I finally got my hands on one of the Obxidian nozzles in 0.4mm.
Reason for wanting one was NOT the impressive flow rate but the overall much better print quality and fewer hassles as observed by so many happy users.
Of course I eventually wanted to see how fast this thing can actually print…
Not really much faster than a standard hotend in most cases…
So I started to tune the 0.2mm default profile to get the most out of it rather than sticking to the defaults and the speedy button.
I can print a GOOD quality benchy in about 28min at 0.2mm layers.
So I started to wonder how it can be THAT slow if the preview shows speeds of well over 400mm/s…
Bambu shows the max POSSIBLE speed for a layer, not the overall speed.
Means things like acceleration, retraction and so on had to be tuned a bit as well.
And while all this resulted in a slightly faster benchy the improvement still was not worth the hype.
Then I started to pick some models where the FLOW RATE was a limiting factor in the past, like for combined infill and the resulting speeds.
And here the Obxidian really took off and maxed out the 39 set for the max flow rate.
The prints still came out in exceptionally good quality…
But with that I observed another conundrum…
For supporting models I have no issues printing the outer walls at 500mm/s and I can do the same with most other print moves, except the critical ones of course.
The ACTUAL reduction in print time however still does not add up here.
For example, a model sliced at a moderate 250mm/s as the overall average compared to the same at 500mm/s DOES show the change in speed and flow rate in the preview…
But if you expect the print to finish in half the time you are mistaken.
As far as I can judge it and that by using very forgiving models there is a reduction of about 20 to 40% possible depending on the model.
The limiting factor however seems to be the firmware.
Yes, the max is limited to 500mm/s but the slicer does not know that and is happy to slice at 800mm/s or more.
The preview is pretty much useless to get a good judgement as only the predicted print time gives an indication here.
I found it not worth the wear and tear to go overboard with the acceleration values for example.
If the machine sounds like it is about to fall apart than you are past the limit already LOL
Which made me wonder if our machines are so limited because that is what Bambu hotends and extruders are capable of or if it is matter of the build and material quality not allowing for more…
Something as seemingly unimportant as a hotend can make a world of difference to 3D printing.
After all, they just heat the filament and let it come out a tiny nozzle…
Could not resist and had to clean the Obxidian so I could take a proper look inside - should have done before using it, I know…
What can I say…
I looks awesome in terms of finish and quality.
All is shiny, slick and there is no imperfections to spot anywhere.
On the other hand the Bambu hotends look like they came from the last decade but not the current one.
Considering that printer manufacturers are certainly aware of the importance of a good hotend I can’t help but think Bambu went intentionally cheap on us.
I was actually proud when my first 3D printer, after lots of tuning was able to print a benchy in just under 2 hours - so many old memories LOL
So doing this with an average print speed of around 200mm/s sec is more than great by comparison.
Bambu however came out of nowhere to provide printers in top ranks for speed on a consumer level.
And all other manufacturers came out with similar speeds in the range of 400 to 650mm/sec.
None of them however also came out with a matching hotend and extruder…
The all only maxed out what a standard hotend is capable of and it seems they made sure the other hardware won’t be able to do much unless you want to risk things falling apart…
If you can accelerate from stand still to 500mm/s instantly sure but 9th grade physics assumptions don’t work with real word physics
even If you send a gcode w >500mm/s movement, the machine has hard limits and will cap its max speed accordingly so you aren’t pushing more than it allows you to. Might be a good idea to retension belts if you hear belt slip/skip sounds on max speed/accel
when you get a e3d hf obx nozzle, limiting factor becomes the heater. Then when you put 90w+ heater on it you run into the extruder limit of either 78 or 79mm^3/s afaik which is already set in bambu studio
I do realise that physics exist, so I am not questioning how the machine runs…
And your heater comment is spot on!
I had a home made hotend for my printer two generations ago.
In terms of years, please don’t ask as it was too many and I am feeling too old already LOL
I did it mainly because I used Nylon filament from the hardware store for most of my prints.
There wasn’t any filament of the real kind available.
Anyway…
I modded the maiboard the quick and dirty way by using the mosfet output for the hotend heater to drive a DC SSR.
The SSR in return supplied 24V DC for two heater cartridges with 50W each.
TOTAL overkill but the point was being able to keep an even temp in the oversized melting chamber with very fast reaction times.
I just wonder why Bambu did what they did in Studio and so on.
I would really like to have a realistic preview and a smart enough UI/AI that takes care of speed and acceleration settings based on max flow rate and quality settings in conjunction with the type of filament.
Bambu is half way there but for some reason just stopped where they could show some great improvements.
Probably thanks to their copy/paste approach for Studio.
I really does not make any sense trying to max out the printer based on that these Obxidian nozzles can handle in 0.4mm.
The hardware simply can’t keep up as it was never designed for those high speeds.
But we are now at the point where hotends are finally good enough to be used for constant flow rate printing.
Speed is actually quite relative this way…
What you might loose compared to traditional slicing in terms of print time is more than made for by the improvements in print quality.
Bambu is half way there with how they handle setting for printing and filaments…
I hope they will make the last jump soon…
Actually, to me it sounds like Bambu perfectly matched capabilities of the 0.4mm stock hotend and the mechanics. They rigorously optimized price at the desired print quality. If an old school hotend is sufficient for that why should they have used a pricier solution? Btw I think the hotend is not that standard at all. the very slim long melt zone and the flat heater help for low thermal resistance from heater to heat block and at the same time low thermal mass. That allows to quickly react to changing heat demand, e.g. because of very dynamically changing speeds. Overall that should help to keep the extrusion temperature much more stable than your average e3d v6 hotend.
Of course the obxidian outperforms the stock hotend by a big margin, but at a prohibitive price as a series solution.
In my opinion, if you want faster prints, the Obxidian only makes sense in 0.6mm, not 0.4mm. That way you can make use of increased flow while the mechanics still can keep up.
Regarding the preview in Bambu Studio: I guess it is just the code from Prusa Slicer, which doesn’t have what you ask for.
I agree that it would be nice to see actual speeds including acceleration etc. But I don’t know of any other slicer that provides that. So while that would be really nice, I don’t feel it missing to be a flaw. Maybe computing and visualizing acceleration would take excessive time and therefore be unacceptable for most users? No idea, but that wouldn’t surprise me.
Good point actually…
Unless cloud computing is used this could put a severe strain on the CPU to create a meaningful preview…
Unless there would be a way to do this over the GPU I don’t think it is feasible on standard home or office PC’s…
Try the BIQU E3D Revo Hotend for Bambu. It has a 60w heater. Also provides the quick change nozzle system where the standard E3D and high flow E3D oXidian nozzles. Once you dial in your filaments from the change, it works really well from my perspective. Alex_vG is right, the heater as 40w is a big limiter in the stock and E3D direct replacement. I have a Slice Engineering Mako coming soon as well, looking forward to see that brings to the table as their Mosquito hotends are amazing.
Without cht I don’t see any performance benifits, especially thermal with the heater being higher up the hotend. Also the cooling zone is a lot smaller.
If they get the same performance as the stock nozzle the higher (claimed) robustness and the ability to just change the nozzle could be good enough tho.
Revo has the same problems, the confined space and length limits the design space and the only open lane is the way bambu did it.