3D Nerd put out an update saying that Bambu sent him a new plate and the problem (at least as of the video) had gone away. Not sure what happened to his original plate. I am hooked on the Frostbite plate at the lower temps so I don’t have much experience on the default plate with the H2C. Might want to check if it is a plate issue.
I know I am late to the game on this one but one mistake I was making with textured plates is using too much glue. If your glue is “filling” the texture, that is too much glue. It has to be a smooth, thin, and even coat. That is hard to do on a textured plate. Another tip is to clean the plate with isopropyl alcohol before applying the glue. Since wiping with Alcohol, I dont use glue anymore; for PETG and PLA… I use glue on Nylons but only on a smooth plate.
I’d like to comment on something I find unacceptable. If I have a machine with multiple print heads, it’s unacceptable that it forces me to use the same diameter throughout the entire print. I make a lot of keychains. If I want the first layer to be printed with 0.2mm and the rest with 0.4mm to optimize printing time while still achieving high detail on the visible surfaces, I have to trick the software. I print the first two or three layers with 0.2mm, and then, in a “new print,” I have to “force” it to print the rest of the job “in mid-air,” using reference pieces.
I don’t understand how:
-
The software doesn’t allow you to combine diameters when you have a machine that changes print heads.
-
The software doesn’t natively allow you to start printing on any layer you want.
I know these are minor issues that can be resolved in other ways, but if I have 7 nozzles available in total, the least I could do is be able to combine them however I want. And if I want to print 7 colors at 0.2 mm, and then 7 colors at 0.4 mm on top of that print… the least I could do is tell it, “Hey! Start printing at 0.3 mm, not 0 mm.”
Thanks everyone.
Best regards.
You’re assuming a lot of things that would be easy, but actually quite hard.
- it doesn’t allow this because the software doesn’t have this capability. It’s not something that the software doesn’t want to do, it’s simply a physics, math and programming problem that is so complex difficult at current stage and not well-solved by anyone yet.
- it doesn’t allow you to natively start printing on any layer, because the motors are not servos, and doesn’t know the build plate’s accurate position. Even if it knows, the print will have a high risk to fail, or at least leave a huge scar, because plastics would deform during repeated heating and cooling down. This is simply physics that you can’t get by.
I’m not saying that there’s no way to solve the problem, I’m just saying that the entire slicing software that took years to develop, was built on some basic assumptions that’s without knowing multi toolhead would be possible in the future. So, the amount of engineering efforts it would take to make multi nozzle size usable would be enormous.
When you bought the printer, was there anything in the spec that said you can use different size nozzles?
I don’t think so, and now you are complaining about it.
the slicer is open source, feel free to code yourself into doing that if it’s that simple to implement as it’s only a minor issue 
sorry but you’re saying this like you are the first person that every thought of this, and that there hasn’t been many attempts that haven’t been able to successfully and consistently achieve multi diamenter in single print, it’s a difficult problem to solve
recently Creality is actually announcing that their upcoming printer (k3 with klitech) they will make this available, but i have to say i’ll need to see it to believe it
If you would have taken some time and wandered over the forum, or better, the entire 3D printer community on the internet, you would have found out that this is not available on any nozzle/tool changer at this time.
There are some slicers around that have some experimental features that support limited functionality.
If it would have been that easy, every vendor would have implemented this already.
And as @h3li0 already mentioned, Creality is promising something in the near future.
The H2C/D are fully capable of doing what you ask, if you can supply the correct gcode.
Gotta say, I don’t love the “I find it unacceptable” tone. Bambu has never once stated this was possible, and while it is being developed, there is no guarantee it will ever succeed to a degree that can lead to a public release, and even if that happens, there’s probably some limits to how far it can be taken.
The fact that nobody’s done this yet should be a good indicator to you how hard this is to crack.
Put differently, if Bambu released full multiple size capability tomorrow, they would immediately present s far greater value proposition than the fized-path extruder changer machines like the U1. The example I use is multiple fine outer wall layers with a 0.2mm nozzle, then a single infill and reinforcing layer with a 0.8mm nozzle. That approach would crush every other printer out there in speed for ultra high detail prints, and Bambu has every incentive to be able to do this - so if they can, they absolutely will.
In the meantime, you can do whatever you want by making modifications at the G-code level - that may give you an idea how complex this problem truly is.
Colleagues, please excuse me. I’m writing in Spanish and using Google Translate. It’s possible some words have a more “aggressive” tone than I intended. I never attack anyone. These are community critiques to improve the software. Due to life circumstances, I’m involved in the design, manufacturing, and programming of industrial machinery, so I have some idea of what I’m talking about. I’ll answer you one by one.
@hotellonely No, these aren’t difficult things. Nothing we’re talking about is difficult. Slow, expensive… yes. I prefer to call it “entertaining.” But in no way difficult.
If no one has solved it until now, it’s clearly because there has never been a machine that offers these possibilities so directly. We could do it by hand… But… BAM!!! Automatic, in 2 seconds… Nothing like that has ever existed, and that’s why no one has solved it. Because no one has had the need to program it.
Whether the motors are servos or steppers is a matter of hardware operation (windings, encoders)… and it has nothing to do with whether the machine knows where zero is or not. When the machine starts, it touches the table several times at various points… it knows perfectly well where zero is. You would only have to program a cell… mind you… this is printed starting from 0.3, if the user had that need.
This can be emulated with tricks, by deceiving the program. So if it can be hacked, it can be done natively.
I completely understand what you mean about the software. It was created with a fixed printhead concept, and modifying that… isn’t a simple copy-paste job that can be done in a couple of days… it requires touching the core code, and I understand that’s the really delicate part, and it’s not worth it for them given the current market share of the H2C compared to all the other machines in their catalog.
@hnodesign Indeed, it didn’t mention anything about a combinable multi-head system. And I’m not complaining about that. My complaint is that, if you’re a leader, if you have a unique machine, you launch it on the market without exploiting possibilities that no other machine, not even the other multi-head machines from the competition, can offer. It would be an option that only the H2C can offer because only it has color change and the ability to change the printhead on each side… It would be something exclusive… That’s what I’m complaining about. But I understand that the time and human costs of developing something like that have to be taken into account.
(continue)
@h3li0 I’m not a programmer, but I do know what needs to be done or said to make the machine do what we want. Programming it isn’t my job. I manage groups of people and give instructions. And things work. On the other hand, I’ve often encountered, for example, programmers who don’t enjoy what they’re programming for. Programmers for newspaper layout software who have never laid out newspapers, or, as in this case, programmers for 3D software who don’t have a 3D printer at home. Obviously, they don’t have to enjoy it; they need to receive the right instructions. But here it depends on how many people there are, how much time they have, and the scope and impact of this implementation. The complex part here isn’t the functionality… it’s the human capital and the final return on investment of dedicating time to this development. If they don’t do it, it’s because it’s not worth it for them. But from my point of view, that’s a mistake. I’m not going to program anything for them… Give me a team of people and you’ll have it done in a month… I guarantee it.
As I said… there’s never been a machine with this potential, and I doubt many people have thought of it, considering that 90% of users download pre-made profiles and parts. They just print them. I design the parts; I like the details, the nuances, and I see a unique functionality of this machine here, unmatched by the competition, which remains untapped.
The Prusa index system is forced to use a nozzle linked to a color until they make an AMS update. The Kitech system seems the same, but I don’t see it clearly… they don’t explain much. It’s not a hardware problem. It’s a matter of telling the slicer… “Hey! This object at 0.2 for details, this other one at 0.4 for the infill that doesn’t have details…” or… “Hey! This entire layer at 0.2 and the others at 0.4.” Combining a color with different nozzle diameters is something only the H2C can do…
I insist… it’s not difficult. You just have to program it. Similar things have been done on CNC milling machines for years. In the 3D world it’s a little different, but it’s the same G-code generation procedure.
@Gerrit This feature isn’t available because there’s never been a machine with the ability to change nozzles like the H2C. That’s why nobody offers it. If you have a machine that can do this, and you don’t utilize it, you’re missing out on a functionality that only your machine can provide. And this, as of today, neither Prusa nor Creality nor anyone else can do… only Bambu… and it’s not even programmed…
I think they’re not implementing it because the impact on the H2C’s user volume is lower than that of the other machines in their catalog, and they prefer to invest their resources in developments or improvements that will benefit far more than just H2C users.
Creality promises new features for when it’s released… no one knows when that will be. Bambu already has the machine on the market. They just need to program it. It’s simply a matter of printing the assigned area at 0.2mm, subtracting this area from the total area of the layer or color, printing the rest at 0.4mm… It’s not difficult. Milling machines perform roughing and finishing passes. It’s the same concept. But I understand it’s not worth it for them.
(continue)
@CarbonForge Please excuse me if the translation has a harsher tone than intended. Bambu hasn’t stated it, but this is absolutely possible.
The limitation is that it’s only usable for H2C users. For other single-head machines, it’s completely unfeasible, unless you want to do all the head changes manually… and that’s not the idea at all.
Nobody has managed it because, as I mentioned, a machine like the H2C has never existed, and there was no need to program something like this. Now it exists… Program it!!! Get ahead of your competition!!! They could have it first, more than 18 months before their competition. Inexplicable.
Indeed, THAT’S THE POINT. If Bambu were to launch this tomorrow, it would be TOTALLY EXCLUSIVE to them for at least another year… Index can’t do it. Creality is coming around 2027… If they don’t do it now, it’s because the H2C (home to home) industry doesn’t have a large enough market share to make the effort worthwhile. Unlike what they did with the full spectrum system, which affects everyone and gets implemented. This only interests the H2C industry. You’ve given the perfect example. It would be unbeatable. Being able to combine nozzles with the same color, even for perimeter details and quick fills… For example!
I know perfectly well what needs to be done and what I have to tell the programmers to implement it. I’m not going to program a single line. Give me a machine and it’ll be done in a month.
Laminate your object. Take the surface of that laminate. These colors, or these vectors, at 0.2. The perimeter of these at 0.2, the necessary walls to make it achievable at 0.4 or whatever is programmed… I don’t know. 4 walls at 0.2, the fill at 0.6. It’s not black magic. You just have to provide the option. Organizing the interface, making it usable and simple for the user, is almost more complex than the software programming itself.
Please excuse my tone again. I’m a product developer and would never attack other developers. But my English isn’t that fluent, so I use translators to communicate with you “fluently.”
I understand that most users won’t use this feature. And many H2C users don’t even need it. It’s just me; I like to push the boundaries.
I’ve managed to get the program to do what I’m asking by tricking it, playing with the parameters. I use “phantom” reference pieces to print “in mid-air” (over previously printed pieces with a different nozzle diameter).
If you’d like, I could make a video showing how I do it. But it’s simple. And really… it’s only of interest to crazy people like me XD. Most people just download things from the internet and print them… and that’s it…
Thanks for your comments and greetings to everyone. Let’s keep talking.
I come from CNC and conventional machining before I started doing 3D printing. I agree with you that conceptually, it’s a straightforward concept - it’s just G-code and toolpathing.
With that in mind, the CNC parallel can help explain a bit of why slicers in general are so much less sophisticated than what the hardware can handle.
There have been basic CNC 2.5D or even 3D toolpathing programs for years, just like there were basic linear motion 3D print slicers for years. On the CNC side, things like waveform roughing, constant load cutting, and resonance compensation all
There’s the odd YT/TikTok-Vid on more or less successful prints. And there’s a bit in Research papers.
Most interestingly though, it seems to (have been?) be an experimental feature in Prusa Slicer for the XL.
Comments are mixed but at least one indicates that it did/does work.
Well, with it being looked at in Prusa Slicer, from which Studio was forked, from which Orca was forked, from which Creality Print was forked…
I fear they are all just waiting to get a free ride. Wouldn’t be the first time…
& 
Edit: Addendum: The Orca folks would probably be able and willing to do it. They’d just need to consider that the usual “minimum” layer height is a legacy figure not rooted in physics. In my trials, I found larger nozzles to be able to print thinner layers as the first actual physical limit is extruder flow control.
=> Contrary to popular belief: Larger nozzle: Lower actual minimum layer height.
=> Just use the smaller nozzles layer height for all and a first release could be ready…
There have been dual nozzle/tool head systems for at least the last 10 years.The IDEX as one of the pioneers in that field. That the H2C can switch nozzles doesn’t introduce suddenly the requirement for running different nozzle sizes. That question is also very old.
And that milling machines can do roughing and finishing, is not the same as using different nozzle sizes in an additive technology. If you would only print cubes, mixing different nozzle sizes is simple, but as soon as tilted wals, overhangs, mapping infills, different layer heights, etc. come into play, the complexity increases.
If it was simple, it would have been there already for a long time. You don’t have to be a programmer to understand the complexity of the topic.
It will eventually come, maybe Creality has already developed a good solution, but only time will tell.
Check this video out to see an interview with Bambu’s CEO. Skip to hour 1 minute 19 for a section on why different nozzles sizes is currently not allowed.
“It’s not mature enough and we’re not confident enough to deploy it” is wiiiiiiild.
Weird that this didn’t stop the Filament Track Switch, the X2D, Bambu Connect, etc. from being released.
I understand that Orcaslicer has supported multiple nozzle size slicing since 2024, I’m so hoping to see Orca get H2C support for this one reason.
Maybe it’s low on their priority list….
It’s the one thing that will make the H2C competitive against the Snapmaker U1 and the K3, and has direct benefits for the H2D and X2D as well.
It should be higher on their list of priorities.
Where do you get materials for 8-14 dollars per kilogram??
Calibrated my filament today for the H2C and the H2D. Set up a fun heavily multicolored model (flying tiger) for the H2C and bypassed the left nozzle so it would only utilize the right Vortek system. It added quite a bit of time but I wanted to push it through the extra changes. So far - outstanding. It’s really printing beautifully. Man, I really do love these printers!