A lot of people are asking about cutting early and returning to the print before finishing the layer. I had that same thought initially, and even mentioned it in the previous video about this subject, but this solution is more complex than it seems at first. First of all you have to stop the layer early to do the cut. This sounds simple but but often GCode has long lines be a single command, and what if the optimal place for a cut is in the middle of a long line? Do you stop before doing the line? Do you try to calculate where in the middle of the line to stop and turn a single GCode command into two so you can pause to cut? Either way you're adding a new seam to your wall, which is unsightly. Then, what if you're adding only small amounts of several different colors to a layer? I'm talking small enough that you actually need to cut for that color, 2 or 3 colors before the last one gets added? That's a lot to coordinate. And you'll need to be sure you haven't missed anything, like the effect of travel moves, or your color could start changing too soon, or too late. And since every color combination needs a different amount of purge, it gets even more complicated! Not saying it can't be done. Put enough time and effort into it and anything's possible. But if there's a more simple solution, that's more time that Bambulab's programmers can focus on other things. For more ways Bambulabs can improve things, check the blogaroo: www.3dpprofessor.com/2024/03/09/is-zero-purge-on-the-bambulab-ams-3d-printers-possible/

If you are printing 4 objects in different colours you can select print by object instead of by layer and this will print each object before moving on to the next, this will save you oodles of wasted filament as there will only be one filament purge per object.

You get a thumbs up just for the tie alone!

Thanks for the detailed explanation of the thought process behind this.

Plastic waste is on my mind more every time I print. I'm going to have to start thinking about various purge "blocks" that I can create where color doesn't matter. Possibly various brackets, gears, or basically any other section that would normally be hidden inside of something doesn't matter if it's a muddy grey or a dirty light color. Also, I have a feeling that the modified g-code would work for most of my prints that are printed with basic PLA. Hopefully it's updated to immediately push the clipped filament back into the hot end before it has a chance to cool down in the heat break and cause a clog. That would be my only concern about using the change as is.

Seems like a simple solution for Bambu, make it automatically retract when using Bambu filliment where they can test them all. Then they can claim use Bambu filliment and you will have less waist. I think it will be fair to say you may have some people saying they are monopolizing there filliment, but I feel it's fair to say they can't test all other manufacturers and at the end of the day you could probably trick your printer in thinking it's Bambu filliment at your own risk.

Always love being smart with what you have to reduce waist. Great video. 👍

Love my AMS, it allows me to use multiple materials and colors one one plate (do print by object, so that it only changes materials after finishing one object, and before starting on the next object). Not needing to manually change spools between your most used filaments is also very nice. I’ve not done a single full multi color print, just a handful of icons, or maybe some text in a different color on the first, or last two layers. Though I have made extensive use of multi material supports, setting PETG as support interface material for PLA, ABS, and ASA, and use PLA for PETG. Using them only as support interface materials gives you super clean, easily detaching supports with minimal waste. Just make sure to increase the purge value by 2x, or set the primary filament to pure white, and the support to pure black regardless of actual color. This is to ensure you get only the pure material in the print. If the mixed filaments are not properly purged away, your print will get a very weak section in the layer above the support, because that layer contains a mix of two incompatible materials that do not adhere to each other, and thus the layers detach almost as easily as the supports do.

Just got an X1 Carbon and AMS and am printing my first multi color print. I watched your other Video on the subject and this came on next. from an hour ago! perfect timing! haha I love my 3 collored articulated FOX and hope I can reduce the poop! XD

A version of the false hydra would be very cool to see in the Cthulhu series.

I wish bambu lab would come out with a multi head printer with each head dedicated to a color and even type of filament, but I imagine you would go from a x1c $1500 printer to double or triple the cost. But there would essentially be no waste.

You could check the ERCF v2 color changer for klipper machines. It does EXACTLY what you expect but is clearly for passionate and advanced makers, not for a customer that use it as an appliance. We can control the amount of cut filaments, tip forming for more reliable feed, cooling and ramming etc..

Good info thx!

🤷‍♂️it's progress, thanx for all your work

Love the tie!!

Biggest factor for me is orientation and using surface text printed horizontally instead of embedded.

I wrote you a message on printables! Thanks for doing a video!

Worth noting that the reduce purge script was included in the latest beta release of bambuslicer

The perfect every time overkill option should be the default, though there should be a setting or option that allows you to do the waste reduction procedure you mentioned, maybe even give it it a little pop up warning and confirmation button and a troubleshooting guide to minimize the amount of support requests this would inevitably cause

Why arent the default settings to purge to infill

even the 3 block purge was higher than a no purge object print on my tool changer lol.

New to 3d printing and I've got a Prusa Mk4. Not tried messing with any custom gcodes but, I've asked ChatGPT for help! Arbitrarily, I've told ChatGPT to retract the filament by 1cm, cut it, then advance the filament again by 0.75cm - these numbers will almost certainly need tweaking with experimentations! Here's the result: *** ; G-code for filament color change with retraction and advancement ; Set units to millimeters G21 ; Set absolute positioning mode G90 ; Set initial feed rate F3000 ; Start printing ; Layer 1 G1 Z0.2 ; Move to layer height G1 X10 Y10 ; Move to starting position G1 E10 ; Extrude filament ; Layer 2 G1 Z0.4 ; Move to next layer height G1 X20 Y20 ; Move to next position G1 E20 ; Extrude filament ; Filament color change G1 E9 ; Retract filament by 1 cm G1 E8.25 ; Cut filament G1 E8.25 ; Advance filament by 0.75 cm ; Layer 3 (with new filament color) G1 Z0.6 ; Move to next layer height G1 X30 Y30 ; Move to next position G1 E28.25 ; Extrude new filament ; Continue printing *** In this G-code: Layers 1 and 2 represent the printing process before the filament color change. Before the filament color change command (G1 E9), the filament is retracted by 1 cm (G1 E9), cut (G1 E8.25), and then advanced by 0.75 cm (G1 E8.25). After the filament color change, the printing process continues with layer 3 using the new filament color. You'll need to adjust the X, Y, and Z coordinates, as well as the extrusion amounts (E values), to suit your specific 3D printing project. Additionally, make sure to test the retraction and advancement distances to ensure proper filament handling during the color change process.

The issue with that custom gcode is that its very filament dependent, which you mentioned. I was excited when I first saw it circling on reddit a while back and tried it for myself.....until the jams started...oh the jams lol. Silk filament is a no-go, as the creator said. But so is certain colors of regular pla, even BBL brand, silver BBL ABS jams, every brand of Galaxy PLA I tried jams, Sunlu Clay PLA jams, Sunlu Matte Light Blue jams, Eryone Fuscia PLA jams. Those are all I can remember. Like i said, it's great when it works. But if it's a filament that hasnt been tested with the gcode, you have to babysit it to make sure it doesnt jam.

Holy cow! You still have some useable 3D solutech? Man, I ran out of the awesome green color years ago.

My primary problem with AMS is the time needed for filament switch. It takes about 1min30s for one change and if print has thousands of these... (no idea how long that is on "AMS Lite").

If I select a filament change on my Tenlog TLD3Pro it slowly pushes it forward then quickly extracts the filament all the way back. I always wondered why they made it do that instead of just heating up and pulling out. Wondering does that purge reduction script do that also? Push a little, quickly pull back, cut and then push forward again then extract all the way.

I don't have one of these but one of my printers can do 2 colors and it's the same thing. It drops all the poop down a little shoot on the side and I end up with a big pile of waste on the table.

Bambulabs could do some things to make this better as well. www.3dpprofessor.com/2024/03/09/is-zero-purge-on-the-bambulab-ams-3d-printers-possible/

I want to print some key chains for my sons school band. I want to print in a hard tpu. Any recommendations? I just want them to be durable and not break easily. I will be buying a new bamboo printer for this. Should this work and the newer less expensive bed slinger or the carbon?

The printer knows when the next layer is coming I feel like an early cut would be the easiest way to reduce waste because the end of the last layer could be finished by whats in the nozzle. The only timing has to ensure the "transition" is the only waste..

One setting you should enable is "purge into infill" (I believe that is the name)

@3dpprofessor Why not cut it before the color change to avoid retractions entirely? Once you have [cutter to nozzle distance]mm + [color 1 purge endlength minimum]mm remaining to the color change, cut. Retract color 1. Push with color 2 for [cutter to nozzle distance]mm, purge for [color 1 purge endlength minimum + color 2 purge startlength minimum]mm. No retractions, and your amount of the first filament in the hot end to purge goes down to just a few micrograms. This would require having your GCODE analyzed to find at what point to do the cut by measuring extrusion lengths... but subtraction is a concept that exists.

11:55 is sort of like how mosaic does it in the palette 3 in the heater cartridge thing.

this is how the mystery flavor in Dumb Dumbs was created.

Is it possible to purge as an infill or a model that say is holding a weapon were the weapon is entirely made from purge material, if this don`t make sense its because I don`t own a 3D printer yet. 😐

Can it purge to an object that's not as tall as the main print, but wide enough to absorb all the waste? (Nozzle would have to lower for each purge).

So you said you lose 2 cm every time the ASM cut filament then got purged, but is it possible to cut the filament and purge 1.5 cm on the print itself before you really need to change color ?

Don't forget you can print per object instead of per line

Maybe this exist but it would be nice if there was a mode to use purge as infill.

Have you tried adding the command for feeding the filament to his script yet?

3:20 Nobody does that...and even if they did, there is a setting to print each part in series...so there would only be 1 color change per part...

you mentioned the transitioning sections, can you not reduce those zones? edit: also I've seen hotend on ali express that has 4 filament going into it, i wonder if that could be of benefit?

Hello, that's very interesting. I don't have any Bamboo but it crossed my mind and it's precisely the mess that made me change my mind. However, I have a question about these blocks that you've printed with the "purges", I can't see what purpose they serve at all, in fact I think that's also waste, could you tell me or show me a video explaining what these blocks are for because at the moment it makes me think of a roundabout way of making waste that in the end amounts to the same thing as "poo"?

The modified G-Code you used should reduce the waste dramatically. If that’s not the case you did not modify the flushing volumes correctly. With the modified G-Code a 0.35 is always fine and in most cases 0.25 is enough purge.

Will printing multiple of the same objects increase the waste or will it be able to do a single purge for multiple objects?

Reminds me of an old saying...... waste not, want not!

Another method would be to properly calculate how much more you need of that particular filament before swapping and then doing the swap early - so that the 'purge' is actually used in the part. The end result would be the same, except you wouldn't need to retract and extrude. The risk would be that if you don't calculate it properly you could get color mixing - but that's the risk with any of this... The retraction idea is good - but I think the risk there is hot-end jams/issues - where as you wouldn't have that risk if you were to say swap from the first filament to the second one before the first one is done printing but not so far before it's done that you end up with the transition in your parts. Downside to this would be that I don't think it would be a simple gcode modification for before/after the cut. You would need the slicer itself to calculate when to do the cut earlier.

I will finally be getting a P1S here in the next few weeks and considering the AMS as well. Does anyone here really use it that much? I think I might wait on the ams for now.

Or get a Prusa XL with 5 individual print heads instead of 4 colors with one head. I own neither so can't vouch on purge waste volume but would think the Prusa would beat the bamboo hands down.

The waste blocks, 100% infill? Thinking that prime could be hidden inside an object being printed maybe

Why dont they use the internal structure as a purging space? no one is going to se it. it seems like the perfect space to lay what would be waste material?

9:55 isn’t it the other way around purge area bottom then transition then overkill. I suggest the diagram is right but the labels aren’t. So you want to shorten the top black area…

15:07 you say there's no bleed over into the white but in this clip you can very clearly see a ton of bleed over into the white. What's the deal with that?

Why not print the purge to infill ?

Why pull the filament up to cut it, why not just pre cut the filament, finish the layer, then retract the pre cut filament and change to the next color?

I was thinking of using the purge inside the object as infill. This doesn't have to be 100% correct color.

What if they made the purge another model?

Would be interesting to purge into infill, shouldn't be "hard" to calculate when there isn't enough internal infill and will purge as waste. On larger prints that might make a significant difference. What i would like to see is a tool change that only changes the nozzle itself. Remove having to have multiple extruder motors and gears, etc. This will also make it far smaller and economical. I can already imagine it, fully functional.. someone send me a few grand to cover materials and living costs for the week or two it would take me to develop it np 😅

what about flush to infill?

Love that tie!

That prime block could be replaced with brass bristles

No reason to "push back" with the same filament because the next filament will do just that.

honestly not sure why this wasn't done from the jump it's the "obvious" way to do it

3:30 literally noone does this, print by model rather than by layer ensures that each model prints to completion and then does a single purge to the new colour before starting the next model. There's more than enough genuine areas to find where someone can save filament without manufacturing fake ones.

2:03 that first hidden one looks oddly similar to Big B 🤔

what about purge infill ?

I have some cash set aside just so i can buy the Carbon XL when they release it

The simple solution to make a filament machine that is affordable, that hooks up to the poop chute and it turns it into filament. If people could just recycle their filament, this would not be an issue at all, and even something to look forward to after prints since you could just turn it all back into filament. Filament machines are too expensive, and the affordable ones are not a plug and play system which makes it really troublesome to use for the average user. If Bambu could make one that communicates with the printer, you could organize colors, filament types, and even mix colors depending on how you set it up.

Just reduce your vertical resolution to 50% and you reduce the waste the same amount.You’re welcome.

"Printer blocks?" Can we all pause for just a moment to acknowledge that these are indeed 100% waste as well? They're just not strings in a bucket.

How about using purge as infill?

VeggieTales!

Not a fan of the Bambu for figures and toys. You get much better results with resin and hand painting.

BambuLab is closed system, therefore no contribution was made here. In addition they are actively trying to patent open source inventions. So..... the premise is false from the beginning,