Reducing waste from Bambu Lab AMS (lite) prints

[@AndySmallfry]

Theres a model on printables called "Bambulab Profile for up to 60% purge reduction" and it does exactly that. Pretty much it moves the filament up close to the cold end before cutting it. This reduces the physical amount the needs to be extruded.

[@dustinbailey1980]

Is there a way to save it? Do I need to set the multiplier and auto calculate on every print or does it keep the change

[@AndySmallfry]

@@dustinbailey1980 for the profile it is a G-Code edit that changes the amount of purge and as long as you save if the profile you'll be fine. But for the auto cal, that's usually done per print.

[@keoghanwhimsically2268]

@@dustinbailey1980Yes, it’s a nozzle profile that you can save as a user profile. Once saved as a user profile you can select it from any print for that user.

[@TROPtastic]

​@@dustinbailey1980 For the Profile mentioned, there's some G-code that will be there for every run. Recalculating will be needed if you change filament combos.

[@rs3dpt]

I did this and it works very well! It's insane the amount of filament saved, plus i'm using only 0,3 on flushing volumes, and the purge tower i give it 10mm width and 10 purge amount

[@woodwaker1]

I like the way you visually showed the results in the bins. I like the AMS, but I try to print with less than 20 color changes and fill the build plate with as many copies as possible. I think your results could be even better if you ran some tests to get the color swap amounts down to the bare minimum.

[@boboscurse4130]

The AMS waste has had a positive effect on how I design. It's a real consideration.

[@maxhammick948]

Would it be possible to add a retraction to the colour change G code just before the filament is cut? If you can pull the blob up until it's just below the blade, you're still cutting off the blob and any stringy bits but you're also saving a lot of filament. I am assuming that the filament path between the knife and hotend is smooth enough that the blob can't get caught on anything, of course.

[@boggisthecat]

See the top-rated comment.

[@hagus42]

You can also purge into an object. So if you have a tool or something where you don’t care about color banding, that can be an option instead of duplicate models

[@jippenfaddoul]

If only the video covered that around the 4:40 mark and showed that the results were actually not that great...

[@poselennov]

​@@jippenfaddoul @hagus42 said about purging in technical model instead of wipe tower, not "flush into infill", which was really to that great for such a small model as Marlin.

[@user-mz6qu3hz6m]

@@poselennov exactly correct. And because it’s an entirely separate object, one can increase the perimeters and.or infill on that object only.

[@TheBullBelgium]

add a object,can be the same model. right click on it in prepare screen,flush options- flush into this object

[@TimSheehan]

​@@jippenfaddoulno the video covered purging into infil of the multicoloured model, not purging into a separate colour-doesnt-matter model you're printing at the same time

[@sonofbrun6033]

Layer height also makes a big difference. Fewer layers means fewer filament changes.

[@nickoli2356]

I was about to say this 50% and a .28 layer height should make a world of difference

[@K0BRAKID]

​@@nickoli2356is 0.28 the highest a layer can be?

[@timbervandenhul9383]

​@@K0BRAKID you generally don't want to go beyond 80% of your nozzle size. So for a .4mm nozzle the max layer height would be .32mm. For a .6mm nozzle the max layer height would be .48mm.

[@ChrisUG]

Orientation can also massively affect tool changes

[@IdeationGeek]

Exactly, orient object to maximize the mono-color layers, so that no filament change be needed to print that layer.

[@CrudelyMade]

Filament could be saved if they had a slicer feature that could do variable infill or use the waste material on a second model where colors don't matter. But a model with 10% infill has plenty of room to use up some of that otherwise wasted filament.

[@jjptech]

Thinking exactly the same here

[@meesterjochem1]

This is possible. Add a second object and with right mouse button choose flush options --> into this object

[@ultramegax]

You can do a second model with the infill...

[@CrudelyMade]

@ultramegax , exactly I'd take a second model, with whatever colors over a pile of refuse filament.

[@CrudelyMade]

@@meesterjochem1 that's good to know. I'd still like to see a 'variable infill with refuse material' feature that would add extra infill up to 100% on the current model. i.e I choose 10% infill to save time and filament. but if i'm not going to save time, and the same filament can make the current model an 85% infill with the refuse filament, I'd feel better about that than a pile of trash.

[@roysigurdkarlsbakk3842]

Just one thing… If you use flushing volumes / autocalc, IMHO that works well, also with at factor as low as 0,3, but I guess that depends on the model. But, if you later change filaments, which you generally do a lot with a 3d printer and forget to go by flushing volumes to update the calculation, note that that will not update automatically and you'll end up with a flushing table written for a completely other world than the one you're in and if you, perchance, start a print at night and go to bed and wake up the next morning to find your flashy new print, it'll look like rubbish. Beleive me on this ;)

[@5mmTech]

For that particular model, I'd be curious to see you shift it to a vertical orientation with supports and see how that impacts the volumes. By having the color changes oriented to the direction the model is sliced, I'd imagine significantly fewer material changes and purges. Obviously, you would now have material in supports, but it would still be interesting to see the impact.

[@JH-zo5gk]

I bet the tool change count would go up. Figure those changes only happen once per layer.... so by making the model taller your increasing the changes. I would guess very few of those layers are going to save you more then 1 change.... on this model

[@mrfoameruk]

I thought that, but you would need to figure out what that would do to the model. ie this one might then need support and it might leave an ugly finish but a comparison would be good

[@markburton5292]

wonder if you did a large retraction but not enough to completely remove the filament and then did the cut if that would help the purge?

[@PritpaulMahal]

That's what I was thinking too. I don't know where the blob would catch but I assume it's in the tubing and not the hot end

[@markburton5292]

@@PritpaulMahal doing a measurement on my unused spare bambu hotend its 23 mm long it would be interesting to do maybe a 17 mm retract before the cut and see what happens both to reliability and purge amount needed. i don't have a bambu AMS. i tried multi color with a pallet 2+ but wasn't really happy with that so stopped doing multi color or multi material. but I still keep an eye on it and the Bambu AMS has been tempting me.

[@astropgn]

Imagine if the slicer had a "low priority print" that only prints with the flushes? It would print the main object with the correct filament, but when it is transition, instead of throwing away the filament it prints another object, something that you defined as low priority in terms of color. That second object will have an arbitrary color, but at least you wouldn't waste material.

[@AntiKipKay]

Yeah would be a great opportunity to make some fidget toys on the side

[@ares7181]

@@AntiKipKay this exists, right click an object and select "flush into this object"

[@joshsherrod2535]

@@ares7181 I use this for extra clips and things basically spare parts that the color doesnt matter its a good way to get some use out of the purge.

[@username9774]

You could try retracting the filament a bit, before it cuts it off in the nozzle. There should still not be that much danger of filament clog since it didn't leave the hotend, you might still get away with saving 1 cm of filament per colour change.

[@TROPtastic]

Some smart person figured that out and gave all the G-code changes (and a very detailed explanation) in a Printables "file" called "Bambulab Profile for up to 60% purge reduction."

[@randomlight5038]

could you modify the "swap filament" G-code to first pull back filament so the glob is just unter the cut line and just purge way less after that ?

[@creeper360bill]

I've seen some gcode on printables that makes the x1 do this

[@sambarney5100]

Your channel and CNC Kitchen are probably the best resources for 3D printing.

[@bloodyfinger5]

Be nice if someone could come up with a way to use a dye to change the colors as the machine is printing to prevent waste .

[@9Blu]

XYZPrinting has their da vinci color series that does (did?) something like that. Could make amazing prints by printing a color layer on top of the 3D print. I am not sure they are around though. Their website doesn't seem to have been updated since early this year.

[@boots3372]

KZbin videos that save me money are my favorite KZbin videos. Very well earned "Like".

[@JamieStuff]

$3500 can buy a LOT of wasted filament...

[@astrolemonade349]

Nice video! I learned something! I am very surprised that the obvious and easiest solution wasn't mentioned: changing orientation. Reorient the printing of the part so that the least color transitions are made on a layer, in this particular example, put Nemo with the nose on the plate. You don't even need supports for this particular model . One should think if the supports result in less waste or less quality when choosing them. I would also consider Z Hopping, video about if from ModBot(Z Hopping To Prevent Batch Failures & Multi Color Same Layer Printing): kzbin.info/www/bejne/in65aqeFj7ZrnLs I would also like some slicer option for this kind of things: flush waste into other object(where color of that object does not necessarily matter). Purging is not actually purging, it continues to use the waste material on something different and makes breaks with that, comes back to the main print and comes back and resumes etc. It would be nice to flush color on object so that you can reuse the color on other object as well that need to be single colored. Unfortunately, this only applies to Bambu Slicer since Bambu printers use this purge and continue mechanism. Prusa does multi color printing only by having multiple extruders.

[@spencerdiniz]

Another factor that wasn’t mentioned, but I imagine would also be able to reduce waste is orienting the part to minimize filament changes. For example, the marlin model, if printed vertically, I think would require less filament changes… with the tradeoff that it would require supports.

[@petersvideofile]

IDEA/QUESTION: Based on the extruded junk at 8:32 and also by the graphic shown at 1:27 is it fair to say that more material could and should be pushed out of the extruder and perhaps it should be cut BEFORE the transition in color takes place. Like couldn't one calculate the amount of material that was between the cutter and the nozzle tip then do a volumetric calculation to prematurely cut and exchange the filament? This almost seems to simple, like it should be being done already. The graphic shown at 1:27 doesn't indicate if it does the cut before it wants to start printing in another color or not, but I was left with the impression it is cutting and exchange at the instant it wants a color change. If the flow of filament is pretty laminar and doesn't mix too much (hard to tell from the 8:32 poop), then perhaps this technique of prematurely cutting and swapping would save even more filament (if it's not already being done). Just and idea, which seems really simple so I kind of have to assume it's already being done and I'm just never heard anyone say it.

[@Exodus1337]

I think one of the most important tuning areas for multi material solutions like the AMS is print orientation. Instead of requiring material changes for every single layer, orienting the model vertically would reduce them significantly. Of course this is not a viable solution for every application, but nevertheless very important to keep in mind while doing multi material prints.

[@K0BRAKID]

Yep multiplier, orientation and layer height it seems

[@wktodd]

If it could cut before the colour change , using the stub of first colour until it is almost changing, then a small purge to the second colour , before returning to print the second colour.

[@Dave_the_Dave]

I believe this can be done in that filament change Gcode section. Basically retract the filament a bit before cutting it, so you only cut off the blobby tip.

[@Prof.Polymath]

I don’t have a bamboo labs yet but if I did; this would be one of the most useful videos I’d ever seen. Great video, great process, and really well presented. Thank you for sharing. 👏🏼👏🏼👏🏼

[@AlexFromNL]

The slicer should use part of the 'flushing' filament for the inner infill, where the bleeding does not matter. For instance have a setting which inner parts can accept bleeding, i.e. after 2 mm from outer edges.

[@oleurgast730]

Actually one major reason bambu (and the SMuFF-Project long before Bambu) uses the cut&poo appoach is, as shown in the video, the risk getting a bad filament tip on unload. So the cut%poo reduces the risk of failiure by a lot (and a failed print would be 100% waste of course). However, Prusa tried to optimice unload doing "ramping", wich needs a bunch of fine tuning - and was a pita, so the MMU2 was never a big success. Until the Revo came and the reliability of the MMU2 got a real boost. Even without a MMU2 you can see how much nicer the filament tip looks on unload. The Nextruder nozzle even have a more nice filament tip than the Revo. Have not tested Trianglelabs Unified Nozzle (TUN) yet as a cheaper alternative. Theese Nozzles with integrated heatbreak are actually an alternative solution to get a nice filament tip on unload. At least with "normal flow" nozzles. Testing a RevoHF nozzle, I got worse results for the filament tip. So my conclusion: For full colour prints I hope the MMU3 on a mk4 should be faster with less waste. While the mk4 does not use HF-nozzles and prints slower, without cut&poo not only the waste should reduce, but also the time to change filament should decrease. Resulting in less time for the multi-colour print over all, at least on smal single objects. However, on bigger prints or batches of multiple prints the faster speed of a high-flow nozzle might be more relevant for the over-all printing time. Still the waste of "classical" change should be less. On the other hand, with multi material (PLA/PETG support interface trick) or just a few filament changes in your prints (colouring just the bottom or printing colour litophanes with just a few layers), cut&poo matters less in time and waste. However, if colouring just the first few layers (for example to embed text and graphics to a display case) I would prefer the Revo anyway, as you might not only use only one colour starting from the 4th layer, but maybe also a bigger nozzle size - so a quick change nozzle is the best for such prints, using 0.4 or even 0.25 for text and graphics in the first 3 layers and than 0.8mm for the rest. Imho there is no "one for all" solution. It very depends what you use a filament changer for - and how often you use it at all. Actually the main advantage of the AMS isn't multi colour at all. The main advantage is the filament being transported by the AMS-feeder to the extruder. The "filament-buffer" (in the A1 integrated on the top of the printhead) synchronice the AMS feeder. The AMS-feeder pushes filament to the extruder. If the extruder does not use it at the moment, a spring gets compessed, a slider moves and a hall-sensor detects the magnet inside the slider - and the AMS-feeder stopps. So the filament always is slightly pushed to the extruder gears. The extruder does not need to pull the filament, all it's force is used to push it through the nozzle. If you ever had under extrusion due to long reverse bowden, a not optimal spool-holder or printing out of a drybox, you lnow the advantage of this solution... Unlike on the MMU2 (MMU3, ERCF, SMuFF, Chameleon 3d...) due to sychronicing the feeders, the AMS feeders are always engaged to their filament. If it comes to over all printing time, the next logical stepp would be to cut the filament tip outside the extruder. Actually disengaging the extruder gears on unload (so the molten filament tip isn't pressed by the gears), of course after ramping. So it could be pulled out reliable and after passing the hub the printer imeditly could print the next filament. While the unloaded filament is pushed into an external cutter, removing stringing an any deformation by just cutting the tip and be loaded directly in front of the hub afterwards, ready to go. This would reduce filament changing time by a lot - maybe to times you would use a toolchanger normaly. For reducing waste one method seems to miss: wipe2object. Still you need a "waste tower", but just to prime the nozzle after filament change (so a small one). Instead of wiping or purging the contaminated filament you print it into an object were the colour does not matter (functional part or an object wich should be painted afterwards anyways).

[@woodwaker1]

Great analysis. I have an X1 Carbon with AMS and it works well, just have to plan to minimize waste. My Prusa XL with 5 heads is a totally different experience. It switches filamnent with a slight delay when the heads swap. I just finished my first print with over 250 changes, worked flawlessly and not much longer than the print with no changes.

[@conorstewart2214]

Maybe bambu could save more filament by retracting before the cut. Pull out to the point where you almost have the nasty end issue but you are still cutting solid filament.

[@eriequiet]

That’s a pretty good idea 👍

[@srmofoable]

TLDW - 1)reduce flush volumes to .50% multiplier from auto calcs 2) Print duplicates at the same time

[@santiagopostorivo]

Error in the graph when showing the multi color model. It says single color as the one in the left

[@Ravera.Drawings]

I'm intrigued by the plastic waste that means using a multi-colored printer. What do bambu labs users do with it? Do they recycle it? Do you throw it in the trash? I think we already have enough plastic pollution in the world to continue throwing residual plastic in the trash. 🙄🙄

[@umake3dprinting]

I use S3D for my Bambus, also for AMS prints. I’ve added a pre-toolchange gcode that makes the filament retract to about 8-5 millimeters below the cutting section before the switching filament starts. There is way less filament to purge, and has been reliable since.

[@fatboy1271]

I love S3D, so I'm happy to see someone is using it for Bambu/AMS!!! I haven't tried it yet...

[@jamalmahroof3298]

Would you be able to share that gcode?

[@gibsonsimpson]

Excellent analysis. Glad to see if you print multiple models, the waste doesn't increase and the time increased slightly. Great info to know!

[@Davis676]

Why not shift the cut command in G code up a few lines, so the filament will be cut, print a few more lines, and then purge. Could probably use a matlab script to locate the cut command and shift it up.

[@Anton_Noname]

Why can't you use a color transition inside the model instead of throwing it away? That is, we print the outer layers in the desired color, and on the inner ones, during the printing process there is a color transition and if this is not enough for a complete color change, then the excess is already thrown away. If the model is not translucent, then the inner layers will still not be visible. Except that in some cases the wall will need to be made thicker.

[@crowguy506]

For a stripy model change the orientation. It’s horizontal changes that matter not vertical ones. The purge multiplies with the number of layers, so low resolution is your friend.

[@WhyplayGaming]

Hopefully purge into infill will get better. Its a shame theres not an option just to add additional infill just for the purge ammount,

[@Luka89856]

if your already purging why use the purge block dosent that just waste more??

[@realgoose]

With the flushing volumes dialed in, the prime tower can be turned off altogether. That saves time and waste. Or setup a sacrificial model for all priming to go into. I have also experimented with not flushing and only priming objects. That saves more time and can work well enough to eliminate bleed.

[@Pamesahne]

That might work, but the reason for a prime tower is not filament flushing but priming the nozzle pressure to reduce stringing as well as underextrusion. Its similar to the initial purge line before every print. Its up to the user to decice if a prime tower is worth it. For me, the stringing is negligible and I reduced the prime tower size and prime volume to 1/3rd, which already helps a ton.

[@gbishel]

Can you change the purge block to be a model instead? (A model that it doesn't matter the color)

[@Mr_Denuto]

This is why I am not keen on the purge style of multi colour. Would love a smaller 2 head option of the Prusa XL though

[@mackenzieallen]

Current color changing tech is a joke. seriously , it works but its not worth making multi color prints if it takes days to complete anything. We should be focusing on more simple tech, like coloring our filament while its being printed. Kind of like how people use permanent markers to make multicolor prints by coloring it as its being pulled into the extruder. keep it simple

[@TeachingTech]

I have a video in sharpie coloured multi coloured printing. It's different to how you describe, though.

[@mackenzieallen]

@@TeachingTech I'll have to check it out thanks

[@rendez2k]

Great video as always. Why do we have the poop blobs and the purge tower? I thought the point of the tower is to flush the old colour? Also I read you can select other models (if you don't care about colour) to print rather than the tower which is probably more useful.

[@oleurgast730]

Cutting the filament inside the extruder would need an enourmous waste tower. Purging needs much less space on the plate. Also it's much faster. Still, the preassure inside the nozzle has to be stabilized. So you should have a (small) prime tower. So Waste Tower = do the colour transition (MMU2/3, ERCF, Chameleon 3d ...), replaced by purge on Bambulabs printer, Prime Tower = get preasure insile the nozzle correct. Of course, any Waste Tower also works as prime tower (so no twin towers needed). Wipe into object can replace the Waste Tower (not the Prime Tower). Of course only on multi colour, not multi material (do not mix PLA and PETG)

[@meytechify]

With such a small model purging into the infill it sounds logical that you would not get a huge change. If you were to scale up the model size where you have more internal volume and then more infil would that give a better percentage of wastage?

[@audigex]

The slicer could definitely do a better job of using the purged material for infill, too. It wouldn't solve the problem but it could certainly help

[@ggaub]

We MMU2 users have been doing these things for ages, of course. One of my go-to methods is sacrificial waste/wipe/purge objects that can be scaled to the size of the main printed object. The four stones from The Fifth Element are my favorite, and I set the whole object as purge/wipe. These can then be quickly sprayed with stone color paint, or even left as-is. I find that people like the crazy colored stones as much as the painted ones. With purge to infill/object, and printing multiples of the multicolor model, per-print waste is greatly reduced. My new XL is putting that to shame, though. It does need some better tuning for print quality, but the only waste is the priming tower, which some people have even disabled with success.

[@suivzmoi]

this video definitively kills the concept of single head multimaterial for me.

[@benjaminvdvyt]

I'm missing orientation. I think a lot could be gained from it. Most of it is in layers, and while the layers wouldn't be flat, it would reduce to like 2 colours per layer instead of 4. And a bunch of layers would just be orange. Of course there would be a hit to print quality, but that's an experiment worth testing.

[@mobilechaosyt]

Was that 5 at once with purging into infill? I would have expected that to be better. We really need a way to use the filament after cutting but before purging so you can use a little more of the clean stuff.

[@Hollingsworth7171]

Thank you that is awesome information, I currently have a print going and these changes would have saved me 11 hrs!!! Luckily this is only one part of a 4 part model and I plan on using these settings on the rest and in the future !!!!

[@martinsalinas7250]

Thank you so very much for this ... i have a PS1, and i hate the waste..so i am going to change the parameter in Flushing... to see if that helps! .. Most Awesome video!

[@Admiral_John]

I assume this all holds true with the X1C? It's the same slicer and same concept for the multi-material printing so I'd think it would apply?

[@TeachingTech]

Yes I believe so. Same type of cutter, same slicer.

[@chicarbiomed]

50% multiplier is amazing! Thanks!

[@eideticgoone7035]

I was hoping to see the special case for optimizing the printing of support material. Often the material for support is only necessary at the interface between support and model. Material changes can be skipped by printing layers of support in model material right up until the interface is encountered. This will eliminate the vast majority of changes and purges for when switching to and from printing a support structure within a most layers.

[@fellwind]

Make that video. People need examples!

[@hassosigbjoernson5738]

Isn't there already an option in the slicer to click that enables that support material only at this contact points?!

[@eideticgoone7035]

@@fellwind I don't yet have this printer. But I found this article and video: "Printing soluble interface supports with Prusa i3 MK2 Multi Material" (search) dated 2017.

[@eideticgoone7035]

@@hassosigbjoernson5738 Looks like there is in Prusa's slicer. Not sure about the Bambu. I don't own one.

[@craigcampbelljr.4097]

@@eideticgoone7035 It's an option in Bambu's slicer as well. I've tried it. Personally I haven't found a lot of use for the specialized support material yet, but I also haven't attempted to use it that much either. YMMV.

[@Nifty-Stuff]

I'd LOVE to see "flush into different object" or "purge into different object" where instead of a wasted purge block and/or poop-chute, all the waste could be printed into a separate color-independent model. So, no waste at all! Perhaps on one corner of the bed could be the main model, and then another corner could be the separate "purge model" (so height wouldn't be an issue, but wouldn't bump into models)

[@startide]

Reading other comments : it's already possible

[@kellyjean4981]

This is great info. I am thinking about getting the P1S and would prefer to not waste so much filament. You gave some great solutions to this problem, but it was also helpful to read through the comments and see suggestion such as having that waste actually going towards making another object… Thanks again

[@davydatwood3158]

I'm a huge fan of Bambu's X1, but I've thought since the beginning that the multi-colour element was really just a gimmick. Now that the printer can fail-over to another spool if one runs out, I can see some utility for the AMS for high-volume machines, but multi-material or multi-colour through a single nozzle will just never be worth the wastage, in my opinion.

[@21142317]

Very much depends on the use case. If the only colour change is an embossed logo for example, it wastes very little but looks great.

[@elchavode6479]

Long term I feel a printers like the prusa XL is cheaper vs bambu due amount of loss material due to the pooping. The bambu is cheaper upfront but not long run.

[@elchavode6479]

Also great video

[@bru_haha]

10:22 How did printing 5 copies of the model require slightly over 2x the ‘model filament’ vs a single copy? I expect it to be 5x unless some other changes were introduced in the slicer settings.

[@alexsnow3319]

Brand new to quality machines. Went from a Monoprice Delta 2 POS to a Bambu X1C. I have been watching and such about the X1* but because I never had one, I couldn't gain muscle memory on this stuff. This video got you a subscribe. Now, DO MORE!!! lol

[@brubakerjm]

Thanks for this video. I've been considering the A1 so I am trying to learn what to expect. Do you think "Purge to infill" would have a more substantial effect on a much larger model? The tiny fish, even at 50% infill, uses very little filament volumetrically. But what about a much larger model? Much more filament would be required to achieve even 5% infill on a large model than even 50% infill on a small model. That said, maybe people aren't printing massive multi-color models that frequently.

[@JK31201]

Great video as always, thank you. There were a few things I think would have been nice to see as well. 1) it would've been nice to see you play with the multiplier and push the limits. 50% works but does 40%? 35? Etc. 2) Something else that I was looking for was a data point on your graph for doing multiple of them but showing it per model. So you had 51g of waste but that's for multiple models. Divide that by the number of models and show that each model only had 10g of waste or however much and then see how many you can fit on build plate (not actually having to print them all) and calculate how low you can get waste per model of you are mass producing. Thank you!

[@connorwhitson9935]

IIRC Prusa MMU changed through filments sequentially (maybe Bambu does the same, Tool1 >2>3>4, etc) You could position your filament colours (tool position) to print in sequence to reduce overall purge volumes (avoiding Black to white)

[@aalborgfantasy]

There is also an option when you print dublicates, you can purge into the dublicates... And at some point, waste material goes down to zero...

[@spock81]

Prusa just released a Dev Diary talking about a new ramming procedure for the MMU3 for Mk4, basically a pseudo cold pull. It should reduce the purge requirement a fair bit. I'd be curious to see a comparison between the Bambu AMS and Prusa MMU2/3 and the XL for multimaterial quality, reliability, and waste.

[@Bastyyyyyy]

i think the biggest problem is the price for the XL :P need ALOT of filament waste to justify the 2000€/$ price difference ^^

[@TGWMS]

its pretty fun to saw that bambu fail :-D i print it on your setting on my prusa XL 5 tool heads and reslts are : 1 h 4m print time and 10,01 g of filament use :-D 5 colors zero waste :-D so my prusa XL is 7 times faster and saves 90% of filament :-D so again this confirmed my choice still use prusa instead of this chinese stolen empty promisses .

[@TGWMS]

my prusa XL 5 tool heads print this marlin with same video settings and reslts are : 1 h 4m print time and 10,01 g of filament use :- D 5 colors zero waste

[@NimVim]

you're a clown

[@AllTheNamesWereInUse]

Good luck with that prusa... It's been quite the let down

[@TeachingTech]

I have run out of patience to some extent but I'm hoping it goes smoother than when I got my Mini.

[@ayo3dprint]

terimakasih atas informasinya, sangat berguna dan sangat mengedukasi.

[@x9x9x9x9x9]

Thanks for this. I am super impressed with this A1 printer. Coming from my old sidewinder x1 to this is wild. I had my sidewinder dialed in very well to where I rarely had to even level the bed. The 1 time I used multi material before this was 2019 with a friends prusa mmu and it was such a pain to mess with. Anyways fast forward to 2 weeks ago and I have tried to avoid going with a bambu machine because I was/am one of those "they're not open source so they suck" type of people. But the fact everyone preaches about these machines, I sucked it up and got one and yeah its great. Its just so easy to use. Anyways thanks for this video as always.

[@Sam-qx7ux]

This was great!! Thank you!! I have my first Bambu A1 mini. I love it but it definitely has a lot of waste when making bigger projects with 4 colors.

[@therealuzr]

Fo this specific model I believe rotation to 90° and print with supports gonna make a diffirence

[@KyleBoyda]

Have you experimented with reducing the prime volume for the prime tower? Maybe this can be reduced since flush into ojects infill is turned on. Not sure if layers with a low amount of infill would cause problems though. BTW, great video!

[@humbledeer]

I'm pretty sure that the amount purged into infill is subtracted from what you need to purge on the tower.

[@Bastyyyyyy]

@@humbledeer that would surpise me, because my waste always looks the same. if that is the case you could just make your purge tower rly big, and the filament purging by the AMS would redcue the waiting time

[@kkrizzz]

What if... you'd have the knife cutting the filament long before it's near the hotend. If you could, by software, predict how much of each color you need, you can cut off after, say 73mm. Then with a direct drive you can still feed the filament reliably but will nee little to no pure?

[@oleurgast730]

You would need to do the filament change before you fully printed the actuall colour, as you need the new filamend to push. In this time oozing could happen. You have to reprime the nozzle to get acceplable quality. And no retractions allowed after cutting. And still, molten plastic inside the nozzle can bleed, so you have a transition. So you have to do cut&merge filaments before entering the hotend. Which is exactly what the Pallette system does - for a price you can get a bambu + AMS. If you do a lot of multicolour, this definitly is the best approach. But no way to splice PLA and PETG to do the support interface layer trick.

[@spock81]

I think the Palette multi material units basically do exactly this

[@Nowyoukn0wit]

Hello! Just curious.. after twitching the 50% infill with 0.5 multiplier, do you still FLUSH INTO OBJECT INFILL? Or just leave unticked?

[@captainobvious1721]

Outstanding video. Thank you. I too am concerned about the amount of waste generated. Especially from ma business standpoint, the waste really cuts into profits.

[@microponics2695]

IMPORTANT NOTE! !!!!!!!!!! Orca slicer V 1.80Beta1 doesn't have the option for "FLUSHING VOLUMES" unless you have your printer turned ON and connected to the slicer on your network.

[@Ab0minati0n]

Changing the wall order to infill/inner/outer should also allow the purge volume to be lower but I am yet to experiment with this

[@TheBoojah]

Yes! Both purge and prime to infill and if that's not enough do it to the inner perimeters too. Could even add more perimeters as needed, better have the plastic in the object than a purge tower.

[@Johncoffee2002]

Interesting Video. Thanks a lot ! What puzzles me is: why do 5 models have 54.3g, where one has 23.8g ? Why not 5*23.8g=119g ?

[@boboscurse4130]

Very informative. The ability to recycle would make all that waste less painful. Some day. You must have a serious operation...$3,500 for that new machine.

[@weeeds334]

One Thing you forgot i like to do. I am not the kind of person who has problems with supports. You could have oriented the model 90 degrees (face down on build plate, tail up in the air) - with supports on. This way you wouldn't have a material change every layer but maybe only every 5th layer or so. (For example: the tail and the face have red segments where the printer could do multiple layers without changing). this method depends highly on the model of course, but in some instances you can cut multible HOURES ;)

[@leanazulyoro]

4:04 what about using the single color print and manually painting the colors? Does that reduce the amount of internal wall? And how would that affect the result?

[@davidoshea7838]

Wow! I used this technique on a two color print and went from 46.44m to 35.4m and 3h 58m to 3h 8m on print time. Thank you so much!

[@vazza69]

Strange i downt haw the flushing volumes setting in my orca slicer. (p1s combo)

[@Jirisvejda6]

I got my AMS today and I got to say I am disapointed. For 33g model I need to waste another 150g of filament. And thats a lot. Of I knew this earlier, i would have gotten the Průša XL in the first place

[@SThompsonRAMM_1203]

Newbie here, and have already printed several models successfully but I do have a question. The block that shows you are printing in multiple colors... does this HAVE to be printed as well? I was hoping this was only a reference to show you have multiple colors chosen. Loved the video as it explained all my extra filament.

[@JJiG]

Great video. I will have to make these changes. I'd love to get a 5 toolhead prusa xl but i can't spend 3500 bux on that for a hobby thats not generating income.

[@otherlutris]

I wonder if you could make it switch early? It looked like there's a long section of the original material before you get to the transition. Printing with a bit of that would remove a bit more of the unnecessary purge. I just got an A1 with AMS Lite; I've slowly been getting more and more regretful of how many caveats there are. It still looks like it'll be useful, but half of what I want to use it for it can't do. Alas.

[@rcmaniac25]

Purge will forever be an issue with single-nozzle multi-filament printers. You can reduce it down, but it will still be there. And so long as something is visible, someone will complain about it. The issue that many miss, and you alluded to, is that purge is finite across the entire print... while the print is variable. So purge into object, purge to infill, purge to support, etc. all help... but if the layer can't purge everything via those methods... then it needs to make a purge tower. And worse, if the layers above don't have the amount necessary, then it needs a purge tower AND needs something printed under it. So the tower will be there even with all the settings. I've wanted better calibration of purge... there are 101 tutorials, blogs, videos, models with instructions on how to reduce purge, but what really is needed is a built-into-printer or slicer way to calibrate AND it needs to be deterministic. I'm tired of so much of 3D printing being "it works for me". Cool, I want to follow some steps, get a result, and then give it to 5 friends and it works as perfect as can be. Sounds like I'm asking too much, but I'm not. People keep saying X printer is almost a consumer appliance... yea, then it needs to act like one and be reproducable. So first, better calibration of purge. Second, I've wanted variable infill (someone else below mentioned this too), where the more purge you have, the more infill you get. And I believe that with a bit of intelligence, the purge can actually continue to additional layers that would otherwise be sparse. I also think that so long as it's not intended to be a moving part, it could be possible to do a Bambu purge poop into the hollow portions of the model. But it leads to maybe 2.5... Second point five: better slicing. Print thicker infill, variable infill, extra support structure, have the slicer recommend to print extra models to reduce purge out the back or into a tower, have smoother wipe operations so priming and purging can go right into infill without worrying about it curled overhangs and such, pickup that the print is going from black to white and try to purge as much out before touching perimeters as possible so bleeding doesn't happen. But also, think of the tools... if you want to change a filament, you probably don't need to pause the printer for some changes. Or you specify in the slicer "tool 1 is red, tool 2 is blue" and the slicer tells the printer "tool 1 is blue, tool 2 is red" because it knows it can reduce filament retraction through that tool then the other. There's a lot that can be done. My fear is that not much work is being done on it. Also, unless things have changed since I last checked, don't use the Auto Calc for filament purge in Bambu Studio (and probably Orca Slicer). It barely works. It's a nice user friendly way to suddenly find bleeding or purging way way too much. I remember looking at the code when they first introduced it and it was literally ripped off StackOverflow (software Q/A forum) for how to figure out the difference in brightness between colors. To make matters worse, it was selected as the answer... but every comment said "the math on this is wrong. You'll get incorrect numbers and in some cases it will be reversed". So it's not good but it makes a great bullet item to announce in release notes.

[@Alkerion73]

I don't think the G-Code modd reduce the real required purge at all, what it does is just saving some filament filament, that's why we have less poops volume at the end. If you retrieve 2cm of filament before purge, it's for sure the amount of filament you won't have to purge, but the real required purge to ensure no colors bleeding will stay the same. The reduced purge value in the slicer is just due to the 2cm of retrieved filament. That can be seen on pictures and tests, the transitions between colors look the same. IMHO it's not really useful and can create more troubles than benefits, probably the reasons why Bambu didn't implement that. I use km of filaments every years, I don't really care saving a few meters, just want perfect prints with no color bleed, reason why I'm not shy on color purge to avoid having crap print results like some people just happy because they saved 2m of filaments. My 2 cents of course.

[@Underestimated37]

See if the infill isn’t going to be visible, my go to would be to have the printers purge process purge during the printing of the infill, instead of on a separate tower, the filament is used, and if you need to push more out and increase the density, it actually strengthens the print.

[@Rawstock92]

When is the "purge to Lego" feature coming?! In all seriousness, a set of usable purge block options (like Lego-compatible bricks) that can be multicolor would be cool ...

[@IdeationGeek]

You should rotate your fish, to make its face face ground :) This way, there will be more mono-color layers, that can be printed fast.

[@craighazen162]

Hello - great video on saving time and money. I have a quick question for you. Have you done any testing with a .2 nozzle? I am going to try the .5 multiplier as you suggested but I was wondering if the .2 nozzle would be even less since the amount of filament should be less in the nozzle. Thanks

[@martinaschulz7740]

so basically in a bigger build (i.e. more volume), there would be way less waste, right? Because it might fully bleed the purge into infill/support.

[@NinjaTheCultist]

would like to see how much you would have saved printing it vertically with support structures if needed. i´m a noob to 3d pritning but i watched quiet a few videos about 3d printing since im considering to buy a p1s for myself :D

[@tgfcujhb7583]

If it could cut and change colour before it finished it's layer it could use more of the current colour before purging.,...🤪👍👍👍👍

[@shanemshort]

turning on infill combination can also half your filament changes as it prints two layers of the same colour at the same time.

[@michaelmendoza1818]

They could improve on the AMS if the filament didn’t have to get pushed and pulled all the way back to the spool a whole yard, that takes quite a bit of time itself and I think would help on wear of the stepper motors. Swapping colors from just that adds like 15 seconds per change.

[@stan110]

At 6:15, I literally spit out my water. I was shocked. I thought my Prusa mmu3 wasted a lot of plastic for the purge, but Bambu's default is twice as much.

[@irmiwolf]

Small question from someone without a multi color printer. Why do you still print a purgeblock eventho the bambulab does the purging on the side?

[@crawlerin]

Doesn't FIlametrix have in their code to pull the filament from nozzle a bit, thencut it leaving just a small tip, withdraw, load new filament, and purge just the small tip left?

[@deathmatt]

Love this video as I wait for my A1...any tips on complex prints like dragons to reduce or speed up?