MMU2S loading/unloading getting worse with time
I'm using prusament and started out by printing the included gcodes of the 3 color lizard with no loading issues. Printed the 5 color pliers/gears with no issues. Still using prusament, I attempted a 2 color frog from thingaverse and had to baby sit for an entire 12 hour print as the filament keeps getting stuck in the extruder ptfe tube...and I mean really stuck. Seems that stringing started occurring with the prusament and some buldging on the filament tip. So in the manual, it say to adjust temperature to get the right tip. It doesn't say "if it looks all stringy, adjust temp up or down." It doesn't say "if it looks all ratty, adjust temperature up or down." It just says adjust temperature usually fixes the tip problems.
So I changed filament to matterhackers pro since it has the same .02 specs as the prusament. It has a glossy finish for my dayglo green and black frog. Not sure about matterhackers reputation but I thought I'd try it. Using the prusa slicer (again) with the prusament presets, the frog started out ok at after first 15 or so layers, then the ptfe tube jammimg started. It will jam on load or unload at random. I finally figured out it is due to the tip stringing since the unjamming was pulling hairlike strings out of the ptfe tube. I took a guess and started lowering temps. The preset for prusament is to run at 205. I gradulally lowered to 202 in increments of 1 to see how it effected the stringing. Eventually, I was able to predict the jams because I could see strings hanging out of the hole of the last retracted filament. So eventually, over a 24 hour period for a 12 hr print, I seemed to get the stringing under control as I kept lowering my temp until I got down to 195. Then the string seemed to get under control and the buldging started....again jamming up the tube going down to the extruder. And this is a 2 color print.
Somewhere along the way, I could not manually feed the filament up to the steel ball...it felt like the cleanly cut tip was hitting something. I then realized it was the blade. So partially through my cool looking frog, I just uncrewed the blade holder and removed it. I can't quite understand the point of the blade and why it's so necessary as my strings are so fine, the blade just pushes them out of the way. And also, using positions 1 and 2 for a 2 color print, it seems the blade will do nothing to cut anything if only cycling back and forth to slot 1 and 2. Wouldn't it have to go past slot 2 and come back to cut something? After thinking about it, I thought why can't a blade be set up to slice off a brand new tip each and every time ...but that wouldn't help on an unload since the tip could be buldging any way.
But these bulges aren't that big and the force needed to remove them from the tube sometime requires pliers. And many times what's left in the tube is a string that's even finer than a human hair. Point being, is why does the tolerance of this tube need to be so tight. Seems like there no margin for error whatsoever. And lowering my temp to 195 to fix the stringing is now reducing the quality of my printed part.
So what is happening over time. I had such good success with my first 4 prints. I'm not ruling out my slicer settings but I now have no confidence in this thing to let it print over night. These jams are ONLY in my extruder down tube and I can tell that it has nothing to do with the upstream friction or resistance of the feed tubes from the buffer. This filament is deforming and getting jammed in the down tube and it could be unload or load it doesn't matter. Maybe its just that the frogs are to complex of a pattern? I don't know. The green one is matterhackers pro the red and black is prusament. I will say the red didn't have too many issues until about halfway through the 12 hour print.
Lizards went without a hitch (1st prints). The pliers stopped at 99% (jammed downtube). The frogs were constantly jamming with the green and black being the worst and the last one printed. Seems like it jammed every 20 minutes no matter how I tweaked the temperature and even slowed down the speed to 90% speed at one point. The good news is I'm getting really good at loading filament with this buffer contraption as I'm pulling fresh filament through after each jam 🙂
RE: MMU2S loading/unloading getting worse with time
And now I've even took out the buffer and just running through the feed tubes. So this isn't related to upstream ptfe resistance imo. I'm practically feeding filament in directly with the exact same results. No better, only getting worse. I looked at my sensors on successful loads, the IR sensor is toggling 1 to 0 on filament load/ unload. What I am seeing is that the FINDA is pretty much staying on 1 even after filament is unloaded. I tried tweaking on it to get it to change state. But I can only get it to either stay on 1 or stay on zero. Doesn't seem like it's really doing all that much functionally. Still doesn't explain the filament stringing/bulging in my limited experience with this thing. The prints are coming out GREAT when it loads. But it will only print about 2% of the provided 3 color lizard and then there's that stupid flashing light again with filament stuck in the downtube. I don't get it.
RE: MMU2S loading/unloading getting worse with time
What I am seeing is that the FINDA is pretty much staying on 1 even after filament is unloaded.
You mean after the filament has been unloaded past the FINDA? If you are absolutely sure that there is no stray bits of filament jamming things up and you have the FINDA adjusted properly, what firmware do you have installed for the MMU itself? I believe either the current firmware or the one that is pending release has a fix for some users where the FINDA was not properly resetting it's value. If you have really ruled out physical issues and your firmware is behind you might look into upgrading.
RE: MMU2S loading/unloading getting worse with time
Hey thanks for the feedback on this. Now that you mention hot end ptfe tube, something is beginning to make sense. This machine was bought back in Nov 2018 and shipped to me right away (It was an MK3). I also ordered the MMU2 at the same time. I left the printer in the box until May 2019 when my MMU2 showed up. To Prusa's credit, the "S" upgrade parts for both my printer and the MMU2 were shipped with the MMU2 in May. So during the printer build, I built an MK3S. Tried it out...made sure everything worked, printed out 3 or 4 nice benchy's and basically cleared it for the MMU2S upgrade. I noticed some screws loosened after those benchy prints as I had to take the extruder apart to do the MMU2S.....just mentioning this as it is now common practice for me to check for tighteness after a new build runs a few times. Anyway, I completed the extruder upgrade and the MMU2S build and started out printing the lizard. I was blown away at how good they turned out. I had the buffer wall mounted using the spool trolleys or whatever they are called on a shelf above the buffer. I printed a second lizard...then a third just trying out all the new prusament I ordered. It was then I realized I wanted another Prusa printer and bought an MK3S kit. Now I don't know how else to make this make sense to someone without being all long winded about it...so please bear with me.
Once I finished the new kit, I saw that the hotend ptfe tube on the new printer was down more flush with the black plastic compared to my "successful" MMU2S build. In fact, the ptfe tube on my MMU2S was just touching (ever so slightly) the bondtech gear to the point it slightly wore some of it away as if it was too long or not properly pushed in...whatever. At first I thought I would just live with it since I was getting some awesome lizards. But it began to haunt me. I'm the type that has to get it right the first time and if I don't, it will eat me alive until I go back in and correct it. But during the printer build, there was a pretty clear warning to use the shorter ptfe tube as the tube in the MK3 was longer. What was not clear at the time, and still not clear, was HOW MUCH shorter. And did they even include it in the kit. The shortest ptfe tube I had in that kit was the one that was already installed in the hotend. That's just a side bar.
So knowing that this tube was just a hair too long, I took the extruder apart and with my razor knife, sliced a hair off the inlet of the tube. Put it all back together and NOW having these issues. So the questions are:
1. How much difference in length is there between the hot end ptfe on the MK3 vs the MK3S...in other words, even though it was just maybe 2mm too tall, am I using the correct length tube. Would the MK3 ptfe tube be so obviously long that there would be no question I used the wrong tube?
2. Could trimming a mm or two off that tube cause even more slop to occur (without the spacer you mentioned) to cause this problem that was not a problem until I trimmed the tube apparently.
I know there is a spare tube in my leftover bags of parts. But come to think of it, there was talk about these tubes being too long and needed trimming which is why I did it in the first place. I do know by trimming that, I also shortened the tapered entry of the tube. But its not getting stuck while feeding through. Maybe I need to print another extruder set of plastics?
RE: MMU2S loading/unloading getting worse with time
I wanted to add that I did find the tube in the MK3S kit and verified I was using the right size tube. I tried installing a new tube into the hotend and temp fitting it into place and it still hits the bondtech. So rather than using a new tube, I made the decision to trim the old tube. I'm out of breath. 🙂
RE: MMU2S loading/unloading getting worse with time
Hi gnat...just checked my FINDA sensor with no print running and it's working filament in 1 filament out 0. I'm going to update the firmware to 1.0.6 on the MMU2S. Had a long chat session with support and no luck there either. Going to pull hot end out and see about the connection/clip.
RE: MMU2S loading/unloading getting worse with time
Don't know man...updated firmware to MMU2S to 1.0.6 and even though FINDA calibration is a success, the FINDA doesn't change state during a print on filament unload.....just stays on 1. Not sure if it's supposed to behave like that or not.. But I'm trying another print with no buffer and all my tubes cut down to 6". May just go ahead and order the custom tubes from matterhacker
RE: MMU2S loading/unloading getting worse with time
Updating my progress with this thread as this is turning more into a blog than anything. But seeing how common this issue is with this unit, I thought I would take the suggested, and in some cases missed steps I've discovered along the way.
My last print (Sd card 3 color lizard) after cutting the tubes short and MMU2S firmware upgrade to 1.0.6, failed to load/unload 5 times. Still too much to tolerate imo. So next step was to take apart the extruder to get to hot end. Here's where it gets interesting...for me anyway.
Once I got the hot end out using the instructions, I noticed the the "replace ptfe tube" instructions showed the proper way to install the ptfe tube. Which can be found here for the MK3S MMU2S https://manual.prusa3d.com/Guide/How+to+replace+PTFE+tube+on+the+MK3S-MK2.5S-MMU2S/1418
There is a very subtle step that needs to be done while inserting the ptfe tube to hold it in place. The black locking collar is supposed to be pulled out with your fingernails once the ptfe tube is fully inserted. Once I got my hot end out, I could see that collar on mine was NOT pulled out. And if I pulled on the ptfe tube, there was about 1/8" back and forth slop. I pulled the tube out and reinstalled it per the instructions and no more slop. Not sure if this is the culprit or not but the instructions do say that if this tube is not installed in this manner, it will "not work correctly". And even though the locking collar appears to hold its place, I could see the need for the c clip mentioned by gnat above. With all the loading/unloading/retractions etc, it's easy to see how this collar could become unlocked. https://www.thingiverse.com/thing:3559874 So I printed a few of them to be safe as I also have another MK3S with no MMU2S but I will take that one apart as well to inspect.
I haven't yet attempted a print. I'm just getting ready to reinstall. The point of this is that when these come out of the kit, they have the tube installed. It is unknown by the first time kit builder to inspect this tube for PROPER installation. There is also no clip that I could see as I didn't even know at the time the need for such clip. I will report back once I get this thing back together and try to print another lizard.
RE: MMU2S loading/unloading getting worse with time
With all the loading/unloading/retractions etc, it's easy to see how this collar could become unlocked.
Actually it's far more likely that either it wasn't locked during installation or it was bumped to the unlock position while installing it to the extruder body. Once in place with it properly locked it should not be able to unlock itself, though I wouldn't say it's impossible it is highly unlikely.
Glad you found it. Hopefully that fixes your large end issues.
RE: MMU2S loading/unloading getting worse with time
Ok...I got the hot end back together. Also, during the replacement of the hot end ptfe, I re attached the razor blade using some glue stick to hold the blade in place then put the cover on...so I know it was lined up correctly in it's slot. I ran a 3 color lizard print using matterhackers pro pla for the 2 main colors and prusa silver for the eyes. Everything was running fine for the most part up to about 60% (a big improvement) I started seeing some mild stringing on the filament tips and tweaked the temp back to 202. I am still not sure which direction to go with temp for stringing, but my intuition is telling me if it's stringing, it's too hot. But it does appear that even with this stock tubing, that it does tolerate SOME stringing I guess as long as the tip doesn't bulge. So just for testing purposes, I dialed the temp back to 205 which is the default temp in this gcode. As soon as I did that, I got some really bad stringing and it ended up jamming up the downtube causing the first load fail at about 63%. So back to 202C after restarting the print.
Next thing I was looking at while the print was running was the different "Mode" settings. I noticed that there is a setting for Stealth Mode for the MMU2. It looked like it was set to stealth mode, so I changed it to Normal just to see what would happen. Looks like stealth mode runs more quiet with slower loads and Normal loads faster and whines louder. I thought maybe a faster load would help shove any stringing down the tube with more authority to prevent jammimg in the down tube. But I was wrong. With MMU2 running in normal, it jammed within 10 or so minutes. So I unjammed and set it back to stealth mode....lesson learnt maybe. The print made it all the way to the end after that, but not without stringed tips at random and wondering if it will fail again.
The next thing I wanted to see is how a different filament would behave. I loaded up the same print with Prusament for the 2 main colors and the Prusa silver again for the eyes. I kept the MMU2 setting on stealth. And to make a long story short, this print went the entire 4 hours with no stringing at 205 and no jams.
So here are my takeaways and anybody can correct me if I'm wrong.
1. I don't think that razor blade does anything more than push strings out of the way...didn't see it cut one string. They are just too thin.
2. The razor blade, even if it did make cuts, never goes past the last filament. So if you are using 2 filaments, the blade never goes past filament #2 to make a cut. If you are using 3 filaments, the blade never goes past filament #3 to make a cut....same for #4 and it can't go past #5. So to me, the blade is pointless and only in the way of observing what the tips look like as they are unloaded.
3. Seems like stealth mode setting on the MMU2 not only runs more quiet, but it seems to make the whole process more reliable.
4. The matterhackers pro filament has a very nice glossy finish, prints awesome results, but doesn't seem to play nice with this loading/unloading set up as it likes to string on unload.
5. Likewise, it would appear that prusament was made for the MMU2
6. I would tell anyone to get this thing printing reliably before trying out any kind of buffer setup.
7. Make absolutely sure the ptfe tube in the hot end is pushed all the way in and the collar pulled out. Maybe it's a longshot that the collar will slip, but it only takes literally 2 minutes to print a "C" clip to make certain it won't.
8. I have ordered the capricorn tubing and just haven't gotten it yet. So this whole exercise was with the kit tubing cut to 6 inches on the inlets and the kit extruder tube. I plan on replacing at a minimum the extruder tube.
9. I may just run this whole setup "naked" with no buffer. I have the room and it makes it soooo much easier to load/unload and fix issues on the fly. It ain't pretty to look at, but then again neither am I. 🙃
RE: MMU2S loading/unloading getting worse with time
Adding a cooling move to the filament in Slicer also helps with the stringing in my experience
RE: MMU2S loading/unloading getting worse with time
Yeah I've heard that before. But how do you change settings on the gcode that's on the SD card. I looked at the gcode and did not recognize a cooling move or it wasn't addressed on the initial settings in the gcode. If I bring in the STL file to the Prusa slicer, will the file be the same if I export as gcode with the cooling move tweak? In other words, are these SD card prints optimized or used with prusa slicer defaults?
RE: MMU2S loading/unloading getting worse with time
Yeah I've heard that before. But how do you change settings on the gcode that's on the SD card. I looked at the gcode and did not recognize a cooling move or it wasn't addressed on the initial settings in the gcode. If I bring in the STL file to the Prusa slicer, will the file be the same if I export as gcode with the cooling move tweak? In other words, are these SD card prints optimized or used with prusa slicer defaults?
Cooling moves are generated into the gcode and not done in firmware. You'll need to re-slice the STL with the changed settings.
RE: MMU2S loading/unloading getting worse with time
Yes I realize that...it's what I said in my last post. I do have the STL files for the sample gcodes on the SD card. But the question is: Are the sample files on the SD card using the Prusa slicer default settings? If they are, then tweaking the cooling moves should not have any effect on the print. But if Prusa tweaked settings other than Prusa Slic3r default for the gcodes on the SD card, then what?
RE: MMU2S loading/unloading getting worse with time
Pretty sure the gcode has 1 move only
RE: MMU2S loading/unloading getting worse with time
Jeffrey: Have you found that razor to be needed at all since your last post? Mine popped out for the second time and I've yet to see it provide a function. Just trying to decide whether this particular hassle is worth bothering with. This printer is good, but the MMU portion feels like a war of attrition to some degree 🙂
RE: MMU2S loading/unloading getting worse with time
Jeffrey: Have you found that razor to be needed at all since your last post? Mine popped out for the second time and I've yet to see it provide a function. Just trying to decide whether this particular hassle is worth bothering with. This printer is good, but the MMU portion feels like a war of attrition to some degree 🙂
I took my razor out when I realized it wasn't cutting anything at all. If you ask me, the razor blade is a bit of "Hail Mary" engineering. My MMU2S is working well without it. I think replacing the PTFE tubing with blue Capricorn and making sure the locking collar on the heat sink short PTFE is secure (tubing pushed all the way in and collar pulled all the way out). I made C clips for added insurance on the locking collar. Seems like satin finish filaments like the Prusament PLA yield the least amount of loading unloading problems. The glossys seem more susceptible to stringing in my limited experience. And pushing stringed tips back down through the tubing is where most of the jams start IMHO. Having the blade removed also allows a better view of the filament tip as it's pulled out so you can maybe predict a jam if you start to see strings coming off the tips. The capricorn tubing seems to be more forgiving. If I were to make a judgement call on the MAIN culprit on my machine, I would say it was the slop in the short PTFE because it was not seated properly.
RE: MMU2S loading/unloading getting worse with time
@jeffrey-d10
Amen to your Hail Mary! I used a chisel to better define the seat for the knife blade in my Prusa-printed part. Even still, it has twice become askew and dug into the opposite side of the printer, requiring me to take the MMU2 apart each time. I will reprint this part myself when I get my MMU2 running again...
RE: MMU2S loading/unloading getting worse with time
yesterday I had to troubeshoot my MMU2S also and came over your thread.
I got this stringy tips also and was wondering what's the cause.
In my case, a couple of days before I "improved" my setup a bit.
I have two shelfes over my printer on each shelf is a Filament box with three filament roles.
What I did was to attach ptfe tubes between the filament boxes and the buffer.
That may caused too much friction between the upper box and the mmu.
The rest of my setup is MK3S and MMU2S stock without modifications and using original Pruasment for print.
I re-watched the videos of Chris' basement about the MMU (he has an MK3 and MMU2 without S!) and the Prusa videos about MMU2S.
After cleaning up my MMU2S idler gears, checked the extruder idler and removed the last installed ptfe tubes from the upper filament box.
The next testprint shows lovely made filament tips again with one cooling move and stock prusaslicer PLA profile.
The lesson I learned is, don't change too much at a time and watch after every modification what MMU2S does after it.
This will help you to figure out what's possibly the cause.
RE: MMU2S loading/unloading getting worse with time
@mike242
True. I learned this when I "upgraded" to the MMU2S. It absolutely ruined my experience with this printer, to the point where it is no longer enjoyable to work with the Prusa at all. I regret the entire purchase.