and all bearings line up. See my comment lower, I think the spring screws have to be tightened too tight to make up for the obvious 1mm? gap between the filament and the teethed rollers (pics below other comment)

by

Add a comment

I too have this problem. 1st and 2nd length calibrations worked just fine. Then I tried to print something 10 x, only once did 2nd filament load correctly. Now length calibrations don’t even work anymore. I think the parts settled/and or moved. Look at the daylight between the filament and the MMU2 feed, this can’t be right, they should be almost touching the filament?! A patch might be bigger rollers with teeth? My suspicion is we’re having to crank down so hard to bend the filament that the roller selection can’t work properly?

Block Image

Block Image

Was this answer helpful?

Score 0

Comments:

Talked to support, noticed there was slop in the holes for the pulley motor, if one holds the motor up slightly firmly, then looses and retightens the pully motor screws, this problem goes away, also the optimum spring tension is 1mm? above the plastic for me? (not too sure)

by

Add a comment

I’m still having some issues related to this failure mode, despite having tried everything I can find. My issue looks like this:

The print will run, changing tools reliably, and then, seemingly randomly, the filament won’t reach the Extruder. It will then happen again shortly after and eventually the MMU will either fail to retract the filament from the Extruder or fail to extrude it past the FINDA at all. Upon investigating the MMU at this point and trying to load the filament using the pushbutton interface, I find the bearings to be inappropriately aligned with the drive gears (pulleys). I then find that I can rotate the idler by hand so that that he bearing is positioned to move the filament forward or backward down the tube. The print can the proceed but this failure mode repeats frequently.

It strikes me as odd that I would be able to rotate the idler freely. Shouldn’t it be energized at its position no matter what’s happening with the print? Could there be a problem with the MMU idler motor or its driver?

Things I’ve tried:

  • I’ve made sure there’s no interference with the idler rotation. It’s good.
  • I’ve tried multiple MMU spring tensions from very light to very torqued down. I see the same failure mode, regardless (excepting when the springs are so tight the idler can’t rotate).
  • I’ve make sure there’s no friction in the filament path, or eliminated as much as possible.

Any ideas? For me, the error only shows up every 50 tool changes or so. I’m so close to having this thing work properly. My next step is to run a test print (5 colors) with the MMU open and testing to make sure the idler motor is energized during each swap to see if it “lets go” at some point when it’s under no strain. Any ideas or feedback would be very appreciated. Much obliged, ya’ll.

Was this answer helpful?

Score 0

Comments:

Hey there! I'm not from the Prusa Support team but may be able to help.

You're almost there. A couple things are still very important that I highly recommend you try.

First of all, check your filament path, not for friction, but for bits of filament. Usually, at least for my MMU2, friction in the tubes isn't as significant as one may think. Strings and little ground bits of filament can sometimes get stuck in various places from under the FINDA (causing false positives for loading filament, thus leaving layer skips), under the pulley gears inside the MMU itself (causing friction and once knocked one of the pulleys out of place from on the shaft and loosely twirled for the rest of the print, causing one color to be nonexistent), or on any threaded rod that would cause resistance as the selector unit moves across it.

Sincce this comment is actually super long, I will try to divide it into 3 parts because of the word limit.

Mathieu

by

Another thing would be to adjust the spring tension on the extruder itself and not the idler. However, the best tension for the springs on the idler would be juuuust below the same level as the 3D-printed frame, so it's barely not flush. Perhaps about 0.4mm or so. As for the extruder, perhaps try tightening the springs a bit. It has been said that a lot of the time super tight extruder gears cause inconsistent extrusion, but if you do it just a little bit (mine is so when filament is between the Bondtech gears the screw heads are basically flush with the extruder body 3D-printed part, maybe just 0.5mm or so sticking out of it. This is so when the filament is being retracted/cooled, it is shaped in a way so that the little blob at the end isn't as wide and thus won't cause as much friction in the orange tube (it's orange in my kit) from the extruder to the idler. If this doesn't work, try tweaking settings by increasing the cooling moves under "Filament Settings" -> Advanced.

Mathieu

by

The last thing, and probably the most important, is that you can adjust the bearing alignment in the idler! The idler-barrel part is secured to the shaft by 2 small screws on one end. Try loosening them a lot and shifting the barrel around, until rather the bearings appear to line up with the pulley gears on the shaft below, or so that neither end of the barrel is rubbing against any other thing.

In hindsight, this was quite long-winded, but I think this wll be enough to help you a little. It took me a good 2 weeks to get my first successful multi-color print, and another week and a half before that to get a single-color one. There is a LOT of tweaking involved.

I hope this helps, and good luck!

Mathieu

by

Hi Mathieu,

Thanks for the reply. I've checked all the sources of error you outlined and can confirm that I'm good on those fronts.

As an update: I ran a test print with the MMU open and manually feeding filament back and forth through the FINDA at each swap to see if the idler motor remained energized and held position over time. It did. I then closed the MMU and ran a real test print that stalled out about 100 filament changes into the print with the filament about 4 inches out of the MMU.

I'm not totally sure what happened because I didn't catch it in the act, But when I use the MMU push buttons to try and feed filament and restart the swap, nothing happens. Neither the idler nor the drive motor do anything. The idler is also not energized and I can rotate it freely. My hunch is that this is the cause of the problem: the idler motor de-energized at some point and wasn't able to hold the bearing in the correct position.

I'm restarting the print now and I'm going to try and catch it in the act.

by

Add a comment

Add your answer

Bob will be eternally grateful.
View Statistics:

Past 24 Hours: 5

Past 7 Days: 70

Past 30 Days: 265

All Time: 1,051