The importance of the filament detection sensor calibration (MK3S) and other problems
This is an absolutely fundamental step for the MMU3 to function correctly, and the explanation in the manual is extremely poor. Using words like 'slightly' or 'tighten it a little bit' is not serious, to be honest. Beyond the fact that the explanation could have more concise steps, there is a serious problem in the design of the 'Chimney' and the 'Chimney-base' parts.
Speaking in detail, the 'Chimney-base' has only one gripping screw to the extruder, meaning that while it is adjusted, it has a slight angular movement that affects the indication of filament presence. It is also possible to displace it linearly, which sometimes happens when adjusting the calibration screw.
Then, in the 'Chimney' (referring to calibration point 9: 'Now you need to start loosening the calibration screw as much as possible but the Fil. sensor value should still say "1". If it switches to "0", tighten it a little bit.'), when loosening the screw, there is no mechanical movement produced. This is because there should be a spring that pushes it, so as we loosen the screw, it is necessary to push it manually and check the state of the sensor. In addition to all this, the sensor is extremely sensitive. Even if we use the 1.5 mm hex key and correctly check the state of the sensor, it is necessary to perform many load tests to verify its proper functioning, which can take a really long time and is quite tedious.
The issue with the high sensitivity of the sensor is that the loading and unloading of filament in Prusa Slicer are designed and tested with their own filaments. I had to find the exact configuration for the filaments I can buy in my country. It was trial and error after another until I found a correct setting that doesn't jam the extruder with every filament change because of a bad filament tip
As a piece of advice, two things that helped me a lot to achieve satisfactory prints were: enabling automatic filament cutting after 2 failed loading attempts and activating "High current on filament swap" in Prusa Slicer.
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
i had MMU3 with Mk3s, and I used a filament sensor with a 5 volt LED soldered between the +5 volt pin and the output pin, to give immediate indication of operation.
it has worked well with my noname filaments.
regards Joan
I try to make safe suggestions,You should understand the context and ensure you are happy that they are safe before attempting to apply my suggestions, what you do, is YOUR responsibility. Location Halifax UK
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
My filament sensor also gave me troubles originally. It did not reliably detect the filament as the tolerance of the latch was a tad too high for my build at least. What I did was to modify the sensor latch, extruding the tip that triggers the sensor in 100 µm steps and testing a few alternative versions. I can't remember witch which version I settled but it was extending maybe 200 µm further than the original. With that sensor it has worked reliably ever since, no matter what filament I am throwing at it (ok, the MMU absolutely hates flexibles but that has other reasons).
Mk3s MMU2s, Voron 0.1, Voron 2.4
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
Thanks! May be I'll do it to save time when I calibrate
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
My filament sensor also gave me troubles originally. It did not reliably detect the filament as the tolerance of the latch was a tad too high for my build at least. What I did was to modify the sensor latch, extruding the tip that triggers the sensor in 100 µm steps and testing a few alternative versions. I can't remember witch which version I settled but it was extending maybe 200 µm further than the original. With that sensor it has worked reliably ever since, no matter what filament I am throwing at it (ok, the MMU absolutely hates flexibles but that has other reasons).
Thanks for your answer! Did you modify the extruder idler? Anyway, after proper calibration and with the correct PrusaSlicer parameters, I've been getting good prints. But, well... I don't like the calibration procedure.
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
Yes I modified the tip of the idler a tiny bit, just where it triggers the sensor. Extended that part maybe by 200µm. Then it reliably blocked the light path of the sensor, when filament was loaded, no matter the tensioning or filament type but also reliably kept it free when no filament was loaded.
That was a bit quick and dirty and non-adjustable but very easy to do.
Mk3s MMU2s, Voron 0.1, Voron 2.4
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
I've noticed intermittent issues with loading to fsensor when using one filament (so far.) All I can attribute this to would be that the filament is less dimensionally accurate in some spots. All my sensor calibrations have been fine.
When it cries at me, I squeeze the fsensor screws with my finger/thumb while it loads. That being said, I am still working on a permanent solution, and I think making a new MK4 nextruder fsensor idler like @Thejiral said is a great idea.
X marks the spot.
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
can you please tell me how to fix this piece of shit? the filament sensor does not work at all. ever. i hate prusa so god damn much at this point. i just want to salvage some of the effort ive wasted on these absolute pieces of trash in the last year
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
Specifically what is the problem?
I went through the MK4 kit assembly instructions and carefully re-completed everything about the nextruder and fsensor trying to be as exact as possible. I did calibration and it has worked since then. In particular the nextruder idler tightness (I still need to make a new idler like @thejiral.)
The only issues I run into somewhat regularly is some stringing inside the nextruder, which after bits of filament break off, results in a "FSENSOR stuck filament" error. This happens more if trying to print with multiple filament types (PLA & PETG for instance)
I fix this by removing the nextruder front plate and gear and cleaning it out.
Causes I have seen:
- Wrong/non-ideal temperatures in the filament profile (causing stringing). Sometimes the first layer has a different temp and the issue will only happen on the first layer.
- Cheaper/other filament (not all PETGs have the same tolerances and printing specs for instance).
As a result, the MMU3 to FSENSOR loading/unloading is very picky and using the "generic" profile (loose tolerance) for all my filaments may not work.
So I've resorted to using one brand of PETG and setting up really specific filament profiles for that brand. It helps, but it's not perfect.
X marks the spot.
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
Thanks The Jiral
regards Joan
I try to make safe suggestions,You should understand the context and ensure you are happy that they are safe before attempting to apply my suggestions, what you do, is YOUR responsibility. Location Halifax UK
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
can you please tell me how to fix this piece of shit? the filament sensor does not work at all. ever. i hate prusa so god damn much at this point. i just want to salvage some of the effort ive wasted on these absolute pieces of trash in the last year
I feel the same way.. I really hate the MMU3. Is this a serious product?? It looks cool and stuff but it isnt reliable at all. The main Prusa caracterisitic is its reliability, this is not the case. I used to love printing with my MK3S, It was so so easy. I completely regreat about purchsing the multimaterial, I fight with it every day. Every printing there is a new challenge. In some moment I tought, may be its me, I broke something and I didnt realize about it. So I removed the MMU3 and went back to the MK3S original form and for no surprise at all the magic came back, smooth printing without any problem! I have been 3d printing since 10 years ago, I have experience to say DONT BUY THIS PRODUCT. If you wanna print multimaterial save some money and go to another brand
RE: The importance of the filament detection sensor calibration (MK3S) and other problems
This user mod is helping a little. Mine still doesnt work but at least the number changes sometimes based on whether there is filament or not. https://www.printables.com/model/7789-mmu2s-easy-ir-sensor-calibration
RE:
One thing I have noticed as well. Once I have everything working as well as I can make it, things are still not perfect. The MMU is certainly finnicky. But if I slow my MK4 down by changing tune > print speed = 90% that reduces the number of errors I have quite a bit. I haven't tried it at a lower speed (yet). This is because with things at full speed, the wipe tower gets clumps and strings and lines that won't stick despite all of my settings.
I ran an 11 hour MMU3 print yesterday with around 200 tool changes and 4 filaments and only ran into I think 3 "stuck filament" issues.
What we need is a wipe tower speed setting (in percent) so that it can be reduced by itself without affecting the rest of the printing.
X marks the spot.