Most desktop fused deposition modeling (FDM) 3D printers these days use a 0.4 mm nozzle. While many people have tried smaller nozzles to get finer detail and much larger nozzles to get faster printing speed, most people stick with the stock value as a good trade-off between the two. That’s the conventional wisdom, anyway. However, [Thomas Sanladerer] asserts that with modern slicers, the 0.4 mm nozzle isn’t the best choice and recommends you move up to 0.6 mm.
If you know [Thomas], you know he wouldn’t make a claim like that without doing his homework. He backs it up with testing, and you can see his thoughts on the subject and the test results in the video below. The entire thing hinges on the Ultimaker-developed Arachne perimeter generator that’s currently available in the alpha version of PrusaSlicer.
We’ve experimented with nozzles as small as 0.1 mm and, honestly, it still looks like an FDM 3D print and printing takes forever at that size. But these days, if we really care about the detail we are probably going to print with resin, anyway.
There are a few slicer settings to consider and you can see the whole setup in the video. The part where an SLA test part is printed with both nozzles is particularly telling. This is something that probably shouldn’t print well with an FDM at all. Both nozzles had problems but in different areas.
I know Tom said he can’t tell the differences between these Benchies, but I can tell you he definitely can. We’ve worked for so long diagnosing printers via bad pictures on G+ back when it was a thing that he most definitely can tell the difference, because I can tell the difference – rather quickly too. He is right about the average person not being able to, certainly. But I agree entirely with the entire video otherwise. 0.6mm nozzles are the way to go, and while right now the slicers aren’t tuned for providing high-detail on these larger nozzles (Cura will select a 0.6mm outer-wall width for example on a 0.6mm nozzle, when you can get away with a smaller one for a bit more sharpness) — they will be tuned for this in the near future. 0.6mm nozzles are also far more resilient to clogging, so overall moving to this method of printing is going to be a huge step up in reliability, and speed.
Hopefully we will see some accuracy tests. I think one designed for .4 but printed on a .6 would be great. Maybe a tolerance test
I wish Cura could print the skin at half line and layer size, then the rest at 0.6.
I’ve been using a 0.6mm out of a volcano the last year and it has its ups and downs.
On the plus side, it doesn’t get clogged as much, and it prints a bit faster. On the downside, cooling is more of an issue, retraction settings are harder to get right, supports are hard to remove and clean up, and most slicers/printers are calibrated for 0.4mm only, so going to 0.6mm requires some work/tuning.
If you only have one printer, I’d say personally stick with 0.4mm; I have a few, and the largest 400mm one has the 0.6mm on it only. I find I can just increase the line thickness and extrude more with a 0.4mm if I want to print more like a 0.6mm. I’ve also tried as high as 1.0, which works well if you want structural furniture or something, but not as an everyday choice.
just my opinion at least
The big thing to consider is you’ve been doing it for a year before arachne was fully released. Imagine what will be possible with the arachne engine, I’m super excited for it.
History repeats itself.
When I started in Reprap it was all about the .5mm nozzles. Then it was realized that the .4 made sharper details. I can certainly easily spot the .5mm (and .7mm) prints that I did in the past.
I agree use a bigger nozzle when details do not matter. This has been well known since 2011 at least. But anyone who says “I can’t see any difference between a .4 and a .6mm nozzle” needs to print something other than benchys with the .6mm. There are many…many cases (especially for customers) when a .5mm or larger would not have had the detail needed in order to complete the job in a manner which made the customer happy.
For things like drone bodies, large prints…etc. I have always been in favor of a larger nozzles to save time….but it has always been clear which print came from the big nozzles. I guess I assumed this was common knowledge.
For many years I have always had on hand .25, .3, .4, .5 and .7mm Jhead mk-vb (still my go to hotend).
I also started out life as a 0.5er. I still have tons of items printed with older printers. They have a certain charm.
Now I have two running printers, one with a 0.35 mm and the other with a 0.8 mm nozzle. And for what I do, mostly functional prints and many of the toys for my son, 0.8 mm is _much_ faster and seem physically stronger. I probably print 70/30 on the 0.8 vs the 0.35.
The real story here is variable-width slicing, which looks to help the worst problems with using a bigger nozzle. I haven’t tried it myself, but I know the problem described — that thin continuous parts become disconnected loops or single lines. It sucks, and it’s worse with 0.8 mm than with 0.6 mm, I’d bet.
So trying that out is definitely on my list, because so far I’ll just use the printer with the 0.35 mm attached when I run into that problem.
Everyone should try out fat nozzles. The price of admission is cheap, and you might just like it. For many prints, the finest resolution you can achieve just isn’t relative, and the overall dimensional accuracy is just the same.
Plus, I kinda like the mega-ribby look of tall layer heights.
Variable width slicing seems interesting but I would assume there are limits. Most of the time when using larger nozzles I was setting line width manually (I used to do it with all nozzles…but I got lazy with the .4).
Personally, I have never had good results setting the line width at/below nozzle diameter…but others may have.
Things like circles are always going to be distorted by the larger nozzle (clearly visible on the benchy smokestack). Circles are limited by the “coil” that you can make due to thickness of the extrude…at some point the extra material on the inside of the curve effects the hole size to a point where it is not acceptable.
Most of the main issues you see when printing with big nozzles is explained pretty well here https://hydraraptor.blogspot.com/2011/02/polyholes.html.
Other than the issue of detail, big nozzles are great. If the slicer can force line widths that seem to be acceptable…that is great. But again, I personally have never had good luck with line width at/lower than nozzle diameter.
Somehow mised the point. Arachne….
I have my doubts that slicer tricks can make a nozzle flow a bead narrower than it’s diameter reliably.
Arachne seems interesting…but not at line widths narrower than nozzle diameter.
It may get away with it by underextruding, overlapping xy and ironing in one step…but I have my doubts that it won’t cause more issues than it will save in cases where you are attempting to use a .6 as a .4mm.
I think it will be nice for varying line width on the fly within a realistic line width given nozzle orifice and gap fill…but will it “make my .4mm nozzle obsolete”….not likely.
I don’t believe there is enough evidence so far to demonstrate anything otherwise.
The entire point of the article is that the Arachne slicer is producing prints with a 0.6 nozzle that have comparable quality to traditional 0.4 nozzle prints. Saying “I have my doubts” when you have no evidence to back them up isn’t worth much of anything. Idle suppositions don’t really stack up against people’s actual documented experience to the contrary.
Now, if you had tried it and hadn’t seen the same results, *that* would be worth mentioning.
Time will tell…
“I have my doubts that slicer tricks can make a nozzle flow a bead narrower than it’s diameter reliably.”
It’s easy to test empirically right now and I can gladly say that the trick does indeed work – I think there’s something being overlooked too. In order to do a 0.4mm line width on a 0.6mm nozzle, you still need to follow the layer height rules for a 0.4mm line width. 0.32mm layer height is going to be the maximum you can achieve reliably while using this trick (maybe less, in fact). So it’s going to be rather easy to blow past that using a 0.4mm layer height, and then thinking you can get away with this particular trick. Layer height still needs to be accounted for, as well as some sort of variable that I’m unsure how to measure immediately – the ability for a polymer to not disconnect from itself at-temperature in order for this trick to work. It’s going to be related to tensile strength, I’m almost certain – but it’s nice to see software tricks, utilize physical tricks in the real world in order to get actual benefits for end users. The limit is going to be pretty small – no getting 0.2mm line widths out of a 0.8mm nozzle or anything, but you should be able to reliably
It only can in some situations and falls on its face in others…. for instance a butterfly hair accessory I printed for a friend. The antenna do not print reliably because instead of following a single line up to the top of the antennae… it does a two line print then tries to extend it with one line… resulting in a gap in the print were it has either underextreded or failed to adhere.
A superior way to print this feature fidelity wise would have been a single line up the center + two thinning lines up the sides but it doesn’t do this.
Problems start to arise when you print models with tight tolerences. Many gears, screws, and print in place models hardly tolerate anything bigger than what they were designed to print with. I sure would love to pump out prints faster on a .6 or .8 but sometimes the slicer just ends up with blank spots or loses so much detail that ths print will fail or be ugly af.
But for big blocky things, sure make it thick.
In the video he claims the software improves the detail and even if that were true: why not use the 0.4mm head and get even better details?
That’s due to incorrectly modifying the model to compensate for errors in the slicer or printer configuration. After that the model isn’t actually a representation of the desired object, but some distorted version of it. That it results in a “usable” print on a particular machine shouldn’t be considered a success.
Width of the bead doesn’t necessarily make tighter parts, if that’s what you’re thinking. All slicers compensate extrusion width pretty well, IME.
Constant-width extrusion _does_ make for essentially radiused corners, though, and that’s really noticeable with a 0.8 mm nozzle. Whether that’s a good thing or a bad thing — if your part is designed with 2 mm radiused corners, you won’t notice. But in those tight angular parts like the inside of a square hole, you’ll see the difference.
But this is also a known phenomenon. CNC routers use dogbones b/c the mill size is finite. Could do the same in 3DP if it were an issue.
(Hmmm… why don’t we? Are there any slicers that will let you add dogbones to internal angles?)
I would like to know what settings in a slicer, such as Prusa SLicer, need to be addressed other than the nozzle size. Temp increase? Layer height? Layer Width?
Width. In Prusa Slicer is under Print Settings > Advanced > Extrusion Width.
The defaults are really conservative. I’ve increased my widths to 0.65 on a 0.4mm nozzle. ~150% extrusion width is no problem and significantly speeds up my “SPEED” profile with no nozzle change
thanks Maave. I’ll give that a shot when I get home from my business trip.
It’s much more than that. In prusaslicer you can compare profiles. Best is to compare your profile with predefined MK3 0.6mm nozzle profile. Especially in prusaslicer 2.5 beta. So many things are set differently for 0.6 nozzle.
I bought myself a 1mm CHT nozzle a month or two back to try. Still haven’t got around to testing it, but I plan to leave it on there when I do, because I am nearly always printing crude mechanical prints anyway.
I plan on getting a resin printer soon for detailed stuff, and have the best of both worlds.
I also have a 1mm cht nozzle that I haven’t tried yet, have you managed to test yours?
I find that smaller text, sliced using Arachne perimeters and printed using a 0.4mm nozzle looks quite bad, the starting coordinates are non-uniform and the surfaces have a poor quality. I prefer a 0.25mm nozzle for this purpose. Obviously for larger rough objects a larger nozzle diameter can be very helpful to reduce printing times.
There’s always going to be a level of detail which is unobtainable without going to a smaller nozzle. But 99% of FDM printer users have no need for it, and reducing print time without losing quality is a much bigger win.
In any case, resin printers are much better suited to ultra-fine-detailed prints.
i like to use a nozzle size that is appropriate for the scale of prints. 0.6 is fine for big prints. for fine prints you still want a smaller nozzle.
Arachne is excellent. Printing features that are only 1 or 2 layers thick is better than before, it’ll tweak the extrusion width as-needed. I can really tune my Joycon rails which are around 0.6mm. I’m keeping the 0.4 nozzle but I increased my extrusion width to 0.65 for additional speed.
and I will only ever need a 1/4″ drill bit so throw all others out
and I will only ever need .035 mig wire so throw all others out
and I will only ever need PETG so throw all the other filament out
OK sarcasm aside this is why i stopped watching and listening to these guys a long time ago. Their topics are nothing more than clickbait. Use the tools that you have and work the best for you and what you are doing. Youtube is a sesspool of horrible clickbait anymore that only very seldom do you actually get some content that will actually assist you in your work.
Did you watch the video?
It’s not clickbait. He’s simply suggesting that you can ditch the 0.4mm for 0.6mm and get the same quality in most cases with the benefit of faster print times
Ah…that’s the rub though.
Make a well known “marginally true” statement in order to “get clicks”. Those who don’t know better come to see the new hotness…those who knew this 10 years ago come to say how “that’s not always true”.
All of which generates views…which is the metric that advertisers use to determine if they want to give you money.
This is the definition of clickbait
Clickbait is negative information headlines: “this one weird trick” stuff where they tell you that there’s something you don’t know and you have to click to find out the thing you don’t know.
What you’re describing is sensationalism.
FWIW Thomas has been performing good research and reporting on it for at least 10 years, so I don’t think his stuff is even particularly sensationalistic.
Not knocking the hustle…but it is insulting to assume nobody is aware.
This is common between all influencers. Not pointing out Thomas particularly. In the past I have seen many good reviews he has done.
Slicer tricks can’t always make up for physics. I understand you can get away with a larger nozzle but IMHO…this is a bad fix. Again…just my opinion based on my testing for many years.
So…yes…stating that 99% of the time you can get away with a .6 is almost certainly false in my case…regardless of slicer settings. I suspect that once most people really play around they will find out that you have different nozzles for different things.
sidenote: I was printing with a .75mm Makergear nozzle in 2011…and yes it is true it is much faster and at times “good enough” but the amount of “adjustable line width” is still limited by nozzle diameter (there will be sacrifices).
Did you watch the video?
I spent 2min reading the article instead….and there is quite a bit I disagree with. So I suppose I will leave it at that.
What If i want fine details like with 0.16 or even 0.12 lh?
I’m with you here – I happen to like having nice crisp outside corners and while it may not seem like much, those 0.2mm of corner make a big difference when printing high-density details such as 1.5mm pitch gear teeth or very tightly kerned text.
Can someone please tell this guy to get another camera man.
Those “artistic” 90% out of focus video’s are completely unfit for comparing the results of the different printer settings.
Quote the best youtube channel existing ‘Focus you F…’
I print with a 1.0mm nozzle for almost everything. But then I am printing functional parts where I want strength and rapid print times. 1.0mm nozzle means single perimeter and very minimal infill if any is still very strong.
Arachne is a BIG improvement, not only does it essentially fix the thin-walls issues, but it means you no longer have to think so much about the thicknesses and print orientations when designing a part that you’ll print single perimeter.
It’s pretty obvious (and I’m shocked!) that a large number of people commenting haven’t actually watched the video.
The video isn’t about using a large nozzle size, it’s about a change in software (Arachne) that allows for larger nozzle sizes while the software is trying to compensate for increased diameter, and avoid the very issues that people (who obviously didn’t watch the video) are saying “You can’t use a larger nozzle size because of these issues!”
Oh internet, you always disappoint.
Same conclusion here. We need an extra checkbox in the comment module, “i have fully consumed the linked info” and then a button for us readers to collapse those comments without that ✅
To be fair, Thomas downplays it a bit too. I mean, given that the huge slicer improvement _is_ the deal, he emphasizes 0.6 mm as some magic nozzle width.
(It might be the goldilocks nozzle, for 90% of prints, though. So he’s not wrong.)
The biggest improvement for me using 0.6 over 0.4 is with PETG…less oozing during moves, liner advance seems more effective and at lower values.
Does the larger hole make for less back-pressure in the nozzle?
I dunno; made the switch recently, still working out the numbers.
0.8mm for 90% of my prints and I’ll go up to 1.2mm if it’s justified. No display models here, only functional parts, so nozzle size and design are always considered concurrently. I design my parts so the walls are thicker than 2 lines and small details are rarely required for the kind of functional parts I print, but obviously it is dependent on the scale and what you need to interface with. Fewer layers the better for strength and with careful design it is rare that nozzle size is an issue.
For anyone wanting to give this a try: you can directly print 0.6mm wide lines with a 0.4mm nozzle. Not 100% of the effect of swapping it, but close.
https://hackaday.com/2020/04/13/using-spiral-mode-to-rapidly-print-enclosures/
Sounds like we need a variable nozzle and it was already invented http://sculpman.com/technology
I found this a while ago and it’s a shame they only work with compaines, especially when it seems that the hobbyist FDM manufacturers don’t really touch anything that hasn’t already been tried by hobbyists months/years beforehand.
TeachingTech made a video about this hobbyist -> manufacturer thing a month or two ago.
Variable nozzle looks very cool! Especially b/c it can print multiple (rectangular) widths in the same print. Very clever!
(I was going to say something snarky about not changing nozzles, just having multiple printers. I mean, I still stand by that recommendation, but the variable nozzle is sexier.)
I’ve been 3d printing for 5 years now, and i’ve never used a 0.4mm nozzle in my life… Only used 0.6mm volcano’s and PETG to print parts.
I had this on a cartisian printer, a delta printer and my latest corexy, and i never had an issue, never had a problem on any of them printing.
What is the filament that he is using? It’s exactly what I’m looking for for a project that I’m planning.