34 Days - ZERO successful Multi-Material Prints - Have tried every suggestion out there.
 

34 Days - ZERO successful Multi-Material Prints - Have tried every suggestion out there.  

Page 1 / 2
  RSS
mark.v18
(@mark-v18)
Active Member

MMU2S was delivered on 15 May. It is now 18 Jun. I have had ZERO successful multi-material print jobs. Every time, job will get many filament changes in (usually multiple hours) and then will get random load failure or unload failure. I have reset printer and mmu board multiple times. I have adjusted finda, IR sensor chimney, mmu2s tension, all of it multiple times. IR sensor reads correctly. Single filament jobs work perfect. I have scoured internet including this forum (and others) and nothing has worked. I am only trying to print either Prusa provided mmu models or this one thingiverse model (thingiverse.com/thing:3192133). None of the print jobs fail on first few filament changes for a particular filament - it is only after multiple which makes no sense to me. Maybe I have bad unit? Please help! Thank you!!

Where you upgrading and what machine did you start with:  Started with my first MK3S in early March (first Prusa printer ever, but not first 3d printer).  It also printed perfect.

Are you using Prusa pre-built Gcode?  No, I am slicing in Prusa Slicer 2.0 with all defaults.  Models (except the one thingiverse) have all been prusa provided models.  For the single color prints with mmu2s it is still prusa model as I've been slowly printing the enclosure pieces (now the enclosure v2 pieces).

Posted : 18/06/2019 12:54 pm
gnat
 gnat
(@gnat)
Noble Member

When it fails, what is the cause?

What do you do to try to resolve it and move forward?

What is your spool to MMU configuration?

I had a perfectly working setup, but it was untenable due to space (printer was actually hanging off the desk a bit), so I finally built a frame to holder the spools above the printer. Now it no longer works. Similar to you it gets a few hours in and then quits. 

My issue (and I suspect yours) is that I haven't been able to get the friction on the filament right again. This is causing too much drag which is leading to stringing and sooner or later this causes it to get hung up.

I learned in my original setup that friction/tension/drag between the spool and MMU is a bad thing. Eliminate as much as possible and you should see better prints. 

MMU tips and troubleshooting
Stop SL1 tank leaks with a simple o-ring...
Posted : 18/06/2019 1:23 pm
mark.v18
(@mark-v18)
Active Member

Hi gnat.

When it fails, what is the cause?  "...and then will get random load failure or unload failure."

What do you do to try to resolve it and move forward?  It depends on the failure.  Sometimes it's as simple as pressing the middle button on the mmu2s, getting success indicator from the lights, then pressing right button to resume print.  Sometimes it's lift top part of mmu2s, retipping filament, and then going through reload.  Sometimes it's reset because the mmu2s has lost it's mind and won't give success led indication no matter what.  Failures are random without rhyme or reason.

What is your spool to MMU configuration?  I have tried different layouts without success.  At one point I had it laid out across most of our large dining room table that seats 8.  Right now, it is roughly set up in the newly touted (by prusa) prusa enclosure 2 setup (I say roughly because I am using the prusa spool holders on top and I am still using the leg extensions for when I had the whole printer and spool in one ikea lack (a littler taller than needed but plenty of space for the ptfe tube from the mmu2s to the extruder).  Are you suggesting that I try the auto retracting spool holders?  I have not done that admittedly because I've read the tension will self-adjust (getting looser) and at some point you run into the same issues if you don't consistently adjust tension.

No, I am not having stringing issues that I am aware.  The tip always appear fine.  When I do re-tip the filament is just out of trying something to get the darn thing to work.

What else have I tried?  I have even done the "turn the mmu2s idler screws and extruder tension screws the same number of turns so there is the same tension at both places".  Which is 7 full 360 degree turns for me.

I have digital calipers so have used several settings for the finda based on various posts as well as trial and error.

Does any of this help you?  I do appreciate any and all responses.  Right now I feel like I wasted $300.

 

Posted : 18/06/2019 2:21 pm
gnat
 gnat
(@gnat)
Noble Member
Posted by: mark.v18

Hi gnat.

When it fails, what is the cause?  "...and then will get random load failure or unload failure."

Failures aren't random. There is always a reason for them even if it is not always easy to figure out.

Are you aware that the FINDA lights up like the PINDA does? Some are unaware of this as it is so faint compared to the FINDA. I find placing my hand between the FINDA and the LEDs from the MMU board let me see it better. Is it in the correct/expected state when you have a failure?

Are you suggesting that I try the auto retracting spool holders?  I have not done that admittedly because I've read the tension will self-adjust (getting looser) and at some point you run into the same issues if you don't consistently adjust tension.

So you are using the buffer then? Is the filament flowing smoothly into and out of it?

At all points you should be able to easily push the filament from one point to another (e.g. spool to buffer and buffer to mmu) without counter pressure from your other hand. If it sticks along the way or you have to hold things steady, then you have an issue to address.

If you have the space for the buffer then that should work and that should be the better option. The auto rewind options are more for those of us that don't have the space for that setup. If your current setup isn't working, however, printing a couple and seeing if you can then get through a two color print may be worth a shot.

I use auto-rewind spools and while they allowed me to print, I agree with many of the detractions against them. Adjusting the tension is an art and you have to stay on top of it.

Another option you can try (though it will be tedious) is to insert the short PTFE tubes into the back of the MMU and then feed the spools directly to the MMU (e.g. no buffer) and then try to print. Obviously you'll need to stay on top of rewinding the spools yourself to keep a rat's nest from forming and messing things up.

No, I am not having stringing issues that I am aware.  The tip always appear fine.  When I do re-tip the filament is just out of trying something to get the darn thing to work.

Remove the PTFE holder from the selector and the PTFE tube from the back of the MMU for the currently selected tool path. Then manually push a piece of filament through both directions and see if you knock anything out. I had a failure last night and what I finally found was that while the tip of my red filament looked perfect and it was not involved in the failing tool change, there was string of red caught up in the selector that was fouling things up.

Does any of this help you?  I do appreciate any and all responses.  Right now I feel like I wasted $300.

I know how you feel. I haven't printed much since I got the MMU. Between getting the MMU working and then not having success breaking up single file models I've kinda lost interest in printing right now (compared to the printer running nearly non-stop from mid Dec to late April).

MMU tips and troubleshooting
Stop SL1 tank leaks with a simple o-ring...
Posted : 18/06/2019 2:49 pm
mark.v18
(@mark-v18)
Active Member

Thanks again Gnat!  All good stuff. I say random as the failure is not repeatable when doing the exact same thing with the same model.  Example, start five different times with clean bed/machine with same model.  First time it fails after 47 minutes with a load failure.  Clean bed/machine start that same model, couple hours in it fails on an unload.  Etc. Yes, all of the failures have meaning/cause.  Sorry.....I do remember  reading at one point about the finda lighting up but had forgotten.  Will check that....Trying to stay with Prusa buffer and have the room...Will also double check each lane for obstructions as suggested.... 🙂 I'm still printing plenty of single material but still motivated to get mmu2s right.  Thanks again!

Posted : 18/06/2019 3:05 pm
gnat
 gnat
(@gnat)
Noble Member
Posted by: mark.v18

Trying to stay with Prusa buffer and have the room...Will also double check each lane for obstructions as suggested.... 🙂

Also try to make each PTFE tube as straight as possible from one end to the other. Where curves are required, make them as gentle as possible.

Similarly make sure filament enters the short tubes as straight on as possible so that it's not pulling the tube  in any direction or trying to make a turn to go into the tube.

MMU tips and troubleshooting
Stop SL1 tank leaks with a simple o-ring...
Posted : 18/06/2019 7:43 pm
mark.v18 liked
matt.t11
(@matt-t11)
New Member

I was also having random filament loading/changing issues after the MK3S/MMU2S upgrades (from MK2/MMU2) and made the same change in 2 places to fix the problems.  I noticed that when the tips were not ideal, they could run into the edge of a part surface and could not be guided into the holes properly.

I took a sharp knife and made a slightly conical shape on the top of the "orange Adapter-printer-mmu2s" (as named in the assembly instructions) and also made the same conical shape on the "mmu2-selector" (as named in the assembly instructions).  Since I have made these 2 changes, I have not had any loading failures and hundreds of successful filament changes including 2 successful completed 5 color prints.  

Posted : 18/06/2019 8:31 pm
mark.v18 liked
robert.s57
(@robert-s57)
New Member

Hi, Mark.v18

Do you have bearings that make some noise?

On "fail stats MMU", what are the total values, especially interesting is the MMU power fails.

This post was modified 2 years ago by robert.s57
Posted : 18/06/2019 9:56 pm
Nikolai
(@nikolai)
Noble Member

Hi guys,

I would bet the FINDA is placed either too low or too high.

@mark: Try to understand how the system works so you can clearly identify which step is failing. It helped and still helps me a lot to print successful MMU2. Since the 3.7.x firmware upgrade the printer is always in a defined and recoverable state.  

For example during unload you should see the FINDA light off, once the filament passes the FINDA, the light should turn on. In this case printer knows the filament passed the gate and it will retract couple more cm and switch to next position. Then it tries to load. Same procedure...

Once you see what's failing, then we can help you to resolve it. Without knowing the root cause, it's just not easy to help.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram...
Posted : 18/06/2019 10:27 pm
mark.v18
(@mark-v18)
Active Member

😀 thanks.  No, no bearing noise.  Purrs like a kitten until it fails with either load or unload failure (no other failure types).  Yes, I understand how the printer works.  I bought an AnetA8 first about a year ago to make sure I understood how a 3d printer works.  I bought the MK3s as a kit.  The MMU2s comes as a kit.  I'm an IT guy with 30+ years experience.  If it was the F.IN.D.A. was too high or too low, wouldn't it fail sooner than 3 plus hours and 100+ tool changes?

Ok, as soon as I get done with my current single filament prints in a couple days, I will try again.  I will document here each failure and we can walk thru the failures together.

Please note, my goal is not to babysit the printer to complete a multi-material print (press center button untill filament engages and you get green red confirmation, then press right button to restart print or manually pull filament,then reload, then choose restart print from lcd menu).  I didn't pay $300 to become a babysitter.  I expect the multi-material prints to work like the single filament prints.  I.E. 24 hour plus more prints with zero interventions.

I work 60+ hours a week and have a full family schedule.  Why was I sold something that doesn't work like the base model?

I do appreciate some of the suggestions I've received.

Posted : 19/06/2019 3:36 am
jesse.l8 liked
gnat
 gnat
(@gnat)
Noble Member
Posted by: mark.v18

If it was the F.IN.D.A. was too high or too low, wouldn't it fail sooner than 3 plus hours and 100+ tool changes?

Maybe, maybe not. If it was right on the edge it could react to variations in the filament diameter or vibrations from the idler or printer moving. I would think that would be a pretty edge case though.

Please note, my goal is not to babysit the printer to complete a multi-material print (press center button untill filament engages and you get green red confirmation, then press right button to restart print or manually pull filament,then reload, then choose restart print from lcd menu).  I didn't pay $300 to become a babysitter.  I expect the multi-material prints to work like the single filament prints.  I.E. 24 hour plus more prints with zero interventions.

I work 60+ hours a week and have a full family schedule.  Why was I sold something that doesn't work like the base model?

I think this is an expectation of many (myself included), but as I've worked with it and gone through my own issues I think this is an unfair expectation. The reality is that the idea of multiple filaments from a single extruder is still a new thing and is more akin to where basic printers where 5-10 years ago.

Frankly the more I've learned and the more I've talked to users of other printers I realize that we are spoiled with how easy the MK3 is to use. Unfortunately that helps us set these unfair expectations.

So I don't think it's fair to expect absolute trouble free usage, but it is fair (and possible) to get through long prints with minimal intervention. I think it is fairly clear that while there can be a lot of pain getting the MMU working properly, once it does get working it does work pretty well.

It's just the getting it going that is the trick for many of us 😉 

MMU tips and troubleshooting
Stop SL1 tank leaks with a simple o-ring...
Posted : 19/06/2019 4:15 am
Nikolai
(@nikolai)
Noble Member
Posted by: mark.v18

...Purrs like a kitten until it fails with either load or unload failure (no other failure types)....

So where exactly is the filament once the error is reported? What is the status of the FINDA?

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram...
Posted : 19/06/2019 6:41 am
Whuzz
(@whuzz)
Active Member
Posted by: gnat

Frankly the more I've learned and the more I've talked to users of other printers I realize that we are spoiled with how easy the MK3 is to use. Unfortunately that helps us set these unfair expectations.

So I don't think it's fair to expect absolute trouble free usage, but it is fair (and possible) to get through long prints with minimal intervention. I think it is fairly clear that while there can be a lot of pain getting the MMU working properly, once it does get working it does work pretty well.

It's just the getting it going that is the trick for many of us 😉 

We were not "spoiled". The MK3 in general just operates as you might expect from a (quite expensive) product. We've come to trust Prusa for building a quality product that does what is advertised, without too many issues. Which is normal for most companies.

That is not a "spoiled" expectation, but a perfectly normal one. And that expectation is now broken by the MMU2(S), which in many cases does not operate without issues and in some cases fails to reasonably operate at all. There's nothing "spoiled" in finding that unacceptable. 

If the expectation should be that it's an experimental product that might not work properly, then it should have been advertised as such. It should clearly state on the product page that there are issues to be expected and that the experience will likely be vastly different from what you are used to with their other products. But that is not how it is advertised. Not on the product page, not on any video or in any blog about the MMU2(S). 

If the product page had said: 
"Will likely not be useable as a "fire and forget" printer. Expect manual intervention at nearly every print. Expect manual intervention every 20 minutes when printing solluble materials. Expect tweaking of hardware and settings for nearly every type of filament."

*Then* it wouldn't be fair to expect trouble free usage. I also wouldn't have bought it, if I'd known that. So there's that...

This post was modified 2 years ago by Whuzz
Posted : 19/06/2019 10:19 am
mark.v18
(@mark-v18)
Active Member
Posted by: nikolai.r
Posted by: mark.v18

...Purrs like a kitten until it fails with either load or unload failure (no other failure types)....

So where exactly is the filament once the error is reported? What is the status of the FINDA?

Boy, you all are a tough crowd (not you Gnat).  So I didn't write down every failure (any actually) during the 34 days before going back to single filament prints.  But again, as I said, "Ok, as soon as I get done with my current single filament prints in a couple days, I will try again.  I will document here each failure and we can walk thru the failures together.". But to answer your question it was either in the tube between the mmu and the extruder (unload failure) or on the tumbler bondtechs in the mmu (load failure).  I've already described what I did each time in previous post.  No, I don't know what was the status of F.IN.D.A

Nikolai-R if you are sure it is the F.I.N.D.A. what is the micrometer measurement for where the F.I.N.D.A. should set for trouble-free multi-hour multi-tool prints?  The handbook has suggestion but there's plenty of proof out there that there's no real tolerance or true setting.  The mechanical variance is enough that the product is faulty at best (IMO).  If it was a car, it would be back at the dealership due to lemon laws.  People in the US would expect reparation either from manufacturer or a court or someone.

Now, please note, I am still engaged and working on this.  I haven't given up.  But please read all the posts and provide constructive feedback if you don't mind.  Thanks.

Posted : 19/06/2019 11:47 am
gnat
 gnat
(@gnat)
Noble Member
Posted by: Whuzz

We were not "spoiled". The MK3 in general just operates as you might expect from a (quite expensive) product. We've come to trust Prusa for building a quality product that does what is advertised, without too many issues. Which is normal for most companies.

From what I can tell even the printers that are 3-5x more expensive than a pre-built MK3S are more fiddly (from print issues to hoops you have to jump through to get stuff to stick to the plate). Admit it or not, but the reality is that Prusa has set a high bar that even the big companies are struggling to meet.

It's certainly justified to have high expectations for the MK3S replacement, but to hold the same expectations on the SL1 or MMU is simply unfair at this time. When they've had the development and real world usage of the i3 I'm sure they will be every bit as reliable, but right now they are literally years behind the i3 development.

If the expectation should be that it's an experimental product that might not work properly, then it should have been advertised as such.

I do agree that they market it as more ready than it really is and I've said as much in other threads dedicated to such discussions. As such they haven't effectively managed expectations properly. I also do not think that frustration with them over the MMU performance is unwarranted. 

MMU tips and troubleshooting
Stop SL1 tank leaks with a simple o-ring...
Posted : 19/06/2019 3:15 pm
Nikolai
(@nikolai)
Noble Member
Posted by: mark.v18

Boy, you all are a tough crowd (not you Gnat). ....

Sorry if you feel that way. We are trying to help as we are all in the same boat (customers).

As you mentioned to have a lot of experience, I hoped for more detailed error description. This way we could help you better.
In regards to your question to the FINDA position, it's correct. There is no micrometer placement which is valid for all machines. Like the PINDA you have to find the sweet spot for your machine.

Please don't feel pressed to provide anything.If it's just another anger vent thread, I'm fine with that.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram...
Posted : 19/06/2019 5:18 pm
mark.v18
(@mark-v18)
Active Member
Posted by: nikolai.r

Please don't feel pressed to provide anything.If it's just another anger vent thread, I'm fine with that.

Thank you Nikolai.r!  No, this is not another vent thread.  I will post results as soon as I'm back to trying multi-material prints if they fail (there have been some suggestions that I plan on implementing first.)  Thanks again!

Posted : 19/06/2019 5:20 pm
Antimix
(@antimix)
Estimable Member

Mark,

the i3MK3S is an almost reliable unattended printer, as result of years of improvements. But it never managed large scale unattended unload and reload of the filaments. I think that adding this new functionality has been under-evaluated for its impact on the reliability.

I posted a long report of all possible issues I had  in another message, but after some days of thinking all can be re-conducted to two base causes:

  1. Filament mixture are different, by melting point and viscosity.  It is almost impossible to get a perfect conic shape on the filament when it is extracted by the melted mass. There are too many variable factors, that create as result blob shapes, and strings on the end of the extracted filament. 😫 The PrusaSlicer and even less, the firmware, can't have the perfect extraction.
    This must be accepted as a fact. Unless to have a difficult to obtain filament tuning parametrization for each installed filament, there is no certain on how the shape will be.
  2. Filament diameter are slide different and even PTFE tubes inside diameter are. You may not believe but my original PTFE buffer tubes have all slide different inside diameter (+/-). This generates friction with the filament, based on how the PTFE tube is flexed, and based on how much the filament inside is torquing itself.
    In this situation the filament is blocked, the MMU gear rotates and just grinds the filament creating powder. MMU Failure. 😞 
  3. The blade. This is not a reliable part. There are no mechanical warranties that it stays in its place.
    Based on the PET deformation and vibrations, after a while it moves and blocks the MMU.

The point (1) affect all the printer in general:

  • Blobs big head, jam the extruder when they are formed into the heath break metal (2.0mm diameter) and try to fit into the adjacent PTFE tube (1.85mm only). The diameter is too small to let pass the head.  The head blocks just under the PTFE tube and the filament is blocked.
    Sometime the head enter for 5 or 6 mm inside the PTFE and then blocks itself.

THERE IS SOMETHING WRONG WITH THE DESIGN
The diameter of the filament path is not constant. Imagine a gallery 2,50m high, where truck enters, and 500m later during the path the gallery suddenly become 1,80
. Probably you will have a jam of truck inside....
This is exactly what happens to the filament path in a Prusa. It starts 2.0mm, then become 1.85mm. JAAAAaaaammmmm 😊 
E3D had to do a 2.0mm metal diameter to ensure all the filaments could fits. Unfortunately the majority of the PTFE producer have 1.85 mm internal diameter +/- tolerances. I do not know why, the shipped PTFE were 1.85 instead of 2.0mm.

On my printer I decided to align the size. For this I replaced the original metal cooler PTFE tube (1.85) with a PTFE tube with 2.0 mm diameter (I take it from one of the buffer tubes that was out of tolerance, very large 🤗 ). Then I replace also the PTFE tube from Extruder to MMU with a 3.0mm inside diameter, because I noticed that a 2.0mm was not enough if the tube was flexed, so I used a 3.0 mm.

My latest filament jam showed me another weak point of the path. There is a hole just over the extruder gear, that is too small. My latest big ball jammed there. It was able to pass through the metal heath break, pass the inside PTFE, exit from the cooler, but not on this small hole in the extruder just after the gear. I probably will enlarge it.

 

On the point (2),  I simply replaced 1.85 mm PTFE buffer tubes, with 3mm inside diameter. I have almost reset the friction. 🙂 

On point (3) I have removed the blade for now. I will need to think to an easy open source solution for that (e.g. no custom built blades, too easy... 😎 )

About your FINDA: there was a crazy moment in my MMU. Nothing was working. I verified that the FINDA sensor was ALWAYS to 1 state, even dismounting it and hanging it in the air.... 😱 (yes there are no filaments strings in the air.... 😎 )

- FINDA issue ? Board port issue ? Firmware issue ? Who can tell ? The only way to go back to the normality was to turn the printer off and on.

Regards

Posted : 19/06/2019 9:31 pm
Nikolai
(@nikolai)
Noble Member
Posted by: Antimix

...FINDA issue ? Board port issue ? Firmware issue ? Who can tell ? The only way to go back to the normality was to turn the printer off and on.

In this case you can just reset the MMU unit only. It will sync again and you can proceed printing.

This post was modified 2 years ago by Nikolai
Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram...
Posted : 19/06/2019 9:43 pm
gnat
 gnat
(@gnat)
Noble Member
Posted by: nikolai.r
Posted by: Antimix

...FINDA issue ? Board port issue ? Firmware issue ? Who can tell ? The only way to go back to the normality was to turn the printer off and on.

In this case you can just reset the MMU unit only. It will sync again and you can proceed printing.

Sometimes that works. Other times, not so much.

MMU tips and troubleshooting
Stop SL1 tank leaks with a simple o-ring...
Posted : 19/06/2019 9:58 pm
Page 1 / 2
Share:

Please Login or Register