Bambu H2C and Prusa Core One L: a comparison

[skyrippersystems]

Here are some results from prints that I’ve done with Creality ABS on the Bambu H2C and the Prusa Core1 L.

The parts printers are stock as purchased. No upgrades or changes, with more-or less default profiles, on their native slicers. I Used the .015 speed profile on the Core1L, and the .016 standard on the H2C. 4 walls, and for what little infill was used, 3D honeycomb.

The major difference being, I also used the Vortek nozzle for a zero-gap support interface material on the H2C. For the parts on the Core1L, I used organic trees with the default .17 top z gap.

The parts were identical, one being a symmetrically opposite mirror of the other.

The H2C

The Core1L:

Both slicers calculated their actual print times perfectly.
Core 1L, 10 hrs.
H2C, 13 hrs 20 min.

The H2C used more material, because it had to print the rather large prime tower.
Core 1L, 148g
H2C, 193g

Both prints completed without errors. The H2C took about 18 minutes to heat and run through its setup. The Core1L was quicker (less than 10 minutes), but I had started the preheat cycle a few minutes prior to sending the print.

Support removal:
H2C, supports came right off in large pieces. No struggle whatsoever.
Core1L, supports came off easily with a little effort.

Finished part weights were within 4 grams of each other.

Finish quality, top side:
Both parts are excellent, both dimensionally and visually. Easy to proudly send these to the customer.

Core1L, Left, Bambu H2C, Right.

Finish quality, back side:
I had expected the parts from the H2C would be better in the areas where the support interfaces touched the part, but they were only just slightly better than the Core1L in some areas, slightly worse in others. Both parts are very nice, other than where the supports touched, and these areas on both parts are visually distracting.
It is probably that if any of these aeras were dimensionally important, just the mathematics of the zero interface would mean the H2C parts would be more accurate.

Core1L Left, Bambu H2C Right.

In conclusion:
I resliced the H2C part again without the support interface material and a .16 top Z distance, and it reported a print time of 8 hours 40 minutes, material usage of 138g. I didn’t print this part.

So apples to apples, the H2C would have printed this part about an hour faster than the Core1L, and I would suspect that both parts would look and perform nearly identically.

I’m sure support interface material has its uses - but perhaps this model wasn’t the one. For the extra 3.5 hours of print time and 60g of material, the slightly easier to remove supports wasn’t worth the extra time.

[GinBerry3D]

What I don’t understand is that bottom part; mine does it too. It’s like it’s very hot at the edges and leaves them lumpy. Mine does it too with ABS H2C… it’s like those layers aren’t getting enough cooling, but it doesn’t make sense because in such a large piece there should be plenty of time to cool down.

[skyrippersystems]

I agree. That area is especially ugly.

[alitopuz]

Yes, I completely agree with you on this. The H2C unfortunately fails in this regard. I’ve done dozens of prints at a 45-degree angle, and unfortunately, it gives this error regardless of the filament. Moreover, I’ve tried every setting imaginable. When I print with the X1C using the same settings, the print is perfectly clean.

[GinBerry3D]

I think it has to do with the temperature of the bed. It happens with Silk, for example; if the bed is too hot, the filament doesn’t cool properly. Printing with a slightly cooler bed and using some kind of adhesive might improve it, along with increasing the layer and auxiliary fan speeds. Keep the chamber temperature at 60 degrees Celsius. I’ll try to run some tests when I have time.

[alitopuz]

In the prints I made, the bed temperature was 90 degrees for ABS and the room temperature was 60 degrees. For ASA, the bed temperature was 100 degrees and the room temperature was 65 degrees. For PLA, the bed temperature was 55 degrees, and the glass and doors were open.

[skyrippersystems]

The pictures show a finer detail than the parts in my hand and my aging eyeballs do. After reviewing the photos and zooming in, my final conclusion is flagship printer vs. flagship printer, the Core One L printed this part better than the H2C in pretty much every way other than perhaps speed. But the extra hour on the Core One L would have been worth the time based on the final part quality and the better uniform surface finish of the Prusa part.

The H2C has some catching up to do. Both machines used their generic ABS profile,

[GinBerry3D]

@alitopuz

In the prints I made, the bed temperature was 90 degrees for ABS and the room temperature was 60 degrees. For ASA, the bed temperature was 100 degrees and the room temperature was 65 degrees. For PLA, the bed temperature was 55 degrees, and the glass and doors were open.

And they all have the same problem? Then it’s something to do with the layer and auxiliary cooling.

[skyrippersystems]

These are ~45 degree overhangs. Do you think it could be a speed issue? It’s replicated all over this part where that condition exists.

[GinBerry3D]

@skyrippersystems

The pictures show a finer detail than the parts in my hand and my aging eyeballs do. After reviewing the photos and zooming in, my final conclusion is flagship printer vs. flagship printer, the Core One L printed this part better than the H2C in pretty much every way other than perhaps speed. But the extra hour on the Core One L would have been worth the time based on the final part quality and the better uniform surface finish of the Prusa part.

The H2C has some catching up to do. Both machines used their generic ABS profile,

Is the ABS the bamboo kind? Or is it a generic one? This happened to me when using the black color.

I’m going to send a photo of a part that’s printed with the settings that Bambu specifies for their white ABS.

[GinBerry3D]

@skyrippersystems

These are ~45 degree overhangs. Do you think it could be a speed issue? It’s replicated all over this part where that condition exists.

Could we do some testing… with a small part?

[skyrippersystems]

Well, the Core One L did this area pretty well. I could look at the speed and cooling for both machines and see what they do different. Let me check.

[skyrippersystems]

Core 1L speed:

H2C speed:

Fan Speed, Core 1L

Fan speed H2C:

[skyrippersystems]

Fan speed is a whole lot higher on the whole print for the H2C. Could it be blowing the newly laid layers around in these overhang areas as it’s being put down?

[hotellonely]

yes it’s a game of micro adjustments. either reduce or increase cooling for overhangs.

[skyrippersystems]

I don’t play games, man. Prusa got it right.

[GinBerry3D]

I’m sure a few minor adjustments could improve it, but the H2C still needs patches and tweaks to the slicer, and quite a few of them. I hope they’re working on it; the X1C came out more polished from the factory.

[Bullocks]

Those “lumpy” edges, if you watch them form in real time, are usually soft and molten as the extruder passes by, and curls up as soon as it leaves. Next layer pushes it around some and does the same. Etc etc. imo it needs more cool or more slow, or both, depending if the part will take more fan or not.

But overall very nice parts and a good quality compare thanks

[skyrippersystems]

Okay, I reprinted the lower section that was so bad, and reduced the max fan speed from 80% to 50%

Result: Top, original, bottom, reduced fan speed.

I did get a warning that I should turn a nozzle glob sensor on when I made the fan speed change.
No I have to decided if I want to print the whole part with this setting or not….

[GinBerry3D]

@skyrippersystems

I did get a warning that I should turn a nozzle glob sensor on when I made the fan speed change.
No I have to decided if I want to print the whole part with this setting or not….

That warning pops up all the time now; I got it too when printing nylon. I would definitely try printing the whole piece; it looks better now. And if you lower the speed a bit on the overhangs, I think it will be even better.