Have you ever noticed that you printed TOO FAST? Oh, and check out our CNC Kitchen products at cnckitchen.store/ or at our resellers www.cnckitchen.com/reseller
Nice work, just a small thought, if the nozzle need a longer distance for the filament, then maybe the path of the filament in the nozzle could be changed to a circular downward path and end in the tip. A bit like the heat pipe in a water tank.
@Stefan, You asked for an idea for your test pieces - how about Archimedean chords as 100% infill and 0 perimeter? This should allow you to reach the speed more likely. General thoughts on flow rate: When I have many identical parts to print, I often take the effort to set the layers to a different temperature for increasing the flow rate even further. It would be nice if there were a function in the slicer for this - a temperature dependency on the flow rate, or, in reality, a temperature compensation, because our material does not reach our set temperature. Perhaps we have a deviation between the measured value in the hotend and the actual temperature in the material due to the high extrusion rate. If you follow this idea further, it should be possible to establish a relation between the heating element's power and the extruder's feed rate, taking into account the thermal conductivity to consider the resulting delay in temperature correction. Therefore, I do not believe that these settings can be adjusted via firmware, as temperature changes must be made in advance. Unfortunately, I am not a programmer, but I am really interested in whether this could be used to tune the flow rate of ANY 3D printer without making changes to the hardware. Tüdelü 👋
To print stronger functional parts faster I use a tip from one of your earlier videos - print wider lines. For a LOT of my printing these days I'm using 0.8mm lines with a 0.4mm nozzle. With Cura 5, thin walls don't suffer, since the slicer will generate thinner lines when it needs to. You still need a higher flow hot end, of course, since printing wider lines at the same linear speeds means more plastic per unit time. Luckily my Artillery Sidewinder X2 has a Volcano hot end.
I printed my Trident 350 parts with my X1 using ABS+ on generic ABS settings. I didn't notice any part strength concerns during assembly, but now I know to do PM checks more diligently. Thanks for the info Stefan
@@gustavbondeus6767 I know, I tested both a monthe ago, clone and original (two monthes ago), and come to the same conslusions, I even get in touch with bondtech about this and they was pretty interrested by this sucject ;)
Great video! For what it is worth, my old Fortus 250 (industrial printer) runs ABS at 305C and has a meltzone nearly 40mm long. And that machine has extremely tight temperature controls and uses quality thermocouples for temperature monitoring. Because of that I have become way more comfortable with pushing Temps through the roof while printing.
Amazing video (as usual) Stefan. I'm glad that I can always depend on you doing proper experiments, because SOMEONE in the community needs to do it, and I don't think I have the stamina or work ethic to do this level of experimentation :) Keep up the good work.
When I first got the X1C I banged out a few parts at the default settings, and noted that the print quality was pretty good. Then I cut everything in half (from Bambu's defaults) and discovered this made the print quality *amazing*. As someone who's mostly printing ABS and then sanding and finishing it to look like not-3d-printed, running the X1C at 50% of nominal and 0.12mm layers still gets me parts in half the time the Ender-5 did, and those parts end up needing about a quarter of the sanding and finishing work that the Ender-5's production did.
@@justinmurray8582 The standard 0.4mm. I'm usually printing at 0.12mm layers and trying to capture fairly small details on many things, so a larger nozzle isn't worth it most of the time. The exceptions don't come up often enough to bother with building up a hotend with a bigger nozzle.
I just got my first printer and was trying to think about what was causing the change in gloss between layers when I stumbled on your video before bed. You saved me at least an hour of parsing through unhelpful reddit threads.
Still a newbie and a former machinist I thank you for insight into issues I've been experiencing with my X1 printer and it's settings. My Ender 3 S1 Pro with Sonic Pad tuned produces faster, good finished and parts of solid structural integrity of a phone holder, clip style. Using the same material, settings (temp/flow) X1 produced gummy weaker parts that failed under the stress of the design. Again being new (December 2022) I wasn't understanding the "why's" for such failures. I will be looking into your suggested mods and settings. Thanks for your insightful efforts.
I’ve had a lot layer adhesion problems with Polymaker ASA with stock Bambulab ASA profiles. Ended up slowing things down and turning down part cooling to get acceptable strength. Definitely going to look into this mod! Thanks Stefan!
Thank you for the scientific and especially concise context. Everyone should draw the necessary conclusions for themselves. Personally, I find 12mm³/s to be completely sufficient. Great video!
I really love the content made by this channel. Steffan (I hope I spelled it right) is not only a great engineer, but also a very good teacher. I learned a lot from this channel.
AWESOME video!!! Upgraded to 0.6 nozzle and print quality went to the toilet! Thanks for the tip on the extrusion rate, will try it! Thank you! Really appreciate your work! Hope one day someone makes an after market hotend that can use standard V6 nozzles!
You can replicate almost all of the benefits of a 0.6 nozzle by changing extrusion widths to 0.6 while using a 0.4 nozzle. You lose the ability to do layers that are 0.4mm+ but you gain being able to do 0.1 or less. You also get much better quality. In almost all cases, the 0.6 nozzle print speed increase is a result of the larger widths, not the higher maximum flow rate. It sounds like you might be pushing that. The cht style nozzle has a significantly higher boost to flow rate than going to a larger nozzle. Same for the volcano style setups. The trick is to print your outer layers at a smaller width and the inner layers much, much larger. I'm able to get away with half of my prints times are typically from the external perimeter, even with infill values of 50%. You almost can't see the layer lines. Only the seams are readily visible.
Excellent video Stefan. I really think Bambu has hung their hat a little too heavily on "Speed" and they need to tune in "Quality" presets. My own tuned profiles are significantly slower but more reliably produce better looking & stronger parts (anecdotally). My X-1 gets a lot of use with these profiles and I'm much happier for my needs.
Hi there. Can you share your updated profile settings for different material PLA,TPU,PETG ,etc? It would help a lot of us that are getting into 3d printing with this Machine. Thank you.
Bambu Labs said some months ago that they will be releasing their own *high-flow* nozzles. Hope they offer hardened nozzles for abrasive materials, as most current high flow nozzles aren't hardened.
Wondering if they will also release a new nozzle design similar to the A1 series that doesn't require a full hotend. Maybe on the next X1 release? Hopefully also backwards compatible.
Awesome video. Made pretty much the same experiences with the aftmermarket hotend + CHT. Currently my go to hotend + nozzle combo for the x1c. Looking forward to a hardened version.
Finally. A video about something I realized a long time ago from many videos prasing klipper-enabled machines and their ridiculous speeds. But I still did not hear an explanation of how the increased part cooling fan speeds, that comes with the increase in print speed, affect layer strength...
Awesome video as always. What this actually makes me wonder, is whether the good old 2.8mm filament might be due for a bit of a comeback. Changing the diameter of the filament affects the feed rate by the inverse square of the proportional change in filament diameter. Meanwhile the distance from the hotend wall to the center of the filament only rises linearly (but so also does the surface area available for heat transfer). So the question basically becomes: Is plastic such a bad conductor of heat that fully melting a 2.8mm filament takes more than 2.56 times as long as 1.75mm filament. If that is not the case , the achievable flow rate for 2.8mm filament should be higher in a heating zone of the same length. Thinking about this further this is probably one of the reasons why for bigger extrusion systems the pellets are usually not melted by heating them externally but rather through the friction imparted by the feed screw. In a system like that the flow rate of material and the amount of energy transferred into the material should be largely coupled as long as you adjust the turn rate of the screw to always have the same pressure in the system.
As always, an objective no nonsense fact based scientifically tested approach to solving a real problem. This is my go to channel when I need to research issues related to 3D printing.
Ah I have noticed this as well up to this point I have just been setting my print rate to 50% on the X1 for my "quality" prints but controlling the volumetric flow rate is way smarter. Thanks for the tips and the research!
When I make parts for press fitting bushings , I use concentric infill to reduce accelerations and keep print speeds faster at 100 percent infill. Should help with your coupons too.
I have bought the hotend and nozzle linked in the description and was able to reliably get to around 38mm^3 before visible changes between layers. Incredible upgrade for my P1S!
I have had a smoothieware delta for the last about 7 years. It has been my workhorse till it was damage in a earthquake. With the stupid amount of time I have put on this printer (wish i had a hour meter but just not possible at this point). It was the first printer I owned with the capabilities of going into the 200mm/s - 250mm/s range. I feel like your findings mirror what I learned over the years with that printer. If I wanted to print fast, part strength often suffered, and even details at times. I would often run it much slower than it could go, even more so when printing smaller plate volumes. You might also want to consider revisiting annealing the parts that you found previously weak. Not as big of a deal on PLA. But for ABS, and Nylons you can see noticeable differences. Its more noticeable on these materials if you run a high bed temp and chamber temperature. As you can experience annealing while printing, when doing multi hour prints in these materials.
After pushing my hbot to max extrusion rates my parts were coming out all matt and I really liked it but never knew they were weaker. thanks for the info!
I love your practical scientific 3D printer testing. This is very useful information, and you provide a great education on important subjects most people don't consider, allowing us to understand our 3D printers and the 3D printing process to gain optimal results. Thank you!
Really interesting flow tests! Great to see those aftermarket options performing so well. The standing tests are pretty useless though. Since they’re in batches, the layer time is many times higher and there’s lots of time to cool down. Printing 12 samples 200mm/s is going to end up extruding on plastic pretty much just as cold as 12 samples at 50mm/s
Another nice video that's very informative. I've already ordered the replacement hotend so that I can use my existing V6 nozzles in my X1C if I want to but now I'm ordering a couple of those CHT nozzles made to match the Bambu Lab nozzle length. Yes, the replacement hotend/heatsink with removable/changeable nozzles is expensive to start with but when you consider it can used for many inexpensive nozzles rather than having to replace the comparatively costly Bambu Lab options every time your nozzle wears out, they end up much less expensive in the long run and offer more variation. That and they make changing nozzle sizes/types much faster, no more having to break down the whole carriage assembly and swap entire hotends just to change nozzle size. Bambu only offers stainless steel and hardened steel while V6 style nozzles come in brass, plated copper (my favorite) and various specialized nozzles like ruby, titanium or obsidian plus many other variations/combinations of tip style/metals. I'll never need most of those since I normally use plated copper or hardened steel when it's needed but I like having the option of using my collection of plated coppe/brass nozzles back, not to mention being able to use the V6 style hardened steel nozzles that I already own rather than having to buy Babmu hotend replacements. I'll keep ordering the Bambu Lab hotends for the hardened steel nozzles when I need them, the AliExpress aren't much cheaper and I trust the quality of the Bambu slightly more. AliExpress can be great but you never know what you might end up with (off center holes, poorly machined surfaces/threads, low quality materials, etc.).
Polymaker high speed pla is a good option to handle the high speed issue. This was an excellent video that showed me where to focus on these issues on my Bambu.
The standard Tips & Tricks of successful FDM printing really seems to change once you cross that 100mm/sec threshold. Thanks for starting us on the proper education!
About avoiding crazy acceleration rates: all you care about here is relative performance, so you don’t have to be constrained to your standard coupon geometry. You could make layer-adhesion coupons with much larger diameters in the critical region, which would give proportionately lower accelerations, or conversely higher speeds for the same acceleration. To keep the forces in the range your universal test machine can handle, you could make the structures hollow. Either start large, using a different holders on the machine, or use the same clamping diameter, but expand larger and solid right above/below the clamps, then go hollow and neck down. (It’d be easier to explain if I could draw it here, but I think you’ll get the idea easily enough. When you start making the hollow part, widen the internal opening slowly (similar shape to the exterior of your current coupons) so it won’t need any internal supports.) There are obvious limits, so I’m not sure you’ll be able to go to a large enough diameter to make a difference (the total cross-sectional area of the neck in your current coupons doesn’t give you a lot to work with), but it should help some. - And given that your machine can handle tensile testing in the in-layer direction, it might be strong enough to handle a much larger cross-layer cross-section. (Again, for this test you only care about relative numbers, so don’t need to worry about comparability to your standard tests.)
Such an informative video. Thanks! This is a worthy upgrade for a printer that is possibly the most capable in the market at the moment so thank you for showing this.
So now we need super slicer to let us vary extrusion temperature based on volume flow rate with a comp value for heat up time just like with the fan. I’ve wanted this feature to lower the temp during bridges for a while.
Awesome video :). always get a bit annoyed when people only talk about how fast they can run their steppers! It's fun to do that but practically matters for me personally
Take into account every possibility for practicality. Fast and light prints are amazing to test fit something you've designed before moving on to a full strength design
Stefan, I always enjoy your videos and the dedication you put into making them. Here is an Idea for a video. Testing linear systems, for instance difference in drag between genuine Hiwin rails vs Aliexpress ones, difference between smooth rods and linear rails, difference between carbon fiber rods vs smooth ss rods vs linear rails. Also maybe how big of a difference does lubrication of choice affect smoothness of gliding. and in the end which solution is the best regarding weight to least amount of drag during sliding.
Thanks for the explanation and test. I confirm using the default bambu ASA setting results in poor layer adhesion. I was able to get acceptable results by setting the speed to 50% of the preset. I feel your suggestion of limiting the max flow rate is a simpler and more reliable way to achieve the same.
For the strength test on the fast speeds, you could increase the radius so that it has time to get up to speed. Then you could use your lathe to get the profile correct, while cutting out the acceleration/ deceleration zones
For better high flow nozzle tests. Instead of printing the hour glass shape you could print cylinders of the major diameter, then machine them down to the hour glass shape. Add what's called a live center recess on one end and make them slightly longer so they can be chucked up in a lathe and supported on the opposite end. They do make desktop CNC lathes that can handle this easily. I am a career CNC machinist, if you want any help with programming and setting it up I would be happy to help. But, I have a feeling you could figure it out on your own. good luck and keep up the great content.
Nice video, thank you 😊 why do you want to avoid drastically increasing temperature? I have been doing this for years with fat nozzles without changing to volcano. Works for this printer as well. I go 20-40 °C above max recommendation for nozzle temp. Printing 20-30 mm³/s for PLA, ABS, ASA, PETG, PCTG with hardened 0.4 mm nozzle . Perfect strength, shiny as always 😊
I've been slowing down print speeds on anything "decorative" to 50mm/s on only the external layers for my bambu x1c. While it definitely slows down prints (but still faster then most printers), it makes parts way nicer and eliminates any of the color changes.
bambu should give the option to bulk purchase parts like costco. the key to printing fast is lower cooling and flow. high flow 0.8mm clones on all three of my machines. love that transition into the sponsor message.
I've thought about posting about this. I'm glad you are sharing. Tuning to a flow rate is what I've always done, since my super racer is faster than most materials can handle. So what I do to utilize the speed is maximize the flow rate and then lower my layer height until I either hit my max speed or smallest layer height.
I've noticed this after printing 330 hours with my bambu x1c. Been printing with PETG a lot and when i switched to PC the performance were worse. I was trying to anneal the prints, with PETG , to compensate for that with SOME improvement but not enough. Test done with ERYONE PETG 255°C hot end, 200 mm/s (standard profile) 10% Fan MAX. So the parts printed with PETG at those settings are now much more brittle than 60 mm/s and tend to break like PLA. Instead of deforming and yielding like PETG printed on a ender3 at 60 mm/s (for example). I absolutely want to try the 3rd party hotend with the CHT "like" nozzle. THANK YOU STEFAN!
I wonder what that stabilization systems are like in the Bambu Labs printers, and other printers. I think that you could use centrifugal force to change the direction of the movement/inertia in the Printer Head to keep it from shaking. Kind of like how it's easier to move your hand quickly in circles or ovals without moving your body, than it is to move your hand quickly left to right. Great video, Stefan! I hope Bambu Labs listens and learns from your experiments and findings.
@@Nerlin They would be stupid to not develop and offer drop in replacement highflow units. Most people who have bambulab printers dont want to muck about with aftermarket stuff. Just being able to buy the solution directly would be very advantageous.
Great video as usual!! My existing printers are dialed in and working well. I decided to sit back and watch (and save up my money) as this next generation of printers like Bambu are perfected. Then I will jump in and enjoy the benefits.
ive been waiting for this video after seeing all the speed benchys, really interestign result and great production as always. might be worth trying out a thin "wall" like sample or a vase mode based one in an attempt to reach the higher speeds
A good way to avoid the acceleration issue in the test coupons would be to use perimeters for the primary fill of the part, as the concentric pathing will avoid the harder acceleration points based on the linear infill, and will be more consistent as it avoids the inevitable changes in extrusion speed required to keep consistent bead width when doing harsh direction changes.
I am super jealous that you can just take a hobby, perform detailed analysis of it, and make a living out of it. I really appreciate your videos and your dedication to transparency and detail. Well done.
CNC Kitchen, can you do a vid on price to performance with filament? like what is the cheapest filament that still prints decently or what is the best filament for high accuracy. I'm still pretty new to 3D printing and there is a lot of brands out there and I have no idea what I should buy.
For strength tests, I'd be interested to see what results you'd get if you used higher extrusion temperature for your high print speed tests! And then higher enclosed ambient temperature on top of that!
Yeah I am pretty sure you can solve these kinds of issues with a higher wattage heater cartridge & turning the temperature up which decreases the time needed to melt the plastic. A copper heat-block also helps because it both increases your heat throughput and reduces the lag between the heating cartridge and the hot-end. I print at about 23 mm3/sec and don't have any issues with an e3d v6. It's like $20 in upgrades.
Thx for proving what im saying since the start of "fast 3D print" Speed is Not the goal ! What we realy want is to have as great as possible cubic mms of material ozzed in one second for as litle speed as possible to reduce wear, vibrations ect..
Can you please do an updated "getting started" video for newbies who just want to print things like shop vac adapter parts, or casings for drill bit organizing or the million different clasps and hooks and strange shapes that would be helpful for wood shop organizing? I've never used CAD software before - I'm looking for something like "buy this printer, with these accessories, install them this way, use this software, buy this filament for X job, this filament for Y job", etc
As someone who has been trying to decide between a Bambu X1 and a Prusa MK4 this has been really informative. My main concern is print quality, strength and dimensional accuracy so I am starting to lean towards the Prusa..
You could always just like, slow down the bambu a little bit. Even with it set to a slower speed it'll run circles around the prusa in terms of print times.
@@kylequinn1963 and quality aswell most likely, plus no printed parts and the X1 has an enclosure, actually enabling you to print ABS and other harder materials out of the box if you needed to you may think you never need to, and still wonder why you're running a shitty open printer one day, and i Know that for a fact out of first hand experience
Consider whether you ever want to print engineering filaments (ABS/ASA/PC etc.). If you do, go for the X1 as the chamber makes a huge difference to the printability, quality, and strength of such filaments. You can always slow down the defaults too... The P1P is really the competitor to the MK4.
Bambu X1 is way better machine than prusa, just the build quality is way better. Amd you always can slow it down. Is like having a ferrari, you can drive it at legal speeds and you will have a better car than one guy with a fiat.
wider nozzle sizes allow more material to lay on each other at a heated temperature for a longer time, making it stronger. They make the . This is why even with ultra high flow nozzles, the layers need time to cool down and not too fast, otherwise the part will collapse on itself. Have a look at the approach of filament innovations in the US. The make 2.5mm nozzles with very long heat blocks on they typhoon extruder and still produce very strong parts.
*The rule of thumb is very simple* 1) When slicing, check if the flow is less than your hotend is capable. For example: V6 is capable of 11 cm3/s, Volcano - 18... Your flow can be calculated as simple as this: feed speed (mm/s) * layer hight (mm) * extrusion width (mm). For example if you have 0.2mm layer hight and 0.6mm line width, you should not exceed 18 / 0.2 / 0.6 = 150 mm/s (for Volcano) 2) Never bump up the temperature to increase speed. If you stick with these rules your plastic at the nozzle tip will have the same temperature all the time at any layer.
@@coltenmeredith8899 If you do this to increase the flow capabilities, your printing will behave differently in different parts of the print: if the printing is slower (or line is narrower) your material will be hotter than desired and you will have glossy surface and worse overhangs, maybe even material degradation. If your flow is not exceeding some certain value, your material will have exactly the same temperature out of the nozzle. Which will result in reliable printing.
Yes indeed: printing speed does have an impact on part strength and there are two reasons for this. Lower speeds means there is more time to heat up the material so it can reach a more uniform temperature from the outer surface all the way to the core of the material but it also allows the nozzle to act more as an iron when the flat ring of the nozzle tip irons the extruded material onto the previous layer. I still find it strange that one nozzle geometry is used for all those different materials. The area of the nozzle tip and the length of the orifice play a major role.
Such a great video! And thanks for linking to our blog post. Pretty cool to see that your empirical tests line up with what my thoughts were after reading those papers 😅
Excellent comparitive analysis, thanks to the hardwork on the same❤👍. What I feel might be the issue, is the max power capacity of the heater, and also the quickness of the PID system controlling it. That being the reason for better performance at elevated temps. So, the heating element itself should be beefed up, along with a more responsive PID, coupled with a higher output power supply for the heater element. Essentially, it boils down to the max possible volumetric melt mass within a given time duration. Just some 💭
Thank you Stefan for another super informative video. The work you do is so valuable. I just bought the new Ankermake M5 and have experienced the same issues as addressed here. The Bambu may be my next and having this info makes the decision much easier.
This explains so much. I was wondering why some of my parts were coming out with inconsistent surface quality. I have been tweaking the max volumetric rate, with better results. But I still didn't understand why it was happening. I love any excuse to mod my printers, so I guess it's time to make some aliexpress purchases!
Danke Stefan für die Bestätigung. Ich hatte gerade genau die gleichen Problem, vor allem mit ASA. Die prints sahen gut aus, waren aber einfach zu schwach und ich war sehr enttäuscht von ASA. Schlussendlich war meine Lösung genau die: langsamer drucken und weniger Kühlung. Die Standart Profile haben einfach zu viel Kühlung nur damit schnell gedruckt werden kann. Gerade wenn man mehrere Teile auf einmal druckt, wird der X1C sehr schnell, aber die Teile sind einfach nicht stabil gewesen. Also Kühlung runter. Ich habe es aber vor allem über die layer time geregelt. Also längere layer time und damit den Print künstlich verlangsamt.
TBH, I'm totally fine with the stock nozzles from 0.2mm to 0.8mm. I was considering aliexpress nozzle, but I got used to swapping nozzles enough to not care about dirt and hotswapping with risks of burn. I rarely move my x1c to sport more, and only twice tried ludicrous for lols.
Came across this whilentrying to piece together why the bambu labs printers are so fast. At its core 3d printers melt plastic and deposit it in layers. My hesitation in buying one has been in figuring out if they just print such that the parts look good or if they are able to achieve excellent layer adhesion as well. Thats something that i havent seen ANY other channel address at this point (may 2023). Thanks!
This speaks to why I haven't upgraded to a faster machine yet. I mostly print in ASA for the heat and sunlight resistance. I more often slow down a print for better layer adhesion, particularly if it's small, rather than speed up. It seems like for functional parts, particularly printed in materials other than PLA, high speed printing isn't very useful. I would like to be wrong about that, though. It would be nice to print faster.
@Stephan- have you ever taken a look at temperature accuracy on your X1C? I have spent well over a month going back and forth with Bambu on this, after finding that the hotend thermistor/electronics report back a temperature about ~20C higher than reality, resulting it a "real" hotend temp of only ~250C when you ask for 270C, etc etc. This means the default 220C profiles for PLA are trying to push it out at only about 200C. It's not just a hunch- I've tested with multiple replacement thermistors, different (official) hotends, settings, countless test prints and meandering flow loops and more, and using both a thermal camera (admittedly not ideal for shiny objects, but calibrated using a Mosquito magnum :) and ALSO with K-type thermocouples on different parts of the hotend body (even in the same hole as the bambu thermistor and with either their thermal grease or even with boron nitride paste.) Heck, they even sent me all new electronics for the TC board (the interface board on the toolhead that everything plugs into) and I still see the same thing. I don't know if it's just me seeing this, but I hadn't really seen much discussion in this specific area, and had been meaning to reach out to you for a while now. So, if you set your printer to say, 230C, and sticking a cheap K-Type thermocouple coated with some thermal grease onto the hotend gives you a 230 +-1C reading, then you only have to worry about the "well known" hotend performance limitations that you described very well in this video. But if not, then you may find yourself having to do the constant mental math I've been having to do for the past month or so... By the way, Bambu has only kind-of-sort-of acknowledged my findings (after sending them multiple videos, test reports, troubleshooting sessions, etc) but I'm still missing the big "ahha" from them on this. They still seem to give off vibes that they think I'm just a crazy guy nitpicking things :)
I've felt, with no scientific testing that I was missing something about the temperatures on these machines. If you are right it would make sense to me based on my experiences. I never see stringing and it's like I'm printing at lower temps than I should be. I'm gonna crank the temps up and see what happens.
For the adhesion test i you could do single or multi-walled cylinders and print in vase mode. Since you have a cylinder layer adhesion might still be the weakest link. might be worth a try. The bigger the circle the less the acceleration limit.
Most of my profiles from Bambu for ABS/ASA already have max flow much lower so avoid this issue... clearly bambu keeps updating based upon feedback, which is great.
The video considered the extrusion rate in mm^3/s, length of the melt zone, nozzle design and hot end temperature as variables impacting the temperature of the material exiting the nozzle but that's only half of the issue. The temperature of the previous layer is also critically important to interlayer bond strength. Instead of printing 14 pull test samples at once, allowing the previous layer to cool, try printing the test samples one at a time so each layer is extruded onto a layer that hasn't cooled as much. I print small structural parts in TPU. I have the slicer print them sequentially and the parts are ridiculously strong because at 50 mm/s print speed and minimal cooling, hot TPU is being extruded onto very warm TPU, rather than printing hot TPU onto cold TPU. It would be interesting to have a slicer setting that adjusted fan speed for uniform part cooling depending on print speed and part geometry for more uniform results, hopefully avoiding matte finishes in some areas. I believe so much effort has been devoted to adequate part cooling on the new generation of fast coreXY printers that little to no concern has been expressed for excess cooling in some instances. Maybe the problem isn't printing too fast, but rather cooling too much.
I don't want to sacrifice quality for speed in most cases. I like 1) reliability 2) repairability and low maintenance 3) and good print quality. Speed is not my first desire in a 3d printer although it might be nice to have one where print quality isn't important.
Printing fidget toys, I haven't noticed the matte vs glossy too much. Possibly due to the printer not being able to get to the max accelerations? However even if it was, for the stuff I'm printing it's working great :).
You might be able to increase the diameter of your test piece (beyond the grip point, obviously) as a cylinder, rather than a convex rotation. That would allow for a larger percentage of each print being at the target speed.
To mitigate the acceleration problem, you could redesign your test sample to have a rectangle shaped cross section. If it is rather flat and wide, the printhead will spend way more time at the desired speed and less while accelerating.
I started using my new X1C two weeks ago and I can 100% agree in terms of "speed is not everything". I'm especially struggling with printing TPU (Colorfabb Varioshore) and i already had to de-assembly the hotend and the extruder to remove clogged filament. I didn't had this kind of issues e.g. on my prusa mini. Reducing the max. volumetric speed has solved the problems. I think this long melting zone is not good for TPU.
I found that upping the hotend temps is a better solution than lowering volumetric flow rate. You can still print fast and it doesn't mess up the internal software-hardware tuning that Bambu did. If you are tuning in a new filament, by raising/lowering hotend temp by intervals of 10C then 5C, try to find the sweet spot that surface underextrusions no longer happen. Some underextrusions cannot be solved by flow but only temp, guess it's more about the plastic properties. Another thing is lowering the flow may not actually affect the most print speeds esp. non-infill, if you see similar print times, check the sliced print speed preview and see what speeds are actually being applied on the parts.
I'd like to get a high speed nozzle, but my 3d printer uses m8 threads, which apparently is not standard... I can't even find a source that sells them other than resellers of the original OEM. This was a really great video explaining so many facets related to high speeds. Thank you.
I kind of think you're wrong. I just got my p1p and stuff is coming out stronger than on my ender 3's. I think it's because it uses higher temps which really melds the layers together, and also cools quick with aux fan.
![[LIVE] : ONE ลุมพินี 88 | คู่เอก "ป้อมเพชร vs อัสลามจอน"](http://i.ytimg.com/vi/1ZqTVVbMgbo/mqdefault.jpg)
Have you ever noticed that you printed TOO FAST?
Oh, and check out our CNC Kitchen products at cnckitchen.store/ or at our resellers www.cnckitchen.com/reseller
You don't drive a car at its top speed! 😂
Nice work, just a small thought, if the nozzle need a longer distance for the filament, then maybe the path of the filament in the nozzle could be changed to a circular downward path and end in the tip. A bit like the heat pipe in a water tank.
@Stefan,
You asked for an idea for your test pieces - how about Archimedean chords as 100% infill and 0 perimeter? This should allow you to reach the speed more likely.
General thoughts on flow rate:
When I have many identical parts to print, I often take the effort to set the layers to a different temperature for increasing the flow rate even further. It would be nice if there were a function in the slicer for this - a temperature dependency on the flow rate, or, in reality, a temperature compensation, because our material does not reach our set temperature. Perhaps we have a deviation between the measured value in the hotend and the actual temperature in the material due to the high extrusion rate. If you follow this idea further, it should be possible to establish a relation between the heating element's power and the extruder's feed rate, taking into account the thermal conductivity to consider the resulting delay in temperature correction. Therefore, I do not believe that these settings can be adjusted via firmware, as temperature changes must be made in advance. Unfortunately, I am not a programmer, but I am really interested in whether this could be used to tune the flow rate of ANY 3D printer without making changes to the hardware.
Tüdelü 👋
yes ! very poor layer adhésion issues specially with petg (X1C). days of tunning (slow down cooling and flow rate, increase temp).
Could you do a check on a Voron at high speed? Just to see if this is a generic fast printing issue or unique to the Bambu.
As always, another thorough and professionally produced video. Your content is always top tier.
Much appreciated!
As always, an educated, scientific, unbiased 3d printing informative video.
Glad you enjoyed it!
Thank you for not doing the dumb shit on April fools like releasing a pointless waste of time video. This video is so great and informative!!!
To print stronger functional parts faster I use a tip from one of your earlier videos - print wider lines. For a LOT of my printing these days I'm using 0.8mm lines with a 0.4mm nozzle. With Cura 5, thin walls don't suffer, since the slicer will generate thinner lines when it needs to. You still need a higher flow hot end, of course, since printing wider lines at the same linear speeds means more plastic per unit time. Luckily my Artillery Sidewinder X2 has a Volcano hot end.
I printed my Trident 350 parts with my X1 using ABS+ on generic ABS settings. I didn't notice any part strength concerns during assembly, but now I know to do PM checks more diligently. Thanks for the info Stefan
It will be an upgrade for sure
Second this.
Will definitely be checking also
I came for the Pink Floyd.
Great that you finally make a video on this topic ! Bambulab + CHT Nozzle is amazing, now need a bit more heater and cooling
@@gustavbondeus6767 I know, I tested both a monthe ago, clone and original (two monthes ago), and come to the same conslusions, I even get in touch with bondtech about this and they was pretty interrested by this sucject ;)
Great video! For what it is worth, my old Fortus 250 (industrial printer) runs ABS at 305C and has a meltzone nearly 40mm long. And that machine has extremely tight temperature controls and uses quality thermocouples for temperature monitoring.
Because of that I have become way more comfortable with pushing Temps through the roof while printing.
Amazing video (as usual) Stefan. I'm glad that I can always depend on you doing proper experiments, because SOMEONE in the community needs to do it, and I don't think I have the stamina or work ethic to do this level of experimentation :) Keep up the good work.
When I first got the X1C I banged out a few parts at the default settings, and noted that the print quality was pretty good. Then I cut everything in half (from Bambu's defaults) and discovered this made the print quality *amazing*. As someone who's mostly printing ABS and then sanding and finishing it to look like not-3d-printed, running the X1C at 50% of nominal and 0.12mm layers still gets me parts in half the time the Ender-5 did, and those parts end up needing about a quarter of the sanding and finishing work that the Ender-5's production did.
Hi there what size nozzle are you running on you X1C while printing ABS
@@justinmurray8582 The standard 0.4mm. I'm usually printing at 0.12mm layers and trying to capture fairly small details on many things, so a larger nozzle isn't worth it most of the time. The exceptions don't come up often enough to bother with building up a hotend with a bigger nozzle.
I just got my first printer and was trying to think about what was causing the change in gloss between layers when I stumbled on your video before bed. You saved me at least an hour of parsing through unhelpful reddit threads.
Still a newbie and a former machinist I thank you for insight into issues I've been experiencing with my X1 printer and it's settings. My Ender 3 S1 Pro with Sonic Pad tuned produces faster, good finished and parts of solid structural integrity of a phone holder, clip style. Using the same material, settings (temp/flow) X1 produced gummy weaker parts that failed under the stress of the design. Again being new (December 2022) I wasn't understanding the "why's" for such failures. I will be looking into your suggested mods and settings. Thanks for your insightful efforts.
I’ve had a lot layer adhesion problems with Polymaker ASA with stock Bambulab ASA profiles. Ended up slowing things down and turning down part cooling to get acceptable strength. Definitely going to look into this mod! Thanks Stefan!
I know this is 6 months old, but I have been turning the cooling fan completely off for polymaker ASA and layer adhesion has been great.
Thank you for the scientific and especially concise context. Everyone should draw the necessary conclusions for themselves. Personally, I find 12mm³/s to be completely sufficient. Great video!
US mechanical engineer here, I really appreciate that the strengths were also given in ksi. 👍
I really love the content made by this channel. Steffan (I hope I spelled it right) is not only a great engineer, but also a very good teacher. I learned a lot from this channel.
i swear i could listen to the line 'hi im stefan' on loop for like 10 hours, the way you say it is simply beutiful
AWESOME video!!! Upgraded to 0.6 nozzle and print quality went to the toilet! Thanks for the tip on the extrusion rate, will try it! Thank you! Really appreciate your work! Hope one day someone makes an after market hotend that can use standard V6 nozzles!
You can replicate almost all of the benefits of a 0.6 nozzle by changing extrusion widths to 0.6 while using a 0.4 nozzle.
You lose the ability to do layers that are 0.4mm+ but you gain being able to do 0.1 or less. You also get much better quality.
In almost all cases, the 0.6 nozzle print speed increase is a result of the larger widths, not the higher maximum flow rate. It sounds like you might be pushing that. The cht style nozzle has a significantly higher boost to flow rate than going to a larger nozzle. Same for the volcano style setups.
The trick is to print your outer layers at a smaller width and the inner layers much, much larger. I'm able to get away with half of my prints times are typically from the external perimeter, even with infill values of 50%. You almost can't see the layer lines. Only the seams are readily visible.
Yes! Been waiting for this... I've assumed the speed would compromise strength, so it's great to see it go through your usual battery of tests.
Excellent video Stefan. I really think Bambu has hung their hat a little too heavily on "Speed" and they need to tune in "Quality" presets. My own tuned profiles are significantly slower but more reliably produce better looking & stronger parts (anecdotally). My X-1 gets a lot of use with these profiles and I'm much happier for my needs.
Hi there.
Can you share your updated profile settings for different material
PLA,TPU,PETG ,etc? It would help a lot of us that are getting into 3d printing with this Machine.
Thank you.
Are those profile published anywhere per chance?
If not it would be really appreciated on the discord!
The quality out of the box was better than any of the previous 5 printers I've owned and 3x as fast.
@Broski Snowski I know about Orca,but sometimes getting the profile from somebody who lives and breaths 3d printing is a better base to start from.
Bambu Labs said some months ago that they will be releasing their own *high-flow* nozzles. Hope they offer hardened nozzles for abrasive materials, as most current high flow nozzles aren't hardened.
Oh, I that's cool!
They seem very "agile". Constantly improving and taking in community feedback.
Wondering if they will also release a new nozzle design similar to the A1 series that doesn't require a full hotend. Maybe on the next X1 release? Hopefully also backwards compatible.
Awesome video.
Made pretty much the same experiences with the aftmermarket hotend + CHT.
Currently my go to hotend + nozzle combo for the x1c.
Looking forward to a hardened version.
Finally. A video about something I realized a long time ago from many videos prasing klipper-enabled machines and their ridiculous speeds.
But I still did not hear an explanation of how the increased part cooling fan speeds, that comes with the increase in print speed, affect layer strength...
Awesome video as always. What this actually makes me wonder, is whether the good old 2.8mm filament might be due for a bit of a comeback.
Changing the diameter of the filament affects the feed rate by the inverse square of the proportional change in filament diameter. Meanwhile the distance from the hotend wall to the center of the filament only rises linearly (but so also does the surface area available for heat transfer). So the question basically becomes: Is plastic such a bad conductor of heat that fully melting a 2.8mm filament takes more than 2.56 times as long as 1.75mm filament.
If that is not the case , the achievable flow rate for 2.8mm filament should be higher in a heating zone of the same length.
Thinking about this further this is probably one of the reasons why for bigger extrusion systems the pellets are usually not melted by heating them externally but rather through the friction imparted by the feed screw. In a system like that the flow rate of material and the amount of energy transferred into the material should be largely coupled as long as you adjust the turn rate of the screw to always have the same pressure in the system.
As always, an objective no nonsense fact based scientifically tested approach to solving a real problem. This is my go to channel when I need to research issues related to 3D printing.
Ah I have noticed this as well up to this point I have just been setting my print rate to 50% on the X1 for my "quality" prints but controlling the volumetric flow rate is way smarter. Thanks for the tips and the research!
Thank you for the anti-april-fools warning in the title
Just click-bait 😅
@@CNCKitchen The good kind. April Fool's day gets old quickly.
April fools day is one of the worst.
How do I find out the volume/s?
@you can check the volume in the slicer
When I make parts for press fitting bushings , I use concentric infill to reduce accelerations and keep print speeds faster at 100 percent infill. Should help with your coupons too.
I have bought the hotend and nozzle linked in the description and was able to reliably get to around 38mm^3 before visible changes between layers. Incredible upgrade for my P1S!
I have had a smoothieware delta for the last about 7 years. It has been my workhorse till it was damage in a earthquake. With the stupid amount of time I have put on this printer (wish i had a hour meter but just not possible at this point). It was the first printer I owned with the capabilities of going into the 200mm/s - 250mm/s range. I feel like your findings mirror what I learned over the years with that printer. If I wanted to print fast, part strength often suffered, and even details at times. I would often run it much slower than it could go, even more so when printing smaller plate volumes.
You might also want to consider revisiting annealing the parts that you found previously weak. Not as big of a deal on PLA. But for ABS, and Nylons you can see noticeable differences. Its more noticeable on these materials if you run a high bed temp and chamber temperature. As you can experience annealing while printing, when doing multi hour prints in these materials.
After pushing my hbot to max extrusion rates my parts were coming out all matt and I really liked it but never knew they were weaker.
thanks for the info!
I love your practical scientific 3D printer testing. This is very useful information, and you provide a great education on important subjects most people don't consider, allowing us to understand our 3D printers and the 3D printing process to gain optimal results. Thank you!
Really interesting flow tests! Great to see those aftermarket options performing so well. The standing tests are pretty useless though. Since they’re in batches, the layer time is many times higher and there’s lots of time to cool down. Printing 12 samples 200mm/s is going to end up extruding on plastic pretty much just as cold as 12 samples at 50mm/s
Another nice video that's very informative. I've already ordered the replacement hotend so that I can use my existing V6 nozzles in my X1C if I want to but now I'm ordering a couple of those CHT nozzles made to match the Bambu Lab nozzle length.
Yes, the replacement hotend/heatsink with removable/changeable nozzles is expensive to start with but when you consider it can used for many inexpensive nozzles rather than having to replace the comparatively costly Bambu Lab options every time your nozzle wears out, they end up much less expensive in the long run and offer more variation. That and they make changing nozzle sizes/types much faster, no more having to break down the whole carriage assembly and swap entire hotends just to change nozzle size. Bambu only offers stainless steel and hardened steel while V6 style nozzles come in brass, plated copper (my favorite) and various specialized nozzles like ruby, titanium or obsidian plus many other variations/combinations of tip style/metals. I'll never need most of those since I normally use plated copper or hardened steel when it's needed but I like having the option of using my collection of plated coppe/brass nozzles back, not to mention being able to use the V6 style hardened steel nozzles that I already own rather than having to buy Babmu hotend replacements. I'll keep ordering the Bambu Lab hotends for the hardened steel nozzles when I need them, the AliExpress aren't much cheaper and I trust the quality of the Bambu slightly more. AliExpress can be great but you never know what you might end up with (off center holes, poorly machined surfaces/threads, low quality materials, etc.).
Polymaker high speed pla is a good option to handle the high speed issue. This was an excellent video that showed me where to focus on these issues on my Bambu.
Always happy to learn from your videos Stefan. I hope you and your work-family are doing well with the work life balance.
The standard Tips & Tricks of successful FDM printing really seems to change once you cross that 100mm/sec threshold. Thanks for starting us on the proper education!
Thanks! I have noticed pretty bad layer adhesion and really appreciate the help in understanding the issue
Glad it helped!
About avoiding crazy acceleration rates: all you care about here is relative performance, so you don’t have to be constrained to your standard coupon geometry.
You could make layer-adhesion coupons with much larger diameters in the critical region, which would give proportionately lower accelerations, or conversely higher speeds for the same acceleration.
To keep the forces in the range your universal test machine can handle, you could make the structures hollow. Either start large, using a different holders on the machine, or use the same clamping diameter, but expand larger and solid right above/below the clamps, then go hollow and neck down. (It’d be easier to explain if I could draw it here, but I think you’ll get the idea easily enough. When you start making the hollow part, widen the internal opening slowly (similar shape to the exterior of your current coupons) so it won’t need any internal supports.)
There are obvious limits, so I’m not sure you’ll be able to go to a large enough diameter to make a difference (the total cross-sectional area of the neck in your current coupons doesn’t give you a lot to work with), but it should help some. - And given that your machine can handle tensile testing in the in-layer direction, it might be strong enough to handle a much larger cross-layer cross-section. (Again, for this test you only care about relative numbers, so don’t need to worry about comparability to your standard tests.)
I had the same problems with my X1C.
Thanks for the solution.
You're welcome!
Such an informative video. Thanks! This is a worthy upgrade for a printer that is possibly the most capable in the market at the moment so thank you for showing this.
So now we need super slicer to let us vary extrusion temperature based on volume flow rate with a comp value for heat up time just like with the fan. I’ve wanted this feature to lower the temp during bridges for a while.
Whelp, I was wondering how long it would take, but here we are. That's an instant purchase for me. Thanks for testing Stephan
Awesome video :). always get a bit annoyed when people only talk about how fast they can run their steppers! It's fun to do that but practically matters for me personally
Take into account every possibility for practicality.
Fast and light prints are amazing to test fit something you've designed before moving on to a full strength design
Saustarkes Video! Beantwortet alle Fragen! Vielen Dank dafür!
Stefan, I always enjoy your videos and the dedication you put into making them. Here is an Idea for a video. Testing linear systems, for instance difference in drag between genuine Hiwin rails vs Aliexpress ones, difference between smooth rods and linear rails, difference between carbon fiber rods vs smooth ss rods vs linear rails. Also maybe how big of a difference does lubrication of choice affect smoothness of gliding. and in the end which solution is the best regarding weight to least amount of drag during sliding.
Thanks for the explanation and test.
I confirm using the default bambu ASA setting results in poor layer adhesion.
I was able to get acceptable results by setting the speed to 50% of the preset.
I feel your suggestion of limiting the max flow rate is a simpler and more reliable way to achieve the same.
For the strength test on the fast speeds, you could increase the radius so that it has time to get up to speed. Then you could use your lathe to get the profile correct, while cutting out the acceleration/ deceleration zones
Increase the radius of what? Sorry if a dumb question
For better high flow nozzle tests.
Instead of printing the hour glass shape you could print cylinders of the major diameter, then machine them down to the hour glass shape. Add what's called a live center recess on one end and make them slightly longer so they can be chucked up in a lathe and supported on the opposite end. They do make desktop CNC lathes that can handle this easily. I am a career CNC machinist, if you want any help with programming and setting it up I would be happy to help. But, I have a feeling you could figure it out on your own. good luck and keep up the great content.
Nice video, thank you 😊 why do you want to avoid drastically increasing temperature? I have been doing this for years with fat nozzles without changing to volcano. Works for this printer as well. I go 20-40 °C above max recommendation for nozzle temp. Printing 20-30 mm³/s for PLA, ABS, ASA, PETG, PCTG with hardened 0.4 mm nozzle . Perfect strength, shiny as always 😊
That works when printing at equal speeds for all the print features. If you have to slow down, you will run into issues.
@@alejandroperez5368 what issues? I never had any. You just need to adjust cooling accordingly.
13:17 who else though he would say "with great speeds comes greats responsibilities"?
I've been slowing down print speeds on anything "decorative" to 50mm/s on only the external layers for my bambu x1c. While it definitely slows down prints (but still faster then most printers), it makes parts way nicer and eliminates any of the color changes.
bambu should give the option to bulk purchase parts like costco. the key to printing fast is lower cooling and flow. high flow 0.8mm clones on all three of my machines. love that transition into the sponsor message.
I've thought about posting about this. I'm glad you are sharing. Tuning to a flow rate is what I've always done, since my super racer is faster than most materials can handle. So what I do to utilize the speed is maximize the flow rate and then lower my layer height until I either hit my max speed or smallest layer height.
How fast does that mean you end up printing?
@@travistucker7317 at most, I print quality parts at 400mm/s @60,000mm/ss.
@@ManjaroBlack what layer height does that generally happen at?
I LOVE my P1P, but I have had some problems with strength. Thank you so much for making this video, Stefan!
Same with my X1 Carbon. Certain prints always turn out really weak.. now I know why.
I've noticed this after printing 330 hours with my bambu x1c.
Been printing with PETG a lot and when i switched to PC the performance were worse.
I was trying to anneal the prints, with PETG , to compensate for that with SOME improvement but not enough.
Test done with ERYONE PETG 255°C hot end, 200 mm/s (standard profile) 10% Fan MAX.
So the parts printed with PETG at those settings are now much more brittle than 60 mm/s and tend to break like PLA.
Instead of deforming and yielding like PETG printed on a ender3 at 60 mm/s (for example).
I absolutely want to try the 3rd party hotend with the CHT "like" nozzle.
THANK YOU STEFAN!
I wonder what that stabilization systems are like in the Bambu Labs printers, and other printers.
I think that you could use centrifugal force to change the direction of the movement/inertia in the Printer Head to keep it from shaking.
Kind of like how it's easier to move your hand quickly in circles or ovals without moving your body, than it is to move your hand quickly left to right.
Great video, Stefan! I hope Bambu Labs listens and learns from your experiments and findings.
It seems like Bambu Labs is constantly improving it's printers. Hope they see this and add this to a future revision.
I was thinking the same thing. They should change to a high flow nozzle as the default.
@@Nerlin They would be stupid to not develop and offer drop in replacement highflow units. Most people who have bambulab printers dont want to muck about with aftermarket stuff. Just being able to buy the solution directly would be very advantageous.
I hope they do as default stock.
@@tHaH4x0r I agree. I have 2 X1Cs
Build a VZ-Bot. The Bambu Labs X1C is a toy compared to it
Great video as usual!! My existing printers are dialed in and working well. I decided to sit back and watch (and save up my money) as this next generation of printers like Bambu are perfected. Then I will jump in and enjoy the benefits.
ive been waiting for this video after seeing all the speed benchys, really interestign result and great production as always. might be worth trying out a thin "wall" like sample or a vase mode based one in an attempt to reach the higher speeds
A good way to avoid the acceleration issue in the test coupons would be to use perimeters for the primary fill of the part, as the concentric pathing will avoid the harder acceleration points based on the linear infill, and will be more consistent as it avoids the inevitable changes in extrusion speed required to keep consistent bead width when doing harsh direction changes.
Very useful information. Thanks. I bet the manufacturer is taking notes.
I've seen the banding on PETG much more than PLA. If I go much hotter though stringing is an issue.
Again and again, this is a really interesting subject ! Thank you for all the work
I am super jealous that you can just take a hobby, perform detailed analysis of it, and make a living out of it. I really appreciate your videos and your dedication to transparency and detail. Well done.
CNC Kitchen, can you do a vid on price to performance with filament?
like what is the cheapest filament that still prints decently or what is the best filament for high accuracy. I'm still pretty new to 3D printing and there is a lot of brands out there and I have no idea what I should buy.
I always appreciate your dedication and content, along with your extensive testing. Great video as usual!
For strength tests, I'd be interested to see what results you'd get if you used higher extrusion temperature for your high print speed tests! And then higher enclosed ambient temperature on top of that!
Yeah I am pretty sure you can solve these kinds of issues with a higher wattage heater cartridge & turning the temperature up which decreases the time needed to melt the plastic. A copper heat-block also helps because it both increases your heat throughput and reduces the lag between the heating cartridge and the hot-end. I print at about 23 mm3/sec and don't have any issues with an e3d v6. It's like $20 in upgrades.
It's awesome to see e3d makes bambu high flow nozzles now.
Thx for proving what im saying since the start of "fast 3D print" Speed is Not the goal !
What we realy want is to have as great as possible cubic mms of material ozzed in one second for as litle speed as possible to reduce wear, vibrations ect..
Can you please do an updated "getting started" video for newbies who just want to print things like shop vac adapter parts, or casings for drill bit organizing or the million different clasps and hooks and strange shapes that would be helpful for wood shop organizing?
I've never used CAD software before - I'm looking for something like "buy this printer, with these accessories, install them this way, use this software, buy this filament for X job, this filament for Y job", etc
What about 3d resin printing? How does it stack up against "traditional" 3d printing?
As someone who has been trying to decide between a Bambu X1 and a Prusa MK4 this has been really informative. My main concern is print quality, strength and dimensional accuracy so I am starting to lean towards the Prusa..
we seem to be having the same issue XD
You could always just like, slow down the bambu a little bit. Even with it set to a slower speed it'll run circles around the prusa in terms of print times.
@@kylequinn1963 and quality aswell most likely, plus no printed parts and the X1 has an enclosure, actually enabling you to print ABS and other harder materials out of the box if you needed to
you may think you never need to, and still wonder why you're running a shitty open printer one day, and i Know that for a fact out of first hand experience
Consider whether you ever want to print engineering filaments (ABS/ASA/PC etc.). If you do, go for the X1 as the chamber makes a huge difference to the printability, quality, and strength of such filaments. You can always slow down the defaults too...
The P1P is really the competitor to the MK4.
Bambu X1 is way better machine than prusa, just the build quality is way better. Amd you always can slow it down. Is like having a ferrari, you can drive it at legal speeds and you will have a better car than one guy with a fiat.
wider nozzle sizes allow more material to lay on each other at a heated temperature for a longer time, making it stronger. They make the . This is why even with ultra high flow nozzles, the layers need time to cool down and not too fast, otherwise the part will collapse on itself. Have a look at the approach of filament innovations in the US. The make 2.5mm nozzles with very long heat blocks on they typhoon extruder and still produce very strong parts.
Mega aufwändig, mega schlüssig, mega aufschlussreich! Vielen Dank für dieses Video!
*The rule of thumb is very simple*
1) When slicing, check if the flow is less than your hotend is capable. For example: V6 is capable of 11 cm3/s, Volcano - 18... Your flow can be calculated as simple as this: feed speed (mm/s) * layer hight (mm) * extrusion width (mm). For example if you have 0.2mm layer hight and 0.6mm line width, you should not exceed 18 / 0.2 / 0.6 = 150 mm/s (for Volcano)
2) Never bump up the temperature to increase speed.
If you stick with these rules your plastic at the nozzle tip will have the same temperature all the time at any layer.
Why #2..?
@@coltenmeredith8899 If you do this to increase the flow capabilities, your printing will behave differently in different parts of the print: if the printing is slower (or line is narrower) your material will be hotter than desired and you will have glossy surface and worse overhangs, maybe even material degradation.
If your flow is not exceeding some certain value, your material will have exactly the same temperature out of the nozzle. Which will result in reliable printing.
Yes indeed: printing speed does have an impact on part strength and there are two reasons for this.
Lower speeds means there is more time to heat up the material so it can reach a more uniform temperature from the outer surface all the way to the core of the material but it also allows the nozzle to act more as an iron when the flat ring of the nozzle tip irons the extruded material onto the previous layer.
I still find it strange that one nozzle geometry is used for all those different materials. The area of the nozzle tip and the length of the orifice play a major role.
Such a great video! And thanks for linking to our blog post. Pretty cool to see that your empirical tests line up with what my thoughts were after reading those papers 😅
My pleasure!
Excellent comparitive analysis, thanks to the hardwork on the same❤👍. What I feel might be the issue, is the max power capacity of the heater, and also the quickness of the PID system controlling it. That being the reason for better performance at elevated temps. So, the heating element itself should be beefed up, along with a more responsive PID, coupled with a higher output power supply for the heater element. Essentially, it boils down to the max possible volumetric melt mass within a given time duration. Just some 💭
Thank you Stefan for another super informative video. The work you do is so valuable. I just bought the new Ankermake M5 and have experienced the same issues as addressed here. The Bambu may be my next and having this info makes the decision much easier.
This explains so much. I was wondering why some of my parts were coming out with inconsistent surface quality. I have been tweaking the max volumetric rate, with better results. But I still didn't understand why it was happening. I love any excuse to mod my printers, so I guess it's time to make some aliexpress purchases!
Danke Stefan für die Bestätigung. Ich hatte gerade genau die gleichen Problem, vor allem mit ASA. Die prints sahen gut aus, waren aber einfach zu schwach und ich war sehr enttäuscht von ASA. Schlussendlich war meine Lösung genau die: langsamer drucken und weniger Kühlung. Die Standart Profile haben einfach zu viel Kühlung nur damit schnell gedruckt werden kann. Gerade wenn man mehrere Teile auf einmal druckt, wird der X1C sehr schnell, aber die Teile sind einfach nicht stabil gewesen. Also Kühlung runter. Ich habe es aber vor allem über die layer time geregelt. Also längere layer time und damit den Print künstlich verlangsamt.
Genau, so hab ichs schlussendlich auch gemacht.
TBH, I'm totally fine with the stock nozzles from 0.2mm to 0.8mm. I was considering aliexpress nozzle, but I got used to swapping nozzles enough to not care about dirt and hotswapping with risks of burn. I rarely move my x1c to sport more, and only twice tried ludicrous for lols.
Came across this whilentrying to piece together why the bambu labs printers are so fast. At its core 3d printers melt plastic and deposit it in layers. My hesitation in buying one has been in figuring out if they just print such that the parts look good or if they are able to achieve excellent layer adhesion as well. Thats something that i havent seen ANY other channel address at this point (may 2023). Thanks!
This speaks to why I haven't upgraded to a faster machine yet. I mostly print in ASA for the heat and sunlight resistance. I more often slow down a print for better layer adhesion, particularly if it's small, rather than speed up. It seems like for functional parts, particularly printed in materials other than PLA, high speed printing isn't very useful.
I would like to be wrong about that, though. It would be nice to print faster.
@Stephan- have you ever taken a look at temperature accuracy on your X1C? I have spent well over a month going back and forth with Bambu on this, after finding that the hotend thermistor/electronics report back a temperature about ~20C higher than reality, resulting it a "real" hotend temp of only ~250C when you ask for 270C, etc etc. This means the default 220C profiles for PLA are trying to push it out at only about 200C.
It's not just a hunch- I've tested with multiple replacement thermistors, different (official) hotends, settings, countless test prints and meandering flow loops and more, and using both a thermal camera (admittedly not ideal for shiny objects, but calibrated using a Mosquito magnum :) and ALSO with K-type thermocouples on different parts of the hotend body (even in the same hole as the bambu thermistor and with either their thermal grease or even with boron nitride paste.)
Heck, they even sent me all new electronics for the TC board (the interface board on the toolhead that everything plugs into) and I still see the same thing. I don't know if it's just me seeing this, but I hadn't really seen much discussion in this specific area, and had been meaning to reach out to you for a while now.
So, if you set your printer to say, 230C, and sticking a cheap K-Type thermocouple coated with some thermal grease onto the hotend gives you a 230 +-1C reading, then you only have to worry about the "well known" hotend performance limitations that you described very well in this video. But if not, then you may find yourself having to do the constant mental math I've been having to do for the past month or so...
By the way, Bambu has only kind-of-sort-of acknowledged my findings (after sending them multiple videos, test reports, troubleshooting sessions, etc) but I'm still missing the big "ahha" from them on this. They still seem to give off vibes that they think I'm just a crazy guy nitpicking things :)
I've felt, with no scientific testing that I was missing something about the temperatures on these machines. If you are right it would make sense to me based on my experiences. I never see stringing and it's like I'm printing at lower temps than I should be. I'm gonna crank the temps up and see what happens.
For the adhesion test i you could do single or multi-walled cylinders and print in vase mode. Since you have a cylinder layer adhesion might still be the weakest link. might be worth a try. The bigger the circle the less the acceleration limit.
Wonderfully detailed as usual! Great work Stephan.
Most of my profiles from Bambu for ABS/ASA already have max flow much lower so avoid this issue... clearly bambu keeps updating based upon feedback, which is great.
The video considered the extrusion rate in mm^3/s, length of the melt zone, nozzle design and hot end temperature as variables impacting the temperature of the material exiting the nozzle but that's only half of the issue. The temperature of the previous layer is also critically important to interlayer bond strength.
Instead of printing 14 pull test samples at once, allowing the previous layer to cool, try printing the test samples one at a time so each layer is extruded onto a layer that hasn't cooled as much. I print small structural parts in TPU. I have the slicer print them sequentially and the parts are ridiculously strong because at 50 mm/s print speed and minimal cooling, hot TPU is being extruded onto very warm TPU, rather than printing hot TPU onto cold TPU.
It would be interesting to have a slicer setting that adjusted fan speed for uniform part cooling depending on print speed and part geometry for more uniform results, hopefully avoiding matte finishes in some areas. I believe so much effort has been devoted to adequate part cooling on the new generation of fast coreXY printers that little to no concern has been expressed for excess cooling in some instances. Maybe the problem isn't printing too fast, but rather cooling too much.
Have you ever printed your strenght coupons as all walls or with concentric infill? Could be interesting.
I don't want to sacrifice quality for speed in most cases. I like 1) reliability 2) repairability and low maintenance 3) and good print quality. Speed is not my first desire in a 3d printer although it might be nice to have one where print quality isn't important.
Printing fidget toys, I haven't noticed the matte vs glossy too much. Possibly due to the printer not being able to get to the max accelerations? However even if it was, for the stuff I'm printing it's working great :).
You might be able to increase the diameter of your test piece (beyond the grip point, obviously) as a cylinder, rather than a convex rotation. That would allow for a larger percentage of each print being at the target speed.
To mitigate the acceleration problem, you could redesign your test sample to have a rectangle shaped cross section. If it is rather flat and wide, the printhead will spend way more time at the desired speed and less while accelerating.
I started using my new X1C two weeks ago and I can 100% agree in terms of "speed is not everything". I'm especially struggling with printing TPU (Colorfabb Varioshore) and i already had to de-assembly the hotend and the extruder to remove clogged filament. I didn't had this kind of issues e.g. on my prusa mini. Reducing the max. volumetric speed has solved the problems. I think this long melting zone is not good for TPU.
I found that upping the hotend temps is a better solution than lowering volumetric flow rate. You can still print fast and it doesn't mess up the internal software-hardware tuning that Bambu did. If you are tuning in a new filament, by raising/lowering hotend temp by intervals of 10C then 5C, try to find the sweet spot that surface underextrusions no longer happen. Some underextrusions cannot be solved by flow but only temp, guess it's more about the plastic properties. Another thing is lowering the flow may not actually affect the most print speeds esp. non-infill, if you see similar print times, check the sliced print speed preview and see what speeds are actually being applied on the parts.
Yes. This kind of surface quality problem always means you're printing too cold. Same with part strength/layer adhesion.
Hey, thanks for your amazing scientific, unbiased videos, I love your work.
I'd like to get a high speed nozzle, but my 3d printer uses m8 threads, which apparently is not standard... I can't even find a source that sells them other than resellers of the original OEM.
This was a really great video explaining so many facets related to high speeds. Thank you.
I kind of think you're wrong. I just got my p1p and stuff is coming out stronger than on my ender 3's. I think it's because it uses higher temps which really melds the layers together, and also cools quick with aux fan.
I just increase the temps 10c to keep the volume high, 20+. 230 pla, 270 abs. Using Bambu 0.6 nozzle.