Feeding to right nozzle fitted with TPU High Flow kit using 0.4 High Flow nozzle for TPU. The toolhead enhanced cooling fan upgrade is installed. Factory right PTFE tube left in place but disconnected from buffer and inserted through provided TPU inlet port protruding approximately 60mm outside. Bambu TPU 90A filament in AMS HT feeding from under spool into 50mm tube in TPU port in cover as recommended. AMS HT is elevated and is immediately adjacent to the left rear corner of the printer for relatively horizontal unsupported filament (no tube) feed parallel (left to right) with the rear face of the printer. Unsupported gap between the AMS HT short tube and the printer tube protruding from the TPU inlet is approximately 25cm. Filament manually loaded into right extruder. PTFE tube replaced into extruder port when flow established.
Default process parameters and 0.16 layer height. During good looking first layer an extrusion resistance error is seen. The filament is NOT tight but sagging loose in the tubing gap between the AMS HT and the PTFE tube. Clearing the error and resuming the ongoing layer looks good for a few moments but then the error returns. I am hoping someone has some suggestions to print Bambu’s TPU 90A successfully on my H2D. Thanks.
If you remove the PTFE tube from the print head and feed the TPU thru it manually and then try to pull it, how does it feel?
With TPU90A I’ve made the experience, that it sort of get caught by the entry edge of the PTFE tube. You might check whether it is sort of sharp and add a little inside chamfer to it. Or using a PTFE tube that reaches from AMS HT to the print head but that might be too long already. I haven’t had much luck printing TPU 90A without feeding it to the print head directly without any tube at all.
Not to say this is the issue here specifically, but re-using the normal right side PTFE tube has potential issues:
It’s too short once run outside the printer. (Some say it’s already a bit too short anyway.) This can cause sharp(er) angles in certain head positions, and binding.
Other filament will wear down the tube quicker and possibly leave dust or other residue behind. Of course in that case it might be good to replace the tube anyway, but soft TPU is certainly going to be more sensitive to those conditions.
The unsupported filament entering the tube may be rubbing on the edges, as Daniel suggested, and/or the sag causes extra drag since it has to be “corrected” to get into the tube properly.
Personally (on H2C) I run a new tube from outside all the way to the head, feeding it through the existing clips and a new clip on the end to keep it out of the way when not used. This is long enough to smoothly and comfortably connect to my TPU dry box feeders directly – with around 60cmoutside the printer, and length hasn’t been an issue at all (smooth is king). Switching to TPU, I pull the normal (gray) tube out of the head, park it in that extra tube holder, and stick the TPU tube into the head instead.
Thanks very much Daniel and ‘Max’ for your suggestions. I am thrilled that I’ve actually got some responses. First, I’ll respond to Daniel: I gently pulled on the manually fed filament when I loaded it and there was some resistance. Material flowed freely from the nozzle and was laying down a nice first layer prior to failure. When the head retracted in reaction to the error message there was stringing to the home position, but I imagine that was due to fast retraction of the head responding to the error. I thought about edge catching and chamfered the tube end and also verified tactilely such wasn’t happening and the un-tubed filament was free to move. I see that you also haven’t had much luck with tube fed TPU to the head. It seems that the intent of the TPU inlet was a well-considered means provided by Bambu that should work, at least for 90A and above as they approve. I have the online impression that it does and should. I thought that I had stacked the deck in my favor with a virgin high flow TPU nozzle and the careful physical arrangement of things, only to be frustrated. I’ve been thinking I am the subject of bad luck, that something else is not right. On the other hand, Bambu now offers an external feed assist device, but I imagine that must be for even softer filaments. Max: My H2D is virtually new, the factory tube is pristine. I thought that long tubing would be an issue, but it seems you’ve proven otherwise. How soft has the filament been that you’ve managed to feed with your long tube method? Perhaps I should try it. What tubing brand/supplier do you recommend? Have either of you considered the feed assist? Thanks again.
I use my AMS HT as an external spool system, feeding TPU/PEBA through PTFE tubes from the AMS all the way to the toolhead. So far, it has worked without any issues.
That said, I had to disassemble and adjust the AMS so that the spool could rotate with as little resistance as possible. I also replaced the original PTFE tubing with PTFE Tough. The setup worked both before and after I started using TPU High Flow.
Even the slightest resistance can stretch the TPU, resulting in inconsistent feeding and various print problems. For that reason, I would not recommend copying my setup unless you are comfortable fine-tuning it and understand how sensitive flexible filaments can be.
Moisture can also increase resistance. If the TPU is not dry enough, it may become slightly tacky and stick inside the PTFE tube, which can cause feeding problems. In my setup, I keep the AMS HT in drying mode while printing. This keeps the TPU/PEBA very dry throughout the entire print.
Thanks ‘L0rd’. I assume to achieve success, even with 95A filament, you had to pay extra attention to things like you described? I suppose that the folks who just use this stuff out-of-the-box issue free are lucky, rather than in my case and perhaps yours, being unlucky, where the combination of equipment we got sent and/or a setting just wasn’t quite right? I ran the 18 hour drying cycle on the 90A in the AMS HT prior to printing. I thought that running drying mode during printing would make the filament softer, further negatively affecting feed reliability. Furthermore, it was not obvious to me how to do that, as I understood and experienced that the HT just winds up becoming a ‘dumb’ external spool holder. You wrote that doing so keeps the filament dry during the printing. I did not know that TPU absorbs water that readily. It seems to me that once dry, the decently sealed HT, with a very small volume of internally captive and necessarily dry air after the drying cycle, would have little chance to admit moisture affecting the filament.
TPU 87A worked (haven’t tried 85A yet). Though I mostly use 90A and some 95A. And some foaming PEBA, IIRC that’s something like 90A before foaming.
What tubing brand/supplier do you recommend?
Just whatever got good reviews on Amazon and seller seemed legit (as in might still be around tomorrow). The dimensions are important: 2.5mm ID x 4mm OD.
Have either of you considered the feed assist?
Have one… tried it a couple times. TBH I didn’t see any real benefit to it with my setup. One minor convenience is that it feeds the filament to the head for you… but so far that’s not worth introducing the added complexity. I understand it might help maintain steady feed to the head leading to more consistent prints, or some such thing, but my TPU prints already came out great and the assister didn’t improve anything…
I thought that running drying mode during printing would make the filament softer
I dry TPU first. I’m glad it works for L0rd but I would never heat it up during use, which just seems like asking for trouble. And yes, the HT is essentially a “dumb dry box” in this situation… I don’t use my HT for that because the outlet is in an awkward position for my setup and it’s actually easier to just keep the filament in the same drybox that also feeds it (granted that might not be practical if one has many soft filaments).
Yes, continuous drying does make the filament warmer and therefore slightly softer, but in my experience, the benefit far outweighs that: dry TPU/PEBA is noticeably less tacky.
If I stop the drying function during a print, I occasionally experience the TPU sticking inside the PTFE tube. With drying enabled throughout the entire print, I have had zero feeding issues. This becomes even more important with very soft materials, such as 65A TPU/PEBA, where the tackiness is considerably worse. Even with those materials, continuous drying has made feeding completely reliable for me.
I do not believe this is simply luck, since the same setup works consistently on both my H2D and my X1 Carbon. Although the AMS HT effectively becomes an external spool holder from the printer’s perspective, its drying function can still remain active during printing.
I cannot say with certainty whether the improvement comes entirely from preventing moisture absorption or partly from the warm, dry environment reducing surface tackiness. I can only say that the difference has been clear and consistently repeatable in my setup: pre-drying alone occasionally causes problems, while continuous drying does not.
Thanks ‘L0rd’. I appreciate the details and insight for what I felt was a counter intuitive method. Were you doing that with the glass on top of the printer removed? Thanks to everyone on this posting I have some additional and informed ideas how to proceed. Two main ideas are first to keep the setup as described but using a long uninterrupted tube. For research sake I’ll run it first w/o heat. Not sure I’ll disassemble the HT to free up spool rotation. I’d appreciate some details on that please. Mine seemed reasonable. Not sure if I’ll get the feed assist to have it on hand as an option. The second is to mount the spool (or HT) over the open top for direct feed into the head, which I’d rather not and feel I shouldn’t have to do, at least for approved TPUs such as 85A plus. BTW, I didn’t mention that the Bambu TPU 90A I was attempting to use is Chrystal Blue (51601). I didn’t realize it is a translucent formulation, which may be a factor in how it feeds. I haven’t been able to return to rerunning my print. It appears that it will be some time before I do. In the meantime, other than reading your hopefully coming reply and other’s for which I get emailed notices for, I’ll return to this thread when I have my findings from the next attempt(s).
I keep the glass top installed, with my AMS HT units sitting on top of it. They connect to the bypass port through a 4-in-1 splitter, and I use similar connections for the left and right toolheads during normal printing.
I keep the PTFE tubes between the AMS HT units and the bypass port as short as practically possible to minimize drag and reduce the risk of feeding problems. I have never used Bambu Lab’s recommended arrangement with the glass removed and the filament feeding directly through the open top. That setup is probably more fail-safe, but it is also considerably less convenient if the glass has to be removed every time.
I have disassembled my AMS HT units and lubricated all the parts that rotate or roll using a dry PTFE spray. That is the only modification I have made to reduce spool and feeding resistance.
I have never used Feed Assist and currently see no need for it either. In fact, I am already getting prints that look better than the “after” examples Bambu Lab uses in its Feed Assist advertising.
I should add that I live very far north, so high ambient temperatures are rarely an issue here. In a warmer climate, removing the glass may be necessary, especially on a printer such as the X1 Carbon, which is less effective at dissipating excess chamber heat.
