AMS 2 - Spools and RFID Capabilities

I have quite a few spools of filament I’ve bought over the years that are not compatible with the AMS 2, mostly Atomic Filament. I’ve already printed, broke, and partially reprinted a filament winder and plan on printing up some spools in the near future so that I can use up that stock.

One of the considerations I have is whether I want to try some clip-on RFID tags. I definitely want to print some kind of swatch to go on the printed spool so I don’t have to remember if the spool is Overture, Atomic, Filamentium, PETG, PLA, etc.

I believe Prusa is working on developing an open source RFID standard for spools. Am I correct in assuming that Bambu does not currently support that standard?

I also understand that Bambu signs their RFID tags, so we can’t make arbitrary tags ourselves but I have seen that there are a few people that have worked out how to duplicate RFID tags and have a database. Assuming I buy the correct RFID tags, is it worthwhile to make an Orage/PETG tag, and print it with “Atomic / Neon Orange / PETG” on it to slap on the spool so I don’t forget what it is?

Does the AMS/Printer attempt to track filament remaining on a spool? If so, this seems like it would invalidate an RFID tag after the spool is used up.

I bought a couple spools of ABS to print the spools so I hopefully shouldn’t need to worry about deformation during drying. Very nervous about it… Last time I used ABS was just over a decade ago with a Simple Metal, and that was a truly awful experience.

Thanks!

Hi Henry, just a few thoughts from me…

BL definitely not supporting any open RFID standard at this point, and I haven’t seen any statement about it from them (not that I “follow” their news anywhere besides these forums.)

Hard printed labels look nicer, but painter’s or masking tape + Sharpie works in a pinch also… :wink: Easily removed and re-applied if drying temps are a concern. I could see how it would be convenient to have RFID/automatic detection though, to save having to tell the printer each time what is loaded. And potentially the remaining amount estimate could be handy. I haven’t tried making my own, so can’t speak for “return on investment” at all. Depends on how many and how often, I’m sure!

I’ve been saving the tags from used-up BL filament rolls though, and re-used one so far in an Overture spool (of the same material/color) that I was sticking onto a printed plastic spool anyway (the tear-off-the-sides-and-reuse-the-core method). That seems to work fine, technically anyway (still have to remember/track the actual filament of course).

Interesting point about each BL tag having a serial number.. The AMS just counts rotations for the printer to determine how much filament is left (which is sometimes comically wrong). And as mentioned re-using a tag seems to be no problem in that regard so far. But in theory BL could try to lock that down at some point I suppose… if they can work around the “comically wrong” scenarios which can show a spool going from almost empty to full (for example).

I think PC is a nice high-temperature alternative (I’ve only used BL PC so far which prints great on my H2C, though I understand not all “PC” is created equal). Avoids all the fumes of ABS/ASA, if nothing else. Plenty rigid enough for a spool. That’s what I’ve used. The translucent effects can be neat also. I’ve yet to see a light-up spool with LEDs inside, but it’s possible! Powered by the rotation of the spool, of course… :slight_smile:

Cheers,
-Max

It does, but it does not write to the tag, so worst case, just turn it off and on.

Thank you MaxThreeD and Krellboy.

Wow, the tracking of filament usage via counting spool rotations is laughably bad. The people behind Bambu Lab are definitely incredibly intelligent and talented people, but that is absurd. Every gram of extruded filament is right there in the gcode, and the printer knows how far into the gcode it is.

I’ve used SpoolManager in OctoPrint for 5 or more years now and it tracks my spool’s filament to within about 20g.

@krellboy - Power cycling the AMS/Printer resets the filament to full? Wow… that’s worse than Max indicating that it’s just bad at tracking it.

Still an amazing printer, and my only regret is that my ~4000 credits from buying the printer and accessories will be added to my account the day after the filament sale ends.

@MaxThreeD Does the Overture core fit in a bambu spool? Seems like it is a few mm larger by my measurements. I’m leaning towards printing replica bambu spools, and I’m sure I can print a shim.

To “full”? No, but since it’s not writing to the tag, it simply has to measure the rotation rate again from scratch, so swapping a different spool with a cloned tag won’t start with a wrong assumption. I have no idea how this will interact with filament manager, if that is what you are asking.

Be careful not to ever put an original and cloned reel in at the same time, at least in the original AMS that broken the anti-collision scheme.

Oh, so it’s using the fact than a spool with less filament turns more rapidly to guess at how much filament remains? Clever, and they can dial it in with their own spools, but it’ll fail for any company that uses different geometry. Sounds like it may have been done this way to encourage the use of Bambu spools maybe?

Ok, so if I get two spools of Red PLA in the AMS with the intent to run one out and autorefill with the second, I need to make sure they have different serial numbers in the RFID tag. Thanks for the heads up!

The OT PETG Translucent spool core fit right on a BL reusable spool (or maybe it was a leftover empty one… same thing), no shim needed, even the notch works perfectly. IDK if they use the same spools for all their filaments. It doesn’t have to be super snug, as long as the notch catches enough.

But yea the shims are trivial to make as well if needed. I’ve only printed a few high-temp spools so far since the low-temp reusable BL ones are already so cheap as an add-on, besides already having leftover empty ones of course.

so it’s using the fact than a spool with less filament turns more rapidly to guess at how much filament remains?

Exactly. So it needs at least a few rotations after swapping in new filament or a system restart. And when it gets towards the end of a spool the filament gets all loose on the core, the printer thinks it is full again. :slight_smile:

it’ll fail for any company that uses different geometry

Well it’s a rough/crude estimate anyway, but yea translating relative movement to absolute amount needs some assumed/fixed constants. In reality most ~1kg spools that will fit/work in an BL AMS would be roughly the same anyway.

-Max

That would require tracking the serial number of a spool.

If you then put that spool in a different printer, then how?

Bambu is pushing cloud based connectivity, no reason that usage can’t be stored there.

Do you really want more tracking? I find the current measure of spools good enough. If I’m ever cutting it close, I add a backup roll.

As far as I know, it doesn’t just “count rotations”. It uses odometry, a sensor measures the length of filament fed through the system and compares that value to spool rotations to calculate how much is left on the spool. The estimate is usually reasonably good, certainly not “laughably bad”.

(edit)

I threw a couple of new spools of PETG in the AMS the other day and printed a couple of little filament clips from one of them. Today I noticed in the filament manager that all the unused spools show 1000 grams except for the one I printed from, which reads 990 grams and you can’t tell by looking that it had been used. Pretty accurate, I’d say.

Thats a terrible idea, what if the filament is used in another printer, or you drop it and get a tanglew and have to remove some, or you have a power outage? Or hundreds of other reasons.

The spool rotaion tracking is reasonably accurate (I would say ±50grams), and its consistent as it doesnt depend on inaccurate spool history.

The problem of storage is maintaining an accurate count and providing some method of manually adjusting it. There are many ways to use filament without printing - for example, in the case of a filament tangle, or the many printable storage boxes use a length of “raw” filament as a lid hinge. Personally I would do all of this through the Handy app to prevent having customer service bombarded with requests to change some number on some spool.

But ultimately, it’s a lot of effort and rolling the spool works well enough, while being much simpler.

I put a purchased from Overture this past week refill in a Bambu Spool and it fit like it was made for it. Not saying they all will but I only purchased one refill as a test and will now be purchasing more. That being said Marswork spools are much nicer than BL’s and also accept BL refills so will assume Overture as well so any future spooled filament I buy will be Marswork as I really like their spool design.

Looks intersting. says “designed around bambu users” so i bet their spools do fit,

my question : are their colours the same?

Refills definitely fit, not sure about the colors as I’ve only used their sunflower yellow, a color that seems to be eternally out of stock on BL and only comes in refills. If there’s any correlation the colors do pretty closely match Sunlu’s colors as I’ve mixed their sunflower yellow with Marswork’s and have to really scrutinize to see the difference. My experience with Marswork started out a little rocky as they claim their filament is pre-dried so I tossed it on and started printing, quality was so so. I then dried it and print quality is on par with BL filament, if you do buy Marswork don’t buy into the pre-dried bs…

@piefairy I absolutely do NOT want more online tracking. Local? Sure, I’d love it if the printer could track my filament usage, just like my Prusa/Octoprint setup does, and RFID capabilities would greatly improve on my former setup.

I was just answering someone’s question.

@DMeader “The estimate is usually reasonably good, certainly not “laughably bad”

You’re absolutely right, the idea I had (and too quickly assumed Bambu used) was laughably bad.

@erikt1 and @Eponymous I appreciate your insight and the reasoning as to why Bambu may have chosen the method they did.