Thank you. I do that because I love testing and want to know myself. It is always nice to find the true performance of the products based on proper testing data. :)
There is also a REVO version but it's super expensive and seems geared more to quick nozzle changes. I own the OnXidian and outside the higher flow rate, prints are slightly cleaner and the non stick diamond coating actually works on PETG so the price really is justified. The best thing is all you have to do is change the volumetric flow rate for faster prints. Obviously there are other settings to increase speed but that one setting alone results in faster prints.
@@MrBiggs-wu2kf It's made a huge difference, I have noticed. Especially when printing with 0.12 layer height. The prints are strong as hell (when using PLA+)
@@MrBiggs-wu2kf I haven’t really tested the strength claims, but I have been able to print PETG at 30 cubic mm per second in tests very successfully, and I’ve gone up to around 24mm3/s in actual prints and they seem strong. That’s with cheap Overture PETG. For PLA I’m usually not pushing speed limits because I’m printing production parts that need to be reasonably strong with good cosmetics without post processing. I was happy enough with the first that I got a second for my other X1C.
Hey! Don't forget to anneal your PET-CF filament. As a semi-crystalline material, as printed it will be quite amorphous. If you want the insane stiffness noted in the TDS sheet, annealing is a REQUIREMENT. It's also fairly easy to do
I never looked into the annealing process, but thanks for the info I read up more about it. I looked up what Ultimaker writes about it and it says it gives "a stiffness increase of 10%". I checked the Bambu PET-CF TDS and yes their samples were annealed. I love comments like yours as they help me learn very quickly about important things in my busy schedule, thank you so much. :)
@@PrintingPerspective of course! For bambu its actually fairly simple. Basically, start at 80C hold for a bit depending on part size, and slowly increase in 10C increments every half hour till you get up to 120C. Hold for a bit. Slowly cool down. For everything else pet-cf from other brands. Chuck it in the over at 140C and send it
@@matthiwi6901 which brand Bambu: ramp up to 120c slowly. Everyone else: chuck it in at 140c and hold for a few hours. If you have thin walls, use sand or salt as support
i think at the end of the day layer adhesion comes down to the a) true temperature of the filament when exiting the nozzle and also the b) temperature of the existing layer below. we cannot really measure a), but only the temperature of the block and all these tests we do assume it is equal to thermistor temperature even though we know it is not. that means we have hidden variables like 1)block material, 2)thermistor distance through the block to the heater element, 3)distance to the nozzle, 4) nozzle properties, 5)PID reaction time and maybe other things i can't think of at the moment that account for filament being printed at a lower temperature than requested. i strongly agree with you that better nozzle properties and longer melt zone length is the best move since it appears to be able to wash out all the other variables preventing any of them from becoming bottlenecks. when comparing hotends we cannot really change 1,2,&3 and can treat it as one black box variable and call it the performance of the hotend but perhaps we can stick to the same type of nozzle at least when doing testing so we can eliminate 4) when comparing hotends. i'm looking at 6:57 and wondering why Obxidian adhesion loss saturates at -30% while Bambu is actually accelerating to -70%. i am not satisfied that we cannot separate the nozzle variable from the hotend variables in order to explain this phenomenon. i suspect you feel the same way... i have a thought. i know from your past video(s) that very poor performance can be recovered by increasing the hotend temp. i think if you can measure 4'C difference inside the block itself (6:16) at high flow then the nozzle temp delta will be even higher and (a) will be yet even lower, maybe 10-20C lower with hardened steel nozzle at high flow. to me that is a clear indication that that Bambu's poor performance at 6:57 is suggesting that (a) is too low and potentially near the limit for this particular filament. i think that filaments have this hidden property that needs more investigation: what is the actual temperature required out of the nozzle for a particular filament to suddenly lose adhesion strength, call it "x"? we know from CNC kitchen in a 3yo video that his PLA achieved max layer adhesion at around 230C requested temperature with also pretty ok strength from as low as 200C, but huge loss at 190C, indicating that x exists. it is my guess that your purple PLA has an "x" quite close to the requested 220'C, maybe 200C. so if a hotend-nozzle combo that loses 20C out the nozzle produce strange graphs like accelerating loss of layer strength( 6:57) while a better combination will see layer strength loss saturate to a fixed loss like -30%. the numbers are just a guess of course, my point is your purple PLA maybe needs a slightly higher temperature to behave optimally, combined with the fact that bambu HS is not so good with the requested vs actual temp at high flow. as for b), the temperature of the existing layer below it can be controlled by ensuring the part fan is turned off and the model should ideally have smoothly increasing layer lengths in proportion to the increasing flow rate. this is to ensure consistent layer times at all levels. but i think this variable is not large in comparison to what happens to a) during high flow.
bond strength depends on the new and previous layer temp, but also time. As in, how long the junction itself spends at above glass transition. The relationship is non-linear and also depends on things like strain in the plastic, of which there is plenty in FDM. Its a deep af science this polymer stuff. Half educated guesses, half statistics, half geniuses with their insight coming up with models to predict things.
would love to see follow up video with two additional things; 1) add increasing layer pauses to test your theory of the previous layer being hotter because of faster flow / printing contributing to better adhesion 2) test 0.6 nozzle, and another filament such as PETG, PETG-CF and ASA :)
I run the E3D ObXidian on my Bambu but I run the TZ 3.0 hotend on my TwoTrees SK1 and it performs really well.
4 месяца назад
Wow. Thank you for this. This was actually shocking for me and very insightful! I was shocked by how much layer adhesion drops in the stock hotend with flow rates above 10mm3/s... Given that I print, mostly, mechanical parts, I'll definitely make some changes after this video. Thank you for this
4 месяца назад
I'd love to see more of these filament/settings/hardware comparison videos, since you seem to have quite a nice setup to measure these kind of forces!
It also depend a lot on the filament and even a color of it and the print itself. If you printing something small with a lot of details you won't have any problems but when you start printing bigger practical prints where you spend extruding a lot and not slowing down or traveling much, then the adhesion suffers.
To me it emphasized how important it is to actually pay attention to temperature. If you need a big strong part, go ahead and run PLA at 240. It won't be the end of the world.
I just picked up one of the 0.4 E3D hotends for the X1C and it is wild how much faster you can push the prints. On most of the filaments, it moved it from 13 to 25 mm³/s per second. Honestly, I'm sure I could tune it to be even faster, but this is such a reduction in time that it doesn't even bug me right now. I did notice that the flow is so good that it ends up with extra stringers when it goes to purge filament. so you have to keep closer eye on it so it doesn't clog. I'm sure this could be mitigated with more retraction, but I just haven't tuned it like I said I just got it. I also have used the AliExpress CHT aftermarket nozzles and I wasn't impressed. I ended up going back to the stock bambu lab ones. might actually be a case where the more expensive 3ED was actually worth it.
Thanks for sharing. If you care a lot about layer adhesion I think it is definitely worth it. :) I printed some Voron parts and my god how strong and shiny they were.
Thanks for this! I had the same results, althouh I did not have the numbers to show exactly what happens, this is my experience: - cht clones with hardened steel nozzle printed parts felt significantly weaker than stock hotend, so I stopped using it - cht clones with brass nozzle printed parts seemed fine, a bit better than the stock bambu hotend, I still use it for my Prusa mini - e3d's hotend printer parts seemd significantly stronger, I love it. I should note that I did not increase printing speeds on X1C, I use the standard Bambu profiles speed.
I tried the 3.0 and wanted it to work. Tons of tuning and I still couldn’t get consistent results. 2.0 worked much better for me. Both setups had a CHT nozzle.
I really feel like this channel needs to be bigger, the amount of misinformation about printing benchmarks this guy disprooves is amazing, and as always with the numbers to back it up. Great video keep up the good work :D
Thank you for the testing, I was wondering about the performance of the tz 3.0 but didn't find much, thank you. I will propably stick to the dragon clones.
as he mentioned tz3.0 with nickel coated nozzle is way stronger because of the better thermal conductivity. if you put a brass nozzle i think it will also be better. dont forget hardened steel isnt that good against copper and brass ;)
@@romainfichet5352 Yeah, It does improve quite a bit with the copper plated ones. It's still not as high as the e3d. Would be interesting to see how much these findings translate to other hotend designs (rapido, revo...)
As you had mentioned briefly, I think most of this comes from the nozzle material, the TZ and Bambu use the hardened steel vs coated e3d, I think the Bambu and TZ would perform similarly with a brass nozzle with cht, as you found the TZ with copper plated CHT performed significantly better vs the standard hardened steel, seems to all come down to thermal conductivity of the nozzle material itself. I believe they use standard v6 threads so maybe popping a v6 cht clone would be a quick and dirty test, would need different fan ducts though.
I agree, it’s been known for a while that hardened steel just isn’t that good with its poor thermal conductivity. It surprised me back then to learn that Bambu labs came with them by default.
I picked up the ObXidian nozzle for my P1S and began testing last night.. I normally print Bambu PLA basic at 18 mm3/s. With the ObXidian nozzle the max flow rate test hit 40 mm3/s before showing *any* signs of issues! I ran the test at 210c, 220c, and 230c nozzle temps, 210c looked the best out of the 3 to me. How hot is TOO hot for this combo though? Is 240c a reasonable temperature? I normally print at 220 and have not gone over 230. Your chart at 7:24 shows 230c to have the highest tensile strength.
Nice Video, i use the E3D Hotend for 6 Months or so. The first thing i did, was a flow test, with the old and the new Hotend. Becaus i´m looking for stability, i always break pices out from the flow test to find out how slow i have to go until the pices dont break a long the layer lines. Stock was around 150mm/s and the E3D was just bending at 500mm/s. Thank you that i can finaly see this in numbers.
Okay, yeah, that drastic improvement in layer adhesion really does make the price tag of the E3D hotend easier to swallow. I’m not super concerned with printing fast, but stronger prints? That absolutely gets my interest.
Did you try the bambu clone with copper nozzles? Theoretically, the hardened nozzle does not transmit as much heat as the copper one. Maybe the problem of layer adhesion is there in the clone.
I was thinking back to this video... QIDI has copper plated with HS tip only. If you recover two clone CHT inserts and you enlarge the hole of the genuine QIDI to slide them in, what would be the result? 1) good outer material conductivity 2) good inner insert conductivity 3) HS tip which is the part under abrasive stress, so no big reduction in useful life.
I was about to buy a 0.6mm nozzle from bambu in order to speed up big prints, however harsh reality kicked in as even the 0.4mm nozzle, at 0.28mm layer heights for draft prints, already maxes out my flow rate. So 0.6mm wouldn't save that much time if any (some time I guess could be expected by less movements) - checked this video out, and now the 0.6mm e3d is on it's way.
*I print hotter if I need strength and print at recommended temperature when it needs to look good. I do every thing with a .6mm nozzle these days. Those .4mm nozzles are for beginners!*
Curious, if the flow is increased, and the temp as well, print at a speed where the cooling is at 100%, but only if the model is close to fail at that speed with that cooling. So, in other words, print at say 400mms. set the layer time and cooling parameters such that the model is on the cusp of fail from low cooling, then test the adhesion. Would there be a difference?
Can you add some super high flow hotends to the test? Like a super volcano, Goliath, mosquito etc. those hotends might give very good layer adhesion at 50% of their maximum flow rate.
Ha, so my laziness of not installing my after market cht clone months after receiving it was right on. Yeh laziness FTW! Thank you for this amazingly detailed test! Keep it up!
I’d love to see these type of nozzles available for the A1/A1 mini as it’s easier to change the hot end on these machines. One might not be able to go quite as fast but it’d be great for layer adhesion.
I have the TZ "1.0" and TZ 3.0 both with hardened steel CHT nozzles. I've hoped to get better flow with v3 because of the higher wattage heater. But it didn't make a difference. I still had Nozzle Temperature malfunctions because the Heater couldn't keep the Temperature. I've also used the slice Boronitride paste, but it didn't change a thing. My only problem with the e3d obxidian is that you cannot change Nozzles. I've clogged already two CHT knock offs. If the Obxidian gets clogged, then I've thrown basically €100 down the drain. BTW, thank you for the informative test
Thanks. Well, buying 3 hotends that didn't make improvement to the stock is also basically down the drain. I am in the same boat as I have bought 1x tz 3, 2x tz 2. I would rather have one that actually good. E3D say you can do cold pull and geometry should allow you to do that to unclog it. On clones it is impossible to do because of the design. It would be interesting to know what is E3D policy about clogs that are unfixable.
I just installed a TZ 3.0 and I was thoroughly disappointed. Which led me here…. I wish I saw this first! Thanks for the info. The higher temp can cause worse inter layer adhesion is an interesting fact that I hadn’t even considered. Very well put together. It looks like I’ll be swapping out the TZ 3.0 for the E3D lol Well, that’s a Like & Sub from me, that’s for sure!
Very interesting video. I'm still fairly new to the world of 3D printing so videos like these showing what happens at different speeds and temps is helpful when I start getting into fine tuning the print settings. I came across that Slice Engineering is coming out with a Mako hotend for Bambu Lab X and P series printers. Would be interesting to see what the results of using this hotend would be compared to the others in the video.
The hardened genuine CHT also is using hardened steel insert in a copper shell. I fear it wouldn't score much better than the cheap CHT. I guess I'll have to actually do measurements once I get it. This makes me also think that it may be better to get a volcano CHT and cut 2 mm off it rather than using a QIDI adapter and a V6 genuine CHT. Or simply I'll print abrasives at 10-12 mm³/s
@@PrintingPerspective yea but like you showed, thermal conductivity of metals is at least 20x the one of plastic so it's very likely that the swap of material for inner/outer matters little. For example the obxidian is made of brass, not copper,and beats every other one. I guess the e3d nozzle gets the boost from the shape and amount of surface, not much because of the material. As IMHO direct consequence, swapping copper and HS won't change performance too much between genuine and knock off chat. Tough choice, I'll try to get measurements on the stock before picking a replacement.
I always can ask, it would be another step in figuring out more about layer adhesion. But usually unless companies reach out to smaller creators themselves it is impossible even get any answer lol.
@@PrintingPerspective for completeness, I think the obxidian should not compete with the hardened steel nozzles: I found on the website that it's just a coated nozzle. They say it's harder and more resistant to wear, but also that fiber materials will wear it out. So not really the same field as hardened steel which will last a long time, except maybe with glass fibers.
Nice video, what would be revolutionary is someone working to increase layer adhesion, funny that with all the improvements we all live with a part that is incredably weak in one direction.
It would be interesting for once to fix the hardware and to change the material, since all channels always test PLA. Roughly speaking in % how should I scale for ABS, PETG and maybe another couple of materials? ABS is known to flow better, will it be impacted by speed less?
Testing ABS and PETG is so hard because you can't dump a lot of cooling as that affect layer adhesion. And without cooling test prints don't have time to cool down. Not to mention that for ABS you have to constantly monitor chamber temperature and different brand PETG filaments can heavily impact results. If a hotend can better heat PLA it will also heat better any other material.
I just caught up by watching a bunch of your recent videos, thank you so much for the testing methodology you designed iteratively! Measuring the layer adhesion is the real hotend evaluation approach 🤩
It also show that if you are printing an engineering part and you aren’t sure or can’t measure your layer adhesion then print slower! Failed or useless prints waste much more time than a slower print ever did.
Can someone please explain to me what "Print Speed" setting is being modified? I see so many different speeds in the slicer i.e: Infill, Outer Wall, Inner Wall etc.. How does one speed get extrapolated to these/which one do I change?
I would be interested to know how well extrusion force correlates to layer adhesion. If it does corollate well, it could make tuning settings a lot easier without having to buy or build a strength tester and test with every filament and hotend combo.
Too bad you didn't test with the 2.0 hotend. I feel your conclusion is based on the 3.0 results who seems to be hated by everyone who bought it. So to say that Ali knockoffs are inferior is somewhat misleading. And the E3d hotend costs a whopping 95 € in Europe. Maybe a future update with the 2.0? I do very much appreciate this video with your general findings about layer adhesion. Cheers.
I think you are missing the point of a hardened nozzle. The fact that the nozzle with the copper insert was both (slightly) better performing for layer adhesion AND it can print abrasive filaments is kind of a big deal. Are you going to run Glow in the Dark filament or Carbon Fiber through your expensive E3D hotend? I think you will be quite unhappy with print quality after a couple of hours.
Just slam best parts together. On my K1 max I use Microswiss hotend, stock heat break and Triangle Lab hardened steel CHT nozzle. Almost same flow as Bondtech CHT but I don't have to swap for abrasive materials. Not leak proof anymore but thats where my Wera torque screwdriver comes in, never had leaks after I started tigtehning with 2.5nm. But with the added speed, you need linear rails. Dual rods are binding on fast speeds, k1 or x1c, doesn't matter, you can't align it properly.
E3D really should include layer adhesion as a benefit for their HF hotends. Having purchased two, I wasn't really impressed with the speed gains for the cost. The Bambu printers are already very fast. Knowing that layer adhesion is better maintained at higher speeds, however, does make them more worthwhile to me.
I havnt tried the E3D nozzle but i have got the v1, v2 and v3 of the bambu compatible hotend and for me the v3 definitely can run the best out of those 3 and v2 and v3 can beat the stock in flow too with v3 doing the best. I don't have a method of testing strength like you other than just feeling with my hands but when I print PETG the bambu like to print at 12mm^3 but it will print around 15mm^3, the v2 slightly higher but the v3 will print perfectly fine to 23mm^3 and then runs into heating errors. I know most don't use PETG for speed or flow tests but its all I have tried thus far and so far the v3 with the cht clone seems to work good and the prints don't look any different and feel just as strong. Probably not what your looking for as I would need to test with PLA or ABS etc. but just thought I would say. I do like your tests and would consider getting the E3D but I'm not keen on its price.
e3d seems cool, but MAN that is pricy and it still has the insane downside of not having an interchangeable nozzle size. even for me, buying a v3 after my v2 broke (heatpipe insert stuck) and a new set of nozzles (2x .2 to .8), buying ANOTHER v2 with nickel inserts would still be less than one e3d with still just a single nozzle size.
@@mrrooter601 Couldnt agree more and I see the full kit is now available from E3D now too at a higher cost of course. I just cant justify its cost. My V3 is doing me just fine for now thank God.
@TechieNI PETG inherently has great layer adhesion. Maybe that is because you can't feel the difference in strength? I don't know. I agree that E3D costs a lot, but considering its size the performance is just unreal. It comes down to if you can justify the price in your use case. Increasing the temperature on the stock Bambu hotend gives great results too.
I am printing at around 10mm/s sometimes lower on my pretty stock Ender 3. I chanced to linear rails and to Klipper also I have upgraded the cooling. The heater isn't capable to heat more.
Would this also apply to the Panda Revo hotend with ObXidian nozzle? I'm thinking about switching to the Revo system because of the easier nozzle swaps and the better flow capabilities of the ObXidian nozzles. I just wonder if your results would translate directly from the E3D ObXidian hotends to Panda Revo with ObXidian nozzles. Great video and solid research. This is what we're looking for! Your efforts are highly appreciated!
7:10 Just wondering - how do you remove the filament after printing through a hotend like that? Won't the bifurcated molten filament solidify into 2 strings at an angle that would block the extraction since they cannot both fit in at the same time?
You sir are a true scientist, allowing the results to speak for themselves. However, I left the video confused about what I should buy for the best results. Can you just tell us what machine, what nozzle, and what flow rates get the best results? The video leaves it all up to me, but then again, I am not an expert, so might not make the best choices. I am considering buying a Bambu printer.
Verry nice work! good and simple explanation and perfect that you can visualize the test results. It would be very interesting to see a comparison with the new Panda Revo hotend for P1. It is made in collaboration between E3D and BIQU. the build up of the nozzle seems identical as E3D you just tested. So is there a difference? I love the detail and effort you have put in to thies tests and the detailed graphic so we really can see the difference between the different hotends. A lot of people only test the flow rate but not the adhesion wich I find much more interesting. Good work!
Hey I'd like to suggest an idea. How about every month/quarter, you'd post a video summarizing your findings, some sort of updated "state of the art" and/or have them organized in playlists. That's something I miss with youtube, content keeps adding up and it's difficult to summarize/to know what is still a current finding.
I think what is happening the melt-zone is long enough to melt the plastic but due to the lower thermal conductivity of the steel the temperature of said plastic is not as high thus, the lower strength adhesion. I think thermal conductivity plays just a big role not just temperature when it comes to layer adhesion.
should have bolted your extrusion 90 degrees to that. you doubled your stiffness but if you had taken advantage of moment of inertia you could have 9x your stiffness bh^3/3 so if
I really want to say thank you for doing all the work to do with such extensive testing.I've used the v2.0 and V 3.0 and so far. I like the 3.0, because I've been able to use it to print PETG. At 22 mm per second. Honestly. Caught me off. Guard to strength is that much weaker. I never really noticed it Producing nervously weaker parts. Though now my wallet dislikes you.Because I honestly do you see myself ordering a .6mm .e3d hotend. I do you have a question because I noticed using the .4mm CHT insert hardened steel nozzle would require a great deal more purging. And sometimes even have a small blob from a previous color 5 layers into a new print. Would you recommend e3D nozzle for any multicolored printing?Or is the purge necessary too annoying ? Lastly do you think the e3d is any more susceptible to clogging when printing with carbon fiber filament? Thanks a lot for any info you can give. i'm very interested in getting that upgrade just a little nervous about spending that much but I do like the idea if it doesn't have huge major downsides.
I don't do multi color printing so it is hard to say. But yes you will have to purge more which will provide even more waste. Maybe just increase printing temperature slightly for multicolor. E3D hotend mostly shine for practical prints where layer adhesion is very important. About clogging, I never got a clog using 0.4 nozzle with bambu carbon fiber filaments where they don't say you can't use it. I wish E3D made 0.5mm hotends as 0.6mm one reduce print quality too much.
Put the extrusions in a T orientation for even more stiffness really like the way to test stuff, could you do add a test of the panda revo hotend. i know its pricy :/
Can you maybe share a link to the thermal grease(?) which you used for the thermistors? Is it sticky and keeps the thermistors in place (like a glue?) - because I always have issues with the thermistors popping out of the tiny holes again :(
@@PrintingPerspective Nice to know someone enjoys their kp3s my kp3s pro is so annoying the z linear rail block that connections to the x axis is loose and gets displaced easily, no solution yet. Got an anet et4 for cheap will try to mod that into a good printer though linear rails are too expensive.
Maybe im wrong but layer adhesion is not dependant to the temp only? The nozzle should affect that... so its a matter of setting the right temperature in the TZ so it has a better adhesion... speed and quality is another thing, but is plastic bonding together why the nozzle should affect on that?
bambu P1SC When using a 0.2mm print head, AMS’s material replacement time is very long! Please provide a video tutorial to change the change filament g-code code. Speed up the time to change materials! I observed the material change process: it required 4 material flushes to complete the replacement, but in fact only two material flushes were needed to complete the material replacement. TKS!
