PSA: Don't mess with it once you get it working  

RE: PSA: Don't mess with it once you get it working

I recently had to move my setup from one room to another. Was shocked how everything fell apart. Took me almost week to start getting prints again. Only thing I can think is maybe the humidity and temps changed enough to through it off.

RE: PSA: Don't mess with it once you get it working

Ugh. So with the auto-rewind spools not working I figured I'd give the buffer a try by hanging it vertically from the back of my 2020 frame. Rather than getting through half a Benchy it now won't do a single filament change without help. It's issue is that it isn't unloading into the buffer correctly so that when it gives up and demands attention there is still about 6" of filament in the MMU -> Extruder PTFE.

Even beyond the tight quarters, loading from the buffer to the MMU was a pain as there was clearly a lot of drag (so I didn't have high hopes to begin with). I've ordered more Capricorn XS PTFE as filament does seem to slide much nicer through the piece I already have, but I need to look and see if there is a festo mod out there for the buffer too as that would improve things too.

RE: PSA: Don't mess with it once you get it working

Posted by: gnat

but I need to look and see if there is a festo mod out there for the buffer too as that would improve things too.

You might want to check this one out. I haven't used it as I'm still trying to find the best solution for my Repbox.

https://www.thingiverse.com/thing:3545935

RE: PSA: Don't mess with it once you get it working

Yeah just found and collected that. Festos and tube should be here tomorrow so I'll get them (and a filament guide stick) run off over night.

I think I'm even more frustrated over this now than I was initially since I know it can work...

RE: PSA: Don't mess with it once you get it working

Posted by: gnat

.... It's issue is that it isn't unloading into the buffer correctly so that when it gives up and demands attention there is still about 6" of filament in the MMU -> Extruder PTFE.

Hi Gnat,

as I described in details here also your issues seems related to the PTFE tube friction. Solve the friction and you will solve your issues.
I made several test with the MMU, and I verified that a slide move of one PTFE tube, even a simple flex, causes a change of the friction with the filament, and if this is high, the MMU is not capable to move the filament. You will see that the filament is grind by the gear or shifts (and that explain the last 6" shifted out...)

I also normally avoid to make changes to something working, 😊 but I like the stability and repeatability. So I do not feel comfortable with something that I know it works for chance. I want to be sure to be able to replicate the setup in any moment, and this means to have defined calibration center lines, and positions, and setup. And overall, very important, is to understand why it works (and otherwise , why it not works). 😎 

RE: PSA: Don't mess with it once you get it working

Posted by: Antimix

Posted by: gnat

.... It's issue is that it isn't unloading into the buffer correctly so that when it gives up and demands attention there is still about 6" of filament in the MMU -> Extruder PTFE.

Hi Gnat,

as I described in details here also your issues seems related to the PTFE tube friction. Solve the friction and you will solve your issues.

If you read my post history you'll see that I'm well familiar with the friction issue 😉 

I do know that is the root of my issues here, I solved/mitigated it originally with the auto-rewind spools going directly to the MMU via the short (spool side of the buffer) PTFE tubes. Unfortunately moving the spools above the printer is not letting that solution work and so far my alternate configuration attempts have not provided good results.

Just cause you know the root cause doesn't mean it's easy to solve 😆 

I also normally avoid to make changes to something working,

Yeah and I wouldn't have except that the printer hung off the edge of the desk that it is on and was prone to getting bumped as you walked in or out of the room ☹ 

RE: PSA: Don't mess with it once you get it working

So I had a spare 1m of the PTFE already as well as 5 spare festos so I cut the tube in half and hooked up 2 filaments through the buffer. I'll know for sure in about an hour when the first tool change happens, but I have reason to hope. 

First off is that the festo mod has two immediately noticable improvements. There is now no friction at the entries and exits. It's a nice smooth flow even though I'm still using the white short tubes on the input side. The second improvement is that it is SOOOO much easier to load the filament through the buffer and to the MMU. 

The second thing that gives me hope is that the filament slides nice and smooth through the Capricorn XS tube. 

Hopefully this pans out...

RE: PSA: Don't mess with it once you get it working

Ugh. No joy. First unload failed with most out the filament still in the MMU -> extruder tube then lots of hand holding to get the second filament to load, more for the unload, and then just gave up when the incoming filament failed. 

Its acting like it did last night, but without the resistance in the tubes. So I think I was blaming the tubes for more than they were doing. 

What I saw last night and thought was friction was that the idler was having difficulty moving the filament and I'm seeing that again tonight. What happens is that it move the filament a bit and then the filament stops even though the idler is still engaged. If I nudge the filament in the correct direction it will grab it and move it forward some more. Then rinse and repeat. When the filament is not moving, however, I cannot move the filament in the opposite direction than it should be going. So there is obviously some bite. 

No grinding sounds, however, and no signs of any chewed areas. 

My first thought is that maybe the gears are loose, but if that were the case I would expect to be able to reverse the filament without issue. Also, this seems to be consistent for all filament paths which would be a massive coincidence if all 5 came loose the same amount at the same time. 

My next thought is that the idler motor isn't working properly, but sounds like it is spinning the whole time at a consistent speed.

I'll have to fiddle with things tomorrow so I can actually see into the MMU to see what the gears are doing during a failure and then go from there. 

RE: PSA: Don't mess with it once you get it working

this is why i'm not even thinking about a mmu2S update. mine is working so i'm leaving it alone.

RE: PSA: Don't mess with it once you get it working

"Never touch a running system" does also apply to 3d printer  😉 

RE: PSA: Don't mess with it once you get it working

as does " if it ain't broke, don't fix it"

RE: PSA: Don't mess with it once you get it working

Don't know if it was the modifications I did today or me sitting by the printer, but I got through a 1.5 hour print with 27 of 27 unassisted tool changes!

That is also my first successful use of MeshMixer to break up a single part! I put it on TV if anyone else is interested in it: https://www.thingiverse.com/thing:3706412

It's now trying a 5 color Benchy and I'm not staying down there for the next 10 hours, so we'll see what happens...

My current setup is:

• Simple spool holders with bearings held above the printer by a 2020 extrusion frame.
• Capricorn XS PTFE short tubes about the same length of short tubes that came with the MMU
• A PTFE holder (didn't save the TV link 🙄) is used to help keep the tubes in an identifiable order.
• Into and back out of festos on the factory buffer using https://www.thingiverse.com/thing:3545935
• The buffer is held vertical on the frame with https://www.thingiverse.com/thing:3703814
• The long PTFE tubes are roughly 0.5m and also Capricorn XS PTFE.
• Two of the aforementioned PTFE holders are used to stabilize and organize the tubes.
• The long tubes enter the MMU via festos via https://www.thingiverse.com/thing:3233579
• I also went ahead and replaced the MMU to extruder PTFE with more Capricorn XS PTFE sized to match the original white tube.

I meant to update the firmware before I started tonight, but forgot. Soooo not doing it if this Benchy prints!!! 🤣 

RE: PSA: Don't mess with it once you get it working

Good work.

Since changing my heatbreak my MMU has been working like a champ, but it has to move. Not looking forward to that day.

RE: PSA: Don't mess with it once you get it working

Well I think I'm ready to call it happy again. PS says there were 398 tool changes and only 2 failed. It got about 9.5 hours in and failed on an unload. I suspected a string making things unhappy, but couldn't find anything. I watched it for a few tool changes, but a few more after I walked away and it had another unload failure. This time, however, I did find the string (was in the MMU -> Extruder PTFE) and removed it. It then finished the last 2 hours without incident.

While I would love 100%, a 0.5% failure rate out of those two prints is good enough for me I think.

RE: PSA: Don't mess with it once you get it working

We need an alternative to the teeth gears in the mmu2. Something soft, but grabs like a mofo and doesn't chew up plastic. Like some rubber rollers or something.

RE: PSA: Don't mess with it once you get it working

I had a working MK3 + MMU ... until I decided to do the MK3S/MMUS upgrade. While the reassembly was straight-forward, I have been defeated by trying to calibrate the new IR sensor chimney. The IR sensor works, but any time I tighten the screws for the chimney, it pulls the sensor out of alignment and the door arm stops tripping the sensor. I don't see how this is better than the Mk3S / Mk2.5S sensor arm (without chimney) which works fine for my Mk2.5S.

RE: PSA: Don't mess with it once you get it working

Posted by: kieran.s4

We need an alternative to the teeth gears in the mmu2. Something soft, but grabs like a mofo and doesn't chew up plastic. Like some rubber rollers or something.

I don't know if this is of any idea, but I actually know of a method for an extruder that doesn't use teeth gears. Believe it or not, the GT2 belts can do the same thing as a toothed gear. At my university (virginia tech) one of the students did this to an Ultimaker 2 and the thing produced flawless prints at even the finest layer height. Apparently Microsoft wanted to patent the design but that might have just been a rumor.

Here is the link for anyone who is interested: https://0x7d.com/2014/improved-ultimaker-2-material-extruder/

RE: PSA: Don't mess with it once you get it working

Posted by: richard.l

Posted by: gnat

but I need to look and see if there is a festo mod out there for the buffer too as that would improve things too.

You might want to check this one out. I haven't used it as I'm still trying to find the best solution for my Repbox.

https://www.thingiverse.com/thing:3545935

 

I use the V2 version of that part on both the input and output sides of the filament buffer.  I also upgraded to 6mm PTFE tubing on the input side, which eliminated all the friction issues with my small footprint system. 

I posted an adapter and some photos here: https://www.thingiverse.com/thing:3614012

RE: PSA: Don't mess with it once you get it working

Those of you who visit the forum a lot will see I am putting this link in many threads, but it is to help people who only open one thread and do not use the search feature.  So...here is a collection of MMU2s upgrades on Thingiverse that I have collected so far.  I love the rewinding spool for the Rubbermaid 21-Cup containers.  He also has a larger version of it.

https://www.thingiverse.com/cybrsage/collections/mmu2-mods