Filament head too large after unload (MK3S + MMU2S)
I have a problem with my MMU2S, the filament head is consistently coming out of the extruder too large and gets stuck, either in the extruder assembly, the tube connecting it to the MMU or the selector. And if by some random chance it manages to unload successfully, then it gets stuck on the next load attempt.
I've checked everything mechanically and it seems fine, so the most probable reason is the filament I'm using (Devil Design PLA). I know I need to adjust the toolchange / ramming settings to account for it, but here's the catch. Even though the individual parameters are documented and I understand what they do, I have no idea how they affect the shape of the filament after the unload (e.g. would increasing the number of cooling moves make the head thicker or thinner? etc). And there are far too many parameters to just change them at random and hope for the best 🙂
Is there some better description of the head forming process somewhere? Or maybe could I at least get a few hints on where to start?
RE: Filament head too large after unload (MK3S + MMU2S)
This is my issue, too. Support have tried to be helpful, but went straight for "user error" and had me tearing my machine to bits... twice, then factory resetting and reflashing - none of this did anything. My tips were coming out nicely shaped but too wide at around 2.06 mm and failing to reload at the lowest temperature I could set that wasn't stringing (around 201-210 depending on material). I have almost zero problems at the MMU unit end, it's all about reloading retracted tips causing failures.
I have just now had several successful tool changes in a row. Fingers crossed. Here are the things I remember changing 🙂
Filament Settings: Temp at 210 for other layers. I believe the target to be as hot as you can without stringing (tune this).
Unloading speed at the start: 120 mm/s. I have seen a thread here say it's a placebo and the printer can't actually deliver this. I don't know what to say there, but if I revert it, my success rate dips massively.
Number of cooling moves: 4. Should I get stable results, I will try reducing this.
I tried changing ramming settings... this seemed to not help at all.
I tried going cooler, it seemed to make things worse (wider). I tried going hotter, I got stringing and failed in a different way.
I do feel your pain here, it seems overly sensitive and requires precise settings that must deviate based on potentially material, perhaps some level of precision assembly (aligning things by eye is going to have deviation), maybe ambient temperatures play a part too. I've just witnessed many successful changes in a row whilst typing this. Wish me luck!
RE: Filament head too large after unload (MK3S + MMU2S)
So far for me, I have been able to significantly reduce the number of load/unload failures using what was mentioned previously.
I am using eSun PLA+ as the material and the Prusament PLA as the baseline in PrusaSlicer.
First Increasing the unload speed from 100 to 120 significantly increased success rate. I went from dozens of errors (which would eventually ruin the print at some point) to 6 or 7 at most (on a 40 hour print).
Next, I increased the number of cooling moves to 2 and the next print only had a single error. I was looking at the filament tips and increasing cooling moves seem to make the tip less stringy. My next print will have 3 cooling moves to see if that solves it.
I have found that the MMU2S is very fickle about this, and requires a bit of tuning to get right. It really makes me wonder if there should be some calibration procedure that could be done to fine tune the settings needed in PrusaSlicer. Something like how the first layer calibration is done. It could be a test where it essentially tries different settings based on user feedback and spits out the settings needed. (like how glasses prescriptions are determined, if that makes sense).
Anyways, I digress.
Jonathan Kayne
Virginia Tech Class of 2021 - Electrical Engineering
Thingiverse Profile: https://www.thingiverse.com/jzkmath/about "I am always thinking about making. My future begins when I wake up and see the light." - Miles Davis
RE: Filament head too large after unload (MK3S + MMU2S)
Ok, this helped me quite a lot. Good call! 🙂
RE: Filament head too large after unload (MK3S + MMU2S)
I haven't had time to play with the settings yet, waiting for the weekend. I did notice another problem though - sometimes when the MMU wants to unload filament that is still inside the nozzle (or past the extruder gears anyway) it starts pulling it with the MMU motor, but the extruder motor doesn't move. The filament is gripped tight by the extruder gears, and the MMU motor cannot pull it out, so it keeps grinding the filament in place until it strips about half of it's width and fails completely. If I see it happening, I can intervene by releasing the extruder tension screw, so that the filament is free to move. What could be the problem here? Faulty IR sensor?
RE: Filament head too large after unload (MK3S + MMU2S)
I haven't had time to play with the settings yet, waiting for the weekend. I did notice another problem though - sometimes when the MMU wants to unload filament that is still inside the nozzle (or past the extruder gears anyway) it starts pulling it with the MMU motor, but the extruder motor doesn't move. The filament is gripped tight by the extruder gears, and the MMU motor cannot pull it out, so it keeps grinding the filament in place until it strips about half of it's width and fails completely. If I see it happening, I can intervene by releasing the extruder tension screw, so that the filament is free to move. What could be the problem here? Faulty IR sensor?
must be the IR sensor, if it starts pulling with the MMU units it thinks there is no filament between the gears at this point, do the IR sensor calibration and check if you get consistent readings from the sensor.
RE: Filament head too large after unload (MK3S + MMU2S)
Been banging my head against this problem as I still have a huge fail rate. I've ordered some parts to help me by the weekend.
1: Ordered Capricorn 1.9 tubing. Yes this is a bit bigger than the 1.85 Prusa use - but what are the tolerances of the Prusa tube? Who knows. I can also actually get the capricorn tubing without £10 delivery, which helps.
2: I'm faily sure my heatbreak isn't a Prusa official part now. After looking them up to make sure there is actually a difference (there is) I ordered a pair from e3d as I'm in the UK and they know how to send small things in envelopes cheaply.
3: Switching to auto-rewind spools and much less PTFE tubing.
Hope for a good update on the weekend...
RE: Filament head too large after unload (MK3S + MMU2S)
Been banging my head against this problem as I still have a huge fail rate. I've ordered some parts to help me by the weekend.
1: Ordered Capricorn 1.9 tubing. Yes this is a bit bigger than the 1.85 Prusa use - but what are the tolerances of the Prusa tube? Who knows. I can also actually get the capricorn tubing without £10 delivery, which helps.
Don't use the Capricorn for the hotend. Believe me I've been there. It will cause problems sooner or later.
RE: Filament head too large after unload (MK3S + MMU2S)
Been banging my head against this problem as I still have a huge fail rate. I've ordered some parts to help me by the weekend.
1: Ordered Capricorn 1.9 tubing. Yes this is a bit bigger than the 1.85 Prusa use - but what are the tolerances of the Prusa tube? Who knows. I can also actually get the capricorn tubing without £10 delivery, which helps.
Don't use the Capricorn for the hotend. Believe me I've been there. It will cause problems sooner or later.
Care to elaborate? I can't find 1.85 ID PTFE anywhere apart from Prusa and I'm loathe to pay 10x the item price in delivery charges.
RE: Filament head too large after unload (MK3S + MMU2S)
Sorry I missed this thread earlier.
My first thoughts are A) make sure the PTFE in the hotend is fully inserted in the correct direction and B) try a different filament manufacturer.
If the PTFE is not fully seated it could leave a gap that would theoretically allow the filament to expand if it was still hot. Similarly if it was inserted upside down the interior chamfer would provide a similar expansion area.
I have one garbage PLA (Silk) that does all kinds of weird antics as it cools. I have no intent to ever put it through my MMU (or the printer ever again) as I fully expect it to make a hash of things. Trying a different filament brand may produce different results for you.
Another thing to look into is how smoothly the filament is flowing between the spool and MMU. Drag in this area can help cause any number of load and unload issues.
In regards to the MMU driving while the filament is still in the extruder, that is definitely an issue with the IR sensor. Either the tower needs adjustment to get a consistent reading from the door flag or the door/flag need some adjustment.
RE: Filament head too large after unload (MK3S + MMU2S)
There is a visible to the naked eye difference between the "normal" e3d v6 heatbreak and the Prusa e3d heatbreak, in that the Prusa version has a narrower opening to the nozzle. This explains why my tips were too wide - they were being made in a wider tube. Having those wide tips had enlarged my hot end PTFE, it should have been 1.85 ID but was basically standard width, the capricorn was noticeably tighter despite being 1.9mm, so the capricorn is now in my hotend.
I'm currently watching 2 x two colour benchys (with opposite colouring so it needs to change every layer) go. Zero tool change failures so far *fingers crossed*
RE: Filament head too large after unload (MK3S + MMU2S)
Standard E3D-V6-175 heat break vs Prusa modified version. Standard is 2.0 mm internal diameter full length. Prusa has a stepdown from 2.2 mm to 2.0 mm at the melt transition zone.
RE: Filament head too large after unload (MK3S + MMU2S)
Indeed, I had seen those. What I wasn't expecting to see from those drawings was quite a difference in the diameter of the hole in the heatbreak where it butts up to the nozzle.
RE: Filament head too large after unload (MK3S + MMU2S)
1 skipped layer in 240 tool changes. Not too bad.
RE: Filament head too large after unload (MK3S + MMU2S)
I spend a week dismounting and mounting the extruder. 😕
Often my filament heads were so big (probably 2.2) that they could not pass through the PTFE tube (1.85), and the only way to remove the filament was dismount everything (it did not went neither down if I pre heath, grrrrr! ).
At that point, I realized that ONE of the MMU2S PTFE tubes was out of tolerance (around 2.3), so I cut 5 cm and after setting up the shapes of the ends, I installed it on the extruder cooling section. This resolved the block of the filament inside the cooler. But was not enough. 😒
While I was testing, I noticed that a bigger head was now able to pass through the cooling section regularly, but when it was removed from the extruder, it may jam inside the PFTE tube from MMU2 to extruder, since that is 1.85mm only and the head may be bigger.
So I replaced the 4/1.85 tube from MMU2 to Extruder with a new one 4/3mm. 🙂
Since then, all my extruder blocks ended 😀 (cross fingers). I see often filament with big head that is regularly unloaded and loaded by MMU2.
Probably Joseph would say that it was not necessary and it was a matter of fine tuning the filament parameters, but for me that was an easier solution, since I have too many filament spools, and it would be too time consuming test and tune each single filament.
RE: Filament head too large after unload (MK3S + MMU2S)
I, too, am encountering large filament heads resulting in it getting stuck. For me the problem seems to stem from the MMU not always disengaging the idler pulley after changing colors (reported here: https://forum.prusa3d.com/forum/original-prusa-i3-mmu2s-mmu2-assembly-and-first-prints-troubleshooting/mmu2-sometime-doesnt-release-filament-video/ ). This creates a lot of friction for the extruder to overcome and I *think* this interferes with the tip forming procedure (tips look fine when it properly releases). I don't yet know what the solution, is, though.
RE: Filament head too large after unload (MK3S + MMU2S)
I, too, am encountering large filament heads resulting in it getting stuck. For me the problem seems to stem from the MMU not always disengaging the idler pulley after changing colors This creates a lot of friction for the extruder to overcome and I *think* this interferes with the tip forming procedure (tips look fine when it properly releases). I don't yet know what the solution, is, though.
Are you sure ?
I want to do the devil's advocate. 😎
Imagine that the PRUSA MMU2 firmware detect an high motor torque required to move the filament in the MMU (it technically can) .
It could assume that the motor effort to move the filament only from the Extruder gear will be even higher, and that the filament could slip creating irregular extrusion. So it decides to "help" the extruder gear also with the MMU gear to improve the torque force to print.
And so, an issue may be in reality an advanced feature 😎
RE: Filament head too large after unload (MK3S + MMU2S)
I am sure that the MMU does not attempt to help the extruder and I observe this most of the time (can see the idler bearing get rotated away from the knobbed pulley in the MMU as soon as the filament reaches the bondtech gear in the extruder). Only occasionally does the idler fail to rotate away. I have also been studying the firmware and it clearly calls the motion_disengage_idler function as soon as the IR door sensor in the extruder detects filament. I cannot find any reason why this doesn't always work. It does look like if there is a fault on the tmc2130 the idler may not get retracted. I guess I will compile my own version and try some experiments to get to the bottom of it.
Can anyone else confirm seeing this happen (idler gear not disengaging)? For me it happens about once in ten filament changes and when it happens doesn't always result in a failure.
RE: Filament head too large after unload (MK3S + MMU2S)
I've seen this before... but... never could replicate it reliably enough for a vid. If one of us can catch it on vid, that'd be awesome, buuuut... I'm packing my printer to send back to Prusa.
RE: Filament head too large after unload (MK3S + MMU2S)
See my link in my first post above.