H2D AMS 2 Pro gets wedged if end of filament fails to detach from spool

Two scenarios, one of which (scenario 10 was a problem for me, and I wonder what will happen in the other scenario (scenario 2).

Scenario 1:

I have 3 identical spools of filament loaded into the AMS 2 Pro on my H2D. The 3 spools are on rotation, so that when one runs out, the next in the sequence is supposed to load. The print launches successfully and printing continues for a period of time drawing from the filament in slot 1. The two other spools of the same color and type are loaded in slots 2 and 3, waiting to be called upon when the filament in slot 1 runs out. Unfortunately, though, on this occasion when the filament at the end of the spool on slot 1 reaches its end, it doesn’t cleanly detach from the spool, as would normally be the case. I’ve had this occur on more than one occasion. Sometimes it’s because the end of the filament was taped to the spool by the manufacturer (generally with black tape). On this occasion, though, it simply didn’t pull loose from the spool at the end but instead was mechanically stuck to it. In either case, what happens next is the same: probably suspecting a tangle of some kind, the AMS 2 Pro rewinds the filament onto the spool all the way back to where the front-end of the filament nearly pulls out, but doesn’t. It’s in a state similar to if it had been freshly loaded. Then it tries feeding the filament back to the extruder where, no surprise, the end of the filament stays stuck to the spool and no further feeding is possible. So, it rewinds again, feeds again, rewinds again, feeds again, a number of times until it halts with the end of the filament under tension with the spool, and triggers the alarm.

What I wish it had done instead: rewind the remainder of the filament onto the spool and then move on to the next identical spool in the sequence. It was perfectly capable of doing this, and then the print could have continued without waiting for my intervention.

So, what do I find when I arrive? The system is gripping that filament hard. I can’t simply pull it out and replace the now mostly used-up spool with a fresh one.

What should I do? I’m not sure. When I arrived to rescue the unhappy printer, I wasn’t initially aware of what the problem was, so I cancelled the error and told it to continue. That lead it to repeat what happened prior, doing a bunch of feed and retractions, with the same “tangle”, and ultimately alarming again. It seemed to be stuck in a “mode,” and I didn’t see a way out. So, I punted: I cancelled the print and lost the work which had been completed up to that point. However, once the print was cancelled, the filament was no longer being tightly gripped. I pulled it out and tossed it in the trash. Then I loaded up a new fresh spool in slot 1 and restarted the print from the beginning. End of story.

Just a theory on what might have worked if I had tried it: I’m thinking that maybe if I had used cross-cutters and cut the filament loose from the spool after I had attempted to resume the print and it did a few cycles of feed, retract, feed, retract, then all should have gone well from there, since that is what would have happened anyway in the ideal case on its own, without me being there and cutting. Is this what I should have done, or was there some other solution?

Well, that was a lot of typing, so maybe sometime later I’ll describe scenario 2.

Somehow, last night I had a similar experience with the filament being stuck to the spool.

Since I am fairly new to Bambu and it’s AMS system, I never encountered any spools with black tape at the end. So in May last year (1st Bambu printer), I printed a set of cutter units that cut the filament when the spool comes to it’s end. I believe they were called “AMS saver”. I used them a couple of times until one time during a retraction a piece of PTFE tube was rolled back upon the spool causing other issues. So I never used them again.

Since then all my spools “with the narrow piece of white tape” released the last bit of filament automatically without any issues.

Until yesterday, when I ran into a spool of Black Charcoal PLA Matte that had the same narrow bit of white tape, but the end of the filament protruding approximately 5mm from the other end. This is probably the reason for this problem again, because all other spools of which I have seen, the filament did not protrude from underneath the tape on the other side.

You might have ran into the same situation as me here.

The only difference I see with your situation, is that the stuck filament at the end, caused a “AMS is not possible to feed ….”, or something similar error immediately after retracting the first time. It did not try to reload. Since I was present at that moment to wait for the spool to runout (forgot to add a second spool on forehand) and feed the new spool after the end of filament was detected. I can assure you that it just retracted once. So before selecting “Problem Resolved, Continue”, I simply pulled out the filament (used the release lever) and loaded a new spool in the same position. The job went on without any issue.

So in my case it seemed to have behaved differently, than in your case. Unless you are sure that it repeatedly loaded and re-winded the first time also. Or is this just an assumption because it did so the 2nd time after pressing “Continue”?

I can’t judge for your situation, but I would assume that unloading the filament from the AMS by using the release lever on the front, should have possibly allowed the printer to go on with the 2nd spool. I know it sometimes likes to load from the same location again, so loading a spool there, might have fixed it too. No way to tell what exactly would have happened though.

But I totally agree with you on what should have happened. It should just pick up the next spool and continue with that. no reason not to, unless the previous spool didn’t clear the filament sensor at the exit of the AMS. In that case, the 2nd spool would have gotten stuck.

To my opinion, this is a use case that is probably not anticipated so far, and has no solution yet.

Heh, totally agree, the “automatic” spool switching is useless when the first roll is taped down. Thanks to another thread here I was already aware of the taping issue and both times I’ve used the “auto” switch feature I’ve had to babysit the printer until the first spool is almost done and I can snip the filament in front of the tape with some cutters. Then the switchover works brilliantly.

The 2nd time was with a refill roll on which I had forgotten to check the tape first. On a couple other refills I’ve removed the (thin white) tape, which may or may not have been a mistake… guess I’ll see!

And I agree that the printer/AMS could have switched to the other roll earlier… though I guess if there’s an actual AMS feeding problem then the solution may not be as obvious, and the printer doesn’t really have a way to distinguish.

-Max

Removing the tape introduces the risk of the AMS not able to properly retract the filament onto the spool in case a filament switch is required. The feeding mechanism might still be able to retract the filament into the AMS, but there it would then not be rolled onto the spool but fill the entire AMS.
Result: other spools could get entangled by the floating filament.

2 Likes

This succinctly describes what I was going to call Scenario 2, which, from what I gather, became a real possibility when Bambulab switched from using black tape to tie the end of the filament to the spool without ever releasing it over to using white tape off to the side of the spool, such that it would release the filament when it reached the end of the spool. Right?

I am sure that the force required to hold the filament in place, is a thin line between:

  1. Holding it too tight, causing filament feed errors
  2. Holding it to loose to run into this scenario 2, as you call it

Though my experience with the detection of stuck filament and the procedure to pull back the filament sounds like it would be better that the filament stay stuck to the spool than when it would prematurely release from the spool.

However, the following unload and reload a new spool procedure should be adjusted to what was written into you first post.

I keep forgetting that the solution, at least to a Scenario 1, already exists and is the use of an automatic filament cutter positioned at each spool in the AMS. So, that would require 1 for an AMS-HT and 4 for an AMS Pro 2.

This will fix it permanently

https://makerworld.com/en/models/957400-ams-saver-snag-cutter-v2-0?from=search#profileId-2437367

1 Like

The end of the filament has always come loose from the black tape for me, which is what is supposed to happen. Apparently Bambu had a batch of spools where the tape was on too tight or something, but it has now been corrected. The filament should come loose from the tape at the end, leaving the tape on the roll.

Instead of cancelling your print completely, why didn’t you just cut the last bit of filament just before the tape, and then let the print continue? The printer would then have used up the end, and continued on the next spool…….

1 Like

TL;DR. Agreed. Or, using an automatic filament cutter would have been even better.

I printed those in may last year, but after it rolled back the PTFE tube back onto the spool and messing up my print, I took the off again had only one stuck spool last week. Never had any before.

Some time ago I loaded a Bambu Lab spool with PLA Matte. The tape on the top of the roll was black, so I babysat it. The tape on the end was white! Anybody else seen that?

Oh, no! I’m in the midst of printing the very same:

Is there a better design somewhere?

Anyone know whether the Infinity Flow would work better?

They should at least have a monetary incentive to do it right.

What was the inside diameter of the tubing you used? The project doesn’t specify, but I assume a minimum of 2.5mm would be preferred.

Also, and I don’t know whether it would have helped in your case, but you can chamfer the end of a PTFE tube to make feeding into it easier. A couple different jigs to make doing that easier and/or more precise:

I only know about this because some of the old-school Prusa hotends required it.

Well, anyway, I have 4 cutting blades arriving tomorrow, so I’ll likely assemble the four units then and if no issues either install them into the AMS 2 Pro then or else the next day. A number of the comments on the makerworld page have indicated encountering problems. If it fails, then hopefully there exists some other model that will work.

AFAIK, that is how they come now - chunk of black tape om the top and a sliver of white tape on the cardboard and my AMS has no issues pulling it thru and I do an auto refill basically every other day while printing.

I used pieces of Capricorn PTFE tube, that I had leftover from my CR6, but they have a 1.9mm inside diameter. They are very smooth though as they are designed for Bowden. I could retry them with the standard Bambulab tubes, that are 2.5mm inside. Not sure if I still have them or that I trashed them.

They are a hassle to use though. Given the one failure I had ever since, I might accept it as it is.

1 Like

That is what I noticed also, however the white tape at the end seems to have changed in shape and type.

Where the initial ones looked like insulation tape of approximately 12mm wide that was cut to a narrow piece of 4mm to hold the filament.

But the one I encountered now that gave the issue is 4mm wide overall, and are a woven type of tape.

Need to inspect some more rolls to verify that.

A variant of this scenario finally happened. 1. The end of the print was reached, 2. the end of the filament had detached from the spool (preventing future rewind), but 3. there was still unused filament remaining in the feedpath:

Therefore, throwing an error because it could not rewind the filament at the end of the print, as is its habbit:

I resolved it by “cancelling” the print (which was effectively done already, except for the final tidying up at the end). Then I de-mounted the PTFE tube from the exterior buffer port, pulled out the leftover filament that the machine was helpless to remove, re-connected the PTFE tube, and “Hey, Presto” back to a normal state.

Bottom Line: a bit of a hassle, but this particular scenario was no big deal to resolve.