The Answer
I believe I have the answer - and there is some good news but mostly bad news... (I do not know how to post pictures here so if someone can explain it to me I will be happy to post some pictures)
First the good news: It IS possible to make a reasonable decent multi-color print using the MK3 and MMU2.
But... it is never going to be easy or robust and it will always require extensive setup and pre-print manipulation.
What kind of setup and pre-print manipulation? (After this I will go on to explain the basic problem - which actually correlates most of the other symptom reports we have seen)
The Bowden Calibration is important - and it is critical to repeat the calibration steps many times to make sure that the Bowden Calibration has settled in (there seems to be a hysterisis affect particularly when you have make large adjustments. The Bowden Calibration also seems to shift so you might have to perform the Bowden Calibration multiple times after running some prints. You might need to check the Bowden Calibration before each print - particularly if you have a long or detailed print.
There are setting is Splicer that need to be adjusted - particularly colling and retraction rate so that the melted bulge at the end is minimized. Too much bulge and you get too much friction in the tubes.
You will need to cut off any part of your filaments that has already run through the gears (you will see the indentations on the sides of the filament where the gears dug into the filament)
You will need to re-run the insertion of each filament into the MMU and you will need to relax each spool to make sure you have as little tension and friction as possible. (One of the testes was to cut off a long piece of filament for each station and run it without the spools and with as close to zero tension and friction - yes that does help)
So What's the Basic Problem?
The basic problem turns out to be the relative softness of the filament, often exacerbated by added heat into the system...
This whole device depends on the bondtech gears being able to reliably and repeatedly drive the filament - whcih at first glance is no different than the MK3 in single color mode all by itself - right? ...umm... no... The key differences are that there are now two stages of bondtech gears AND we are expecting the printer and MMU to repeatedly load and unload each filament - add to that the MMU only has one gear (the topis the flat bearing).
What happens is that the bondtech gears slowly eat into the filament and wear it down. Keep in mind that depending on your design and how much of each color you use the same section of filament can be going through those gear many , many times. You will more likely have bnetter success on prints that require few color changes and ones that use a fair amount of each color before changing.
We have seen several posts where people seem to have noticed some correlation between failures and added heat (perhaps from the stepper motors). This now makes perfect sense. Added heat makes the filament softer and more succeptible to the gears slipping and eventually carving out the dreaded half moon where it can no longer advance the material. Yet another factor that all leads to the same conclusion is that the most common time of failure that I have seen is when the retraction of the filament goes from slow to fast - this is when the MMU applies more torque which equals more force into the filamnet and then strips out the filament. Really doesn't matter how well you route the filament it will eventually strip the filament at least to the point where the calibration goes out and it missess the entry point to the extruder gears.
Add to this the occasions when the retration from the extruder adds a bulge or a string and then thereby adds more friction - which requires more force input from the gears to the filament - and you have a system that just cannot work reliably.
Once someone shows me how to post pictures I will share prictures of the test setup I used and I think it will be clear if it is not already.
Perhaps flowing COLD air onto the filamnet might help to keep the filament hard enough to withstand the gears but I think that would be problematic for a number of reasons - not to mention what it would do to the stability of the material.
Sadly, I think this is a critical problem that cannot be solved easily.
Comments?
Re: The Answer
Comments?
The gears on the MMU2 are disengaged completely when the filament isn't being loaded or unloaded, so I'm skeptical that repeatedly running the same filament through two sets of drive gears is really an issue.
There are people, including myself, who have gotten their MMU2 to work reasonably reliably. However, it's clear that there are so many more variables with the MMU as compared to the stock MK3 that getting things to work well requires that you start with a solid foundation of experience printing with a stock MK3, and then build on that a fair amount of practice with the unique challenges of multimaterial printing.
Re: The Answer
[...]requires that you start with a solid foundation of experience printing with a stock MK3, and then build on that a fair amount of practice with the unique challenges of multimaterial printing.
Or use the wonderful people of the forums to help!!! I think I'm on my way towards my first successfully-self-sliced print thanks to everyone here!
Re: The Answer
To add pictures, when creating a post, simply drag the photo and drop it onto the text area. Or click the "Add attachment" button below the text area.
Note that only certain types of attachment are allowed.
We all have our own thoughts about why this fails; mine are somewhat different and revolve around the stringing. Prevent stringing and you will get a good print.
Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…
Re: The Answer
To add pictures, when creating a post, simply drag the photo and drop it onto the text area. Or click the "Add attachment" button below the text area.
Note that only certain types of attachment are allowed.
We all have our own thoughts about why this fails; mine are somewhat different and revolve around the stringing. Prevent stringing and you will get a good print.
Peter
I'll definitely look more into the stringing. Any thoughts on what impacts stringing?
Re: The Answer
Comments?
The gears on the MMU2 are disengaged completely when the filament isn't being loaded or unloaded, so I'm skeptical that repeatedly running the same filament through two sets of drive gears is really an issue.
There are people, including myself, who have gotten their MMU2 to work reasonably reliably. However, it's clear that there are so many more variables with the MMU as compared to the stock MK3 that getting things to work well requires that you start with a solid foundation of experience printing with a stock MK3, and then build on that a fair amount of practice with the unique challenges of multimaterial printing.
I am not referring to when the filament is not being loaded and unloaded... it is exactly when the filament is being loaded and unloaded that matters. Think about it... each time the filament is loaded it is dug into by the gears - that's the only way that the filament can be moved. Unless you use up all the filament that is between the MMU and the extruder each time then you will be going over the same filament area each time. I ran tests and looked at the filament over repeated color changes and was able to see significant wear on the filament. As I noted, I have been able to achieve some good prints - but it is a hand holding process for each print - and that doesn't even get into the huge waste blocks. I'm not trying to criticize the MMU or Prusa. I think the MMU is a very cool design - I just think that the filament media makes repeated transit of the filament very problematic.
Re: The Answer
Any thoughts on what impacts stringing?
The inability of the hardware to consistently retract 15 mm of filament at 100 mm/s with acceleration set to 8000 mm/s^2 whilst creating a serious vacuum.
Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…
Re: The Answer
Preventing filament grinding (settings to ensure the lower gears and the upper gears aren't fighting) and minimizing forces on the filament during changes made ALL the difference.
There's still plenty of tuning to do, but my biggest problem is having to reset the machine 6+ times to get it to boot up right. Once that happens, the other changes ensure a good print. Tips and Friction and you should be good.
I maintain an informal list of San Diego, CA 3D printing enthusiasts. PM me for details. If you include a contact email and I can add you to the informal mailing list.