Printing from an external spool (mandatory for TPU) means that the extruder needs to pull on the filament in order to get it. Unfortunately, that automatically produces under-extrusion (i.e., the extruder doesn’t get all the filament it needs). The under-extrusion happens not only with the softer TPU’s (like NINJAFLEX, which is an 85A), but also with 95A filament such as the excellent Inland TPU (surprising considering the low price).
The ‘solution’ of standing by the printer and feeding it slack filament by hand grows old very fast, specially because the secret to even being able to print at all in TPU is to print very slow, and spending hours feeding filament every few minutes sucks.The alternative of reducing friction, gravity feed, etc., all end up failing in my case: even using my EIBOS filament dryer–which has very low friction rollers), mostly because the entrance to the PFTE tube goes from the bottom up, and ends up stretching the filament just by having to deal with the weight of a a semi-full spool.
Since several of you folks have reported printing TPU on the Mini, would you mind sharing what your approach to this problem is? Has anybody seen (or heard of) a quick make-able device that could do the feeding automatically by monitoring the slack on the filament?
With TPU becoming more and more of a “first-class” filament, with prices dropping fast, a large variety of options, and the large number of usage scenarios TPU has, it seems deplorable that the Mini cannot actually support it well.
I don’t have a mini but I print TPU on my P1P and it has to pull through 2ft of PTFE tube and from my dryer box which sits next to it. I would think the A1 would have no problem pulling from a spool mounted above or from the side.
I’m saying that, the times where I thought I was getting under extrusion turned out to be the nozzle just not being hot enough, with 240°c the minimum it likes. This is for 95A
I use the crealities for tpu but ive heard people had good luck by loosening the extruder tension on bambus. They seem to come maxed out. Probably didnt find out that the tension spring wasnt strong enough so just tightened it all the way (for other filaments)
@StreetSports , thanks for showing me where the tension screw is located. I tried it with various settings, and unfortunately made no difference.
All I learned so far is that it is definitely a problem with the filament moving through whatever lies between the spool and the extruder. Friction, plus the weight of an almost full spool, conspire against the traction of the extruder and the flex of the material.
I made a couple of experiments with bearings, and I think I have found a working idea, I’ll give an update once I get it to work.
You could always turn it into a fun project and build a powered unwinder. May be tricky to get the speed right. You could probably just use the respooler everybody prints. The one that has the option of using a motor. The motor is cheap, very low kv(high torque) and has its own gear box. That on a super low speed setting. Possibilities.
Or a bearing mounted spool holder on a shelf directly above the a1. Take the 4 to 1 adapter out and run a single ptfe tube(if possible. I’m not at home to look)
I love both ideas. I am already thinking about the second one, need to get some free time to execute it (just a way to have the inside rim of the spool sit on two bearings instead of the spool holder axis).
For the first idea, yeah, it would be a fun project. Synchronizing speeds is probably out of my league, perhaps I can manage it by “seeing” the slack on the filament between the spool and the entrance to the PFTE tube and use it as a switch to turn off the motor…
I have tried with inland (95A), Bambu Lab (95A), NINJAFLEX (85A). Best results with Inland, followed by BL, and far behind NINJAFLEX (Very soft). I have tried standard settings, special settings recommended elsewhere in this forum, and my own tweaks. Except for going very slow and very hot (basic idea), all of them fail at some area. Definitely not a PLA printing job…
Bingo! That’s what I had in mind (only two bearings better than four from friction perspective). I assume it will fit on the A1, if not I can remix. Thanks for the reference!
EDIT 12/25 - @maruel I ended up designing yet another spool holder, which replaces the spindle in the standard spool holder. Very simple, but as you said, “ball bearings make all the difference” (I don’t know about ballS bearings, though ). About the difference: see this video
The problem is not the spring but the leaver! but the leaver itself has a geometry that doesn’t touch the bearing with the teeth of the extruder I have printed with success Filaflex TPE 90shoreA and TPU 85shore A i am testing right now TPU from Filaflex 70shoreA
I have made this hack to print PET recycled from recreator and suspect that could help flexibles !
But be careful, I used bearings on the holder of my Anet A8 back in the beginning times for hard filaments - with TPU it worked at the beginning, than there grew a slope between the spool and the printer. After some minutes, the slope was heavy enough and the complete spool unspooled itself on the floor…
I ended up with creating a little “brake”. So the tension was the same, regardles of the weight of the spool.
I finished to print 4 wells for a toy of my boy with Filaflex 70shore A, and printed with 70ShoreA of course is not trouble free but with some patience it works.
I have printed a lot with this mod without problems for now, but more people need to test it out!
@eingemaischt thanks for the positive feedback! I have been traveling out of reach of the printer for a few days, but before I left I printed continuously on TPU, unattended and without flaws. I wish other people saw how great it works, many would adopt it regardless of filament.
That said, two additional comments:
Shore 85 is still a problem (which I plan to fix using @ricaxe ‘s modification to the lever quoted elsewhere on this thread
Doing this mod made me realize that the spool holder ARM (not the spindle) on the Mini is super-flimsy: it slightly bends down when I place a full roll of filament on it. No functional impact, but still looks like it could experience damage at some point. For that, I plan to relocate the spool arm to the cage I built for the printer using 2020 profiles, which I will also use to move the AMS Lite. Tasks for later this week (home tonight)