Irregular surface with dual nozzle print

Halbflauschmammut · May 30, 2025, 8:23pm

Hello everybody

I’m using the second nozzle on my H2D for printing support interfaces, and I noticed an irregular surface on the layers a support interface was printed.

While this is a test piece, I could observe this behavior on all parts so far, where I used support interfaces.

First I thought it is an inaccuracy caused by the nozzle change. But this does not make any sense, as all other surfaces are perfectly aligned. And they align also perfectly after every irregularity (also with respect to each another, below and above this “seam”).

I also printed 2 normal cubes, one higher than the other, with the 2 nozzles. And I expected those irregularities on every layer until the first cube is finished, as the nozzle stops switching then. But the prints were just fine.

So, has somebody an idea what causes this “brims” on the layers a support interface is printed? And how to get rid of this? Looking at the other layers, it does not seem to be an accuracy issue of the axis.

Every input is welcome, as being able to print a different support material with a second nozzle was a main reason for me to get this printer. And in this price range, such artifacts shouldn’t be there…

3DTech · May 30, 2025, 10:12pm

Hello,

What fillament type and project filaments setting. ?

I’m going to take a wild guess you are using Bambu Lab PLA red ?

If that’s the case, check the fan speed in your slicer result. If you notice green strips, try setting the fan speed for overhangs to 80% and then you should see the green strips all gone and and hopefully this is all you need with no extra settings.

You may need to tweak additional settings depending on the filament or part design, as the H2D nozzle and fan controls aren’t as intuitive as those on other printers. However, you can still achieve great prints just make sure to carefully review all your slicer results. This will be your best tool for identifying potential problem areas.

Halbflauschmammut · May 31, 2025, 7:49am

Hello,

Thanks for your input.
Both materials are a generic brand (Purefil) and I’m using the profile for generic PLA and PETG provided by Bambu studio.

There is a uniform fan speed for the part, only the PETG has a different one.

I now also tested if changing the PETG to PLA helps, but the same result. In fact, I can’t find any parameter, that is different on these layers.

Changing the order of walls didn’t help as well. My guess was, if the outer layer is printed accurate, it will stop material from squeezing over the boundary.

Any other input is highly appreciated…

philch · May 31, 2025, 9:18am

Try Increasing the number of walls to see if it helps

3DTech · May 31, 2025, 3:44pm

Alright, now that we have more information, let’s try to figure out what’s going on with your part.

When printing the main component with PLA, with the nozzle at 220c, using the part cooling fan at 100% and the auxiliary fan at 70%, results in a few perfect layers. However, once the support interface layer is added using PETG, everything changes.

The PLA part body, which was previously being cooled at 100%, now receives reduced cooling. Additionally, the part fan speed drops during the PETG interface layer and we have to wait for the interface layer. As a result, the PLA can begin to shrink. When it switches back to the PLA nozzle fully primed and ready it starts extruding PLA at 220c on the now cooled main component that can have layer offset.

The H2D hotend have a different melt zone compared to other Bambu Lab printers and has a powerful part cooling fan, which I’ve been adjusting by lowering the fan speeds for certain filaments even PLA. Keep in mind that the H2D is a new printer, and filament settings are still being updated. You’ll need to adjust your thinking. Some filaments will work just fine, while others might not, as many people are discovering.

You need to keep testing with the H2D and filaments with different nozzle temperatures and part cooling fan speeds. This knowledge helps you evaluate component design and determine what settings are likely to work best based on your testing with plastic temperature fluctuations.

To address this, it’s crucial to optimize cooling fans for both the filament and the part design. When printing small parts or models with numerous walls, holes, or overhangs, lowering the PLA nozzle temperature by 5 to 10c can be beneficial to.

You can perform a quick test by setting all the filaments fan speeds to 60% just to observe how it impacts this part. Start by turning off the auxiliary fan to better assess the effect of the part fan alone, then re-enable the auxiliary fan if additional cooling is needed.

I hope this is helpful and that you find the right combination.

Halbflauschmammut · June 1, 2025, 12:02pm

I run 2 tests now:

I printed a part with PLA and another part with PETG. The PLA part was higher than the PETG one, which means, after the PETG part was finished, I should see a difference in quality on the PLA part. But this was not the case. It seems, that it’s only affecting surface quality when printed in the same part.
I switched off the auxiliary fan and changed the cooling settings in the PETG profile to the same as in the PLA. If I understand you right, this should lead to the same cooling the part receives when printing the PETG interface as when printing PLA. But it didn’t change, these lines are still there.

And according to the slicing results, there is nothing different on these layers. line width, temperature, speed…

I’m really under the impression that for some reason the printer is overextruding. However, the only parameter I can find is the first layer line width. And reducing this has no effect either.

So, I’m not sure whether the cooling is really the problem, or is it a bug in the software? Layertime could it be too, but i can’t get equal layertimes, no matter what i try in the material settings…

3DTech · June 1, 2025, 12:11pm

Can you share the step file ?

Is this like 20x25 with 5mm steps ?

Halbflauschmammut · June 1, 2025, 12:26pm

I can’t upload a step file here.
It’s 40mm wide, 50mm high with 10x10mm steps. However, the exact dimensions shouldn’t matter.

3DTech · June 1, 2025, 1:04pm

The size can affect layer times and cooling durations when printing with plastics.

I wanted to share this link with you in case you haven’t seen it. It shows how cooling on the H2D affects parts, showing that 100% cooling can sometimes be too much for PLA.

Wall Layer Lines

I’ve successfully printed PLA with PETG interface on the H2D without any problems, though I use a variety of different settings rather than the default slicer ones.

I can create a CAD file according to your specifications. I believe I might have some time later on today.

Halbflauschmammut · June 1, 2025, 2:11pm

I read some of that thread, but I was kinda late to the party and therefore didn’t read everything.
I’m running another test print right now, where I reduced the fan speed to 10%. I don’t really see a difference compared to the other tries. I however observed, that printing on the interface is treated like printing a overhang. The fan goes on full speed and the print speed is reduced. Maybe this is the problem?

Here is a quick drawing of the dimensions.

3DTech · June 2, 2025, 2:20am

I didn’t get a chance to run the larger 50 mm block print today, but I did manage a quick test using some two-year-old eSun basic black PLA. I just don’t use PLA very often.

Setting used on this support block
Nozzle 215c, small component and low gloss finish.
Layer height 0.2 - standard process
Prime 5 & 55
Wall gen Arachne
3 walls - I-O-I
Shell layers top 5 botton 4
Bridge 30 speed and 135 direction
20% Rectillinear infill do to pla and no overlap
Top & outer 100 mm/s max
Top surface line width 0.38
Inter wall acceleration 6500
Prime tower infill gap 100
Support fan 60 %
Slow down for overhangs off

Model print time 26m4s

I had to reposition the block to make the transitions lines highlight outdoors.

Front of block by glass door - size 25 x 20 x25 mm

Left side (Aux fan off)

Fan Settings:

Speeds:

Layer Time:

I hope this information gives you some new ideas to try as you figure out what works best for your filament and part designs - Keep in mind, some filaments simply aren’t as good as others.

Halbflauschmammut · June 3, 2025, 6:28pm

I tried with your settings, it’s not perfect, but i think somewhat better.
That some filaments are better than others is true of course. But I also tried ABS and it’s the same result.

I think I have also to consider: what is a good surface? It’s hard to see in your picture, but I think you got some minor lines as well? Perhaps, it’s the best what I we can achieve right now with this printer and software…

hotellonely · June 3, 2025, 6:39pm

It has almost nothing to do with the support interface. It’s mostly the bridging layer (first layer above support) that is causing this issue.

It’s inevitable. It is also why round hole on vertical wall is a terrible idea for FDM.

Bridging requires strong cooling, and slow printing. Even if you have support material, it still has to slow down and print as bridges, again with stronger cooling. Imagine you have a full blown fan blowing on the layer for a very long time? Yes you get a very cold later. And a cold layer shrinks too much.

These combined would cause a significant layer line.

If you want that to be less pronounced, use CF or gf filaments. They are better at anti shrinking, so it would look much better.

3DTech · June 3, 2025, 10:00pm

The side wall is smooth; I had to tilt it under the light to even make the lines visible when viewed straight on, they’re barely noticeable. I used black filament because it highlights color shifts and imperfections more clearly. If I had used white filament, those color changes wouldn’t be visible at all.

Have you done the offset calibration with the white and black pla filament? I can’t remember if you said you had ?

I also have the vision encoder calibration done on my printer it wasn’t far off, but every bit of adjustment can help.

This was just a quick setup if I had spent more time and reprinted it, I would have adjusted the side overhang speed more. But I don’t usually print with PLA for my projects.

The H2D isn’t as forgiving as other printers, but it should improve as the slicer gets better.

Halbflauschmammut · June 4, 2025, 1:57pm

I run flow dynamics and flow rate calibrations on the used filaments. And of course the full printer calibration. Inspired by the z-banding thread I also removed the hotends and checked for loose screws.

I however don’t have the plate for vision encoder calibration. Perhaps I will get one.

@hotellonely I tried to disable the cooling while bridging, and didn’t see a difference. However, I don’t claim to fully understand which parameter I have to change to stop cooling while printing on support.
It would probably nice if the slicer could differentiate between bridging with and bridging without support interface beneath it. Because I have to disagree here: If you have a support interface and no z-gap, there is no need to slow down while printing on it.

3DTech · June 4, 2025, 2:43pm

Have you done this offset calibration ?

Halbflauschmammut · June 6, 2025, 4:52pm

Now I have. Thanks for pointing this out. The print quality improved now with all your tips, even though I’m not yet there where I want it.