E3D Revo(v7) vs HF Dragon Heatup Comparison
ฝัง
- เผยแพร่เมื่อ 28 ก.ย. 2021
- Comparing my Revo Beta unit vs a Phaetus Dragon HF hotend
40w PTC heater on the Revo vs a 50w heater in the dragon
Both are PID tuned in the same(ish) toolhead
Please note, the timeline scale on both charts is different
Recorded on livestream 28th Sept 2021
If you like what you see and want to help support the channel
/ nero3dp
www.buymeacoffee.com/Nero3dp
s.click.aliexpress.com/e/_AlSJoV -aliexpress affiliat link
Looking for a great first 3d printer? Grab an Ender 3 V2 from the affiliate link below, get a great deal on a machine AND help the channel out in the process.
www.banggood.com/custlink/vmK...
Need filament or a printer? Purchase anything from MatterHackers with the link below and a portion of the sale goes to supporting the channel
www.matterhackers.com?aff=7543
I NOW HAVE A TWITTER, follow me @ / 3dpnero - วิทยาศาสตร์และเทคโนโลยี
Not sure if that is really apples to apples..... heating up I get .... but what filament can each push thru ? How can we compare it with filament been pushed thru ? And it's a lot more important that the TEMPS stays up and not drop off at all so a bigger tight block will always with.
Interesting new concept
Thanks for sharing your test of the hotend :-)
You're right about one thing for sure. This IS how you sell things nowadays: flaunt one thing that makes it look vastly superior while not comparing anything else. This demonstration is very reminiscent of Elon Musk unveiling of the Tesla semi truck: "look how fast it gets up to speed! please don't ask how long it can maintain it!"
I need to pick a new hotend if i want to use Stealthburner in my Trident. I came here hoping to get a fair comparison of my two frontrunners.
I really like your videos, but this one seems a liiiitttle biased. I still love you, tho. 😁
Looks like a no brainer upgrade for hemera where it just directly screws in. Cool.
putting the thermistor closer to the heating element gives you a more responsive graph, with more overshoot and undershoot and lag at the actual nozzle temperature. Just looking at the graphs can be misleading. If you try pushing filament by hand while they're heating, you'll probably notice that lag between when the revo says it's at the melting temperature, and when the plastic actually starts flowing.
The revo likely has a prettier graph, but _worse_ temperature regulation if you were to actually measure the temperature at the nozzle. This will mean It'll need a slightly higher temperature setting when at flow rates near its upper limit. (And slower flow on these fast prints will be more overheated)
WRT thermal mass: It mostly doesn't matter either way, assuming PIDs are tuned, and you have an accurate measurement of the nozzle temp. A more massive block will take longer to heat up, and cool down, but since we're generally printing at a constant temperature, it's actually easier to keep its temperature smooth, since the load changes will not affect it as quickly.
0.o? Never have I been even slightly concerned about how fast the hotend gets hot, it's universally the heated bed that takes a long time.
I'm hoping that with the Prusa XL coming late 2022, it has a modular bed of 16 sub-assembly beds that heat independently, that they will heat not only more uniformly but because they are each fairly small they will heat up really quickly.
Not weird, considering the revo has a ceramic heater core heating all around the revo unit, compared to the standard heater cartridge on the dragon.
Would not be surprised to see phaetus releasing a new version with ceramic heater
It's remarkable how similar the heater/thermistor configuration is to the Ubis hotend, which has been around since the first Printrbot.
There is still things not shown publicly yet, but the whole revo hotend heater/block is new and developed in house
Thats like saying a 98 honda civic is the same as a tesla S cus they both have 4 wheels
@@CanuckCreator yeah, I realize there are differences like the ceramic heater, but philosophically they seem similar compared to the old E3D approach of an off the shelf cartridge heater in a block of aluminum.
The single piece (I wonder if it's a friction weld) nozzle and heat break seems to be the big disruptive innovation here.
What is the flow rate?
interesting test but I think measuring heating up isn't quite fair as to how it performs with fast flow changes. the revo has the temp sensor in the heating core, so of course it's fast and stable to regulate. it doesn't mean that the nozzle is actually that fast or stable
the question is though, is it high flow
That’s what was in the box!
Doesn't that just show a low thermal mass?
Yeah, I think this just shows really low thermal mass. I would think slower is better here. In practice higher thermal mass would mean lower temperature fluctuation when printing
@@dakotaadra1044 not understanding this either. :-) would be a different thing if you'd change temps across your print but thats not really a thing as far as i know. and transferring electricity into heat should be totally equally efficient given appropriate cable diameter.
It seems like it does, but if you don't really need the larger thermal mass to get good prints, then you get all benefits.
Yes, however, there is a proper PID tune in place to compensate.
Isn't there a big mass difference?
I would expect if you have more mass to heat up, that it will take longer to do so, however during fast prints, the more mass might be better to keep that temperature.
Would be an interesting test, checking the opposite:
1- Let both hotends heatsoak for like 10 mins at 240º;
2 - Then do constant extrusion;
3 - Check within 30s, how long was heater on.
I have done some testing with seeing how well the hotends handle doing rapid changes in flow rate (basically idle, then extrude 100mm @ 5mms, then idle again)
And the revo consistantly showed a more stable temp chart. A lighter mass reacting quicker to demand works better than a larger mass with larger heater, but being less efficient and slower to respond as the mass takes longer to heat up and cooldown
It's intuitive that a bigger mass is going to be more stable, but there's also something to be said for a responsive heater. You could do more testing for sure, but it's a good sign that the overshoot on the Revo was actually better. It a good sign that it's not hurting from a smaller mass.
@@CanuckCreator yes it reacts quickly, of course, and the temperature variations are smaller. However, "more stable" is quite generic: even if making it smaller is always nice, does it bring any actual improvement in the final print quality?
A normal Dragon after all does not oscillate more than two degrees (more or less?) which is not an issue already.
I don't mean this as criticism, just as request of further data, since you have more experience and more means to test the two to prove the presence of actual practical advantages.
Thanks anyway for this video! I'm curious to try my old UBIS hotend, which used a ring heater already 6+ years ago.
@@CanuckCreator it could be interesting to compare the duty cycle once the temperature is reached: if the duty of the Revo is 1/5 higher than the Dragon, the heat dispersion to air is the same. It's just a measurement of the quality of the insulation, since the Dragon has a sock already
@@CanuckCreator Maybe comparing filament throughput would be nice. With the thermistor and heater in the same package but only losely (compared to thread of a v6) connected to the nozzle-heatbreak combo i wonder if the heat transfer is becoming a problem. In that case more stable chart could mean it did not transfer heat. As Olaf said also duty cycle is important. (Also i wonder how this will work out for hardened steel/titanium nozzles where conductivity becomes even more of a problem)
there are 70W 6x20 heater cartridges out there
So a redefined Trianglelab ceramic heater? that one fits a V6 thread heatbreak, but I have NOT tried it, bought it about a year ago or more..
Could we make a cheap Revo like hotend or would we be limited by low flow of the compatible hotends?
...those were exactly my thoughts when I saw the first videos of the Revo: a souped-up TL ceramic heater kit with a proprietary spring-loaded nozzle.
You can combine the DIY variant with a cloned Hemera titanium heat break/sink combo to get closer to spec ;)
One guy in the comments on AE claims 210°C in 15 sec...
The heater is a completly different type of heater the ring style heaters out there already, its also a PTC heater which is safer and the whole package was developed by e3d internally
@@tomwilczek1759 Maybe I should just try it out, actually started to do some visual research of the hotend compared to whats out there. But the Triangle Lab ceramic heater is limited to a V6 thread, so the question would be what is the best heatbreak metal these days? bi-metal or titanium it seems. If this actually is a good heater it would certainly be cheaper then the Revo with those nozzles. Maybe stacked two of these with a volcano nozzle?
@@CanuckCreator So in other words you have no faith in even trying the round ceramic heater like the one trianglelab have?
Sadly if this new hotend from E3D is good, and will be the new "norm", it would probably hurt the 3D printing community now that they will start patenting stuff.
@@CanuckCreator no doubt that E3D put out a well-designed package - and a properly designed PTC element indeed won't overshoot its equilibrium temperature which can be set to the "tap out" point of around 300°C and everything below is controlled by voltage adjustments aka PWM. It's just hard to imagine they found a ground breaking new concept for thermal control that hasn't been around in the process industry for the last 30 years :)
Nice! Can you please attach 24V directly to it and see if E3D's claims are correct, that it prevents thermal runaways?
its a PTC heater, as temps increase power draw decreases so it cant "run away" like a traditional heater
@@CanuckCreator Yes, please put that to the test! I want to see how hot it becomes! :D
@@certified-forklifter after I get a second one. Dont want to kill my best hotend yet lol
@@CanuckCreator A PTC... That's pretty clever.
Nice, but why E3 can´t get rid off their groove mount. ( I remember ur Video "Choosing an Hotend" )
There are un-announced mounts, heres hoping for a rigid mount
The beta units had the screw top mount, which while not rigid worked quite well in testing (still hoping for rigid down the line)
Which heatblock design is the dragon your using? The TL version (with v6 size and shape) or the phaetus block? And is it aluminum or copper?
Copper Phaetus HF Dragon
Yes it has more thermal mass, but also a larger 50w heater (vs revo`s 40w)
@@CanuckCreator that's pretty impressive for the Revo. It would be nice to see a volcano style down the road. :)
@@CanuckCreator I'm curious what the flowrate of the Revo is compared to this HF dragon setup. Quicker heat up time doesn't really matter if you're still waiting on the bed to heat up (especially when you have a heat soak macro).
@@martijnpeters the hf revo hasnt been shown yet. So current one wont outperform a HF. But im currently running 15-18mm3/s real world on mine depending on the filament plastic type with no issues
If the Revo have a better response and less thermal mass, does this mean lower numbers in pressure advance?
I would argue no, pressure advance is about friction / resistance of the hotend and the needed pressure to extrude filament.
We need another "Choosing a Hotend" video with all the new options that are out there.
Is the revo out yet? If yes then revo. If no. Wait or dragon/dragonfly
@@CanuckCreator Nothing else worth considering? Mosquito, TL Spiral Tower, Dragonfly HIC HF? I know you've said things like the Nova hotend is just a shiny volcano.
@@CanuckCreator I think we need some proper flow rate tests and print examples before we all go buy this thing. So far it seems like it can barely keep up with a standard flow dragon (luckily E3D's official numbers of 10 mm^3/s seems to be wrong), and i'm personally of the opinion that there's zero point in buying a non HF hotend in 2021 with the machines we have available. Not to mention that a Dragon HF prints slow just as well as it prints fast. Why would we settle for less? To me the Revo looks like it's targeting beginners with slow moving bed slingers, which is fine, but then the price is just all wrong.
@@miklschmidt Its keeping up with my normal "pif speed" profile which maxes out around 18mm3/s, it performs a tad better than most SF hotends, but it isnt a HF....those are comming later.
Also, what is flow? e3d says this, slice says that, what temps, what nozzle size, what filament type? Theres so many variables and no set standard for comparison it really doesnt mean anything cept for marketing.
IF i run a 1.8mm nozzle with a plastic that melts at 190c running the hotend at 300, look at how high my flow is!
@@CanuckCreator Yep that's about what i've heard. My SF dragon does 20-22 on a good day, depending on the filament. If/when the HF comes, i would debate if this current model even needs to exist. I've found no advantage to using a standard flow dragon over a HF dragon in any scenario.
Completely agree with you there, we need a proper standardized method to accurately measure flow. I know someone is working on that :) Until then, it would be sufficient to just provide the numbers and the testing methodology for PLA on different nozzle sizes (At least with a 0.4mm nozzle). Then people can reproduce and compare themselves. It will never be apples to apples, but right now we've got nothing. It's pretty complicated to compare flow accurately as it depends on so many factors, but we can still get within some range of "useful" without too much effort :)
Heat up speed is nice but 8d prefer which one was more stable under flow load
Same test done high flow stability as well
It is absolutely stable. (After pid tuning) I had +-0.2°C deviation while printing. Even on higher speeds.
@@danielkrah5129 sweet, good stuff. U talking e3d?
@@felurianmasters4369 yes revo beta unit
@@danielkrah5129 ty
Revo isn't so bad to be honest. Posted on your short about proprietary being shit for the community, and while I stand firmly by that, it seems like E3D isn't a dog company. I watched someone interview Sanjay and he explained it quite well. $120 USD they approximate for this, including the nozzle set. Pretty amazing tbh. Compare that to other dirt bags we see in the 3D printing industry that are trying to milk the consumer.
Im all for access to everyone for as much hardware as possible, i like when we all get to enjoy the new toys
But with the current ecosystem, nozzle, heatbreak, block with a heater shoved in it, all seperate, and assembled like legos...we`ve reached the limits of that system.
Whats the bigges leap in this setup in the past few years? Standoffs?
wooooo
Well it does suck that other than mounting theres no legacy support with the new hardware...but this opens and allows a leap in the tech we would not get otherwise. and since E3D has spent alot of time and money developing the tech in house, and its a whole new thing...yes, they do reserve the right to patent it.
@@CanuckCreator I meant companies that are actively trying to milk every penny they can from the community/industry by locking everything down behind their expensive paywalls.
Look at the companies that Essentium are competing against. They are actively trying to stop other shady companies from ripping the consumer/industry off by charging extra to print different polymers. That's how it should be.
Like I said, hats off to E3D for how they are handling this innovative nozzle design. It's priced very reasonably.
Considering how many headaches I had with Dragon I'm excited for Revo. I gave up on getting the Dragon working without jamming or a ridiculously overpowered fan so I usually just use a volcano on my CR-10 V3. Tempted to get the Revo now just based on the easy switching of nozzles. Z-catch works ok on Volcano but it's still a minor PITA.
Too bad you still need to wait another 30min for the heatsoak…
I have a question about the HF hotend, do you limited the Power of the 50W heater in HF hotend to get a smooth temperature control?. anyway. use a PTC for a 3dprinter heater is SMART. E3D YES
I prefer a dual 40w Maxiwatt volcano.
You still have to wait on your bed to heatup, so does it matter?
Because the faster your hotend can respond to increased and decreased demands ( flow rate changes. Part fan speed changes,) The better and more consistant your prints will be
@@CanuckCreator Even if this was true, and it's a pretty dubious claim, your experiment just shows the Dragon has a higher thermal mass. That will smooth out any thermal changes due to changes in flow. Thermal mass in this case is equivalent to adding a capacitor to reduce ripple in an electrical circuit.
@@MatBettinson the only direct one to one comparison was vs a standard flow mosquito. As thats what i had previously in this machine.
After the swap temps where more stable..prints were cleaner and small slow details more defined after during infil speeds and flow. And overall finish was improved with a bit of a speed bump from a bit more flow capacity
@@CanuckCreator I guess I have to take your word for it but sheesh man, you can print with a delta of a few degrees and the differences aren't huge. The stability you're talking about here, I would assume this is within a single degree. I still don't see how lower thermal pass gives you greater stability, particularly since Klipper's PID is infamously choppy. I want to see these guys succeed but there needs to be some evidence of real benefit to counteract the inclusion of prioprietary nozzles, heater and thermistor (and a shit 104GT at that). It needs to be _tangibly_ better to justify that stuff in my view.
@@MatBettinson pt1000 will come. Later beta units remember. Not everything was finalized
Have you tried any high temp printing with this? I fear the ptc circuit would make the pid unstable at different temps than where the tuning was performed.
I had this issue with a volcano nozzle. Without running some filament through it during the tune I'd get horribly unstable pid settings going plus or minus 20c because my power demands were much higher compared from no flow to all the flow.
Now this was probably my fault due to power supply constraints, but with less thermal mass i wonder if the pid tuning can be accurate enough.
printed up to 280c (nylon) no issues, i have a dedicated machine for higher temps plastic so didnt push it much beyond that
it's cool but for 120$ i rather spend the extra 30 for a mosuito that also has cheaper nozzle replacements and more options like tungsten carbide
Sure $30 more then buy the heater and thermistor as well. Unless your swapping them from another hotend.
It will suck being limited with nozzles for the time being but those will come.
@@jasonwoody8041 it's just too expensive too. Maybe china will clone them. I just can't afford to spend 300$ for just hotends in my printer anyways
@@ameliabuns4058 i hear you there. But its new. They want to recoup the R&D costs asap. And the cost of metal is way up. If your not needing to replace a hot end just hold off till you can afford to lose the money. Im going to keep my hot end until that time, but i love how that comes molexed right out the gate. And the super low weight.
@@jasonwoody8041 oh the molex is great :p, maybe triangle labs will make a clone anyways. For people like me my designs just aren't possible with expensive stuff.
I wish some manufacturers focused on budget products
This test is nice but useless. The Revo has thermistor and hearing element extremely close together. It heats up fast but you don't know when the heat has reached the filament. For the dragon, the thermal mass buffers the heat and the reading and actual nozzle temp are likely closer together. Only a part that's quickly printed and has rapid changes in flow will be able to show which system is better.
Oh boy. My wallet is hiding from me again.
Daaaaaaaaaaaaang
So what? Ceramic heaters aren't exactly new tech.
The heater is an in house design that they`ve spent quite awhile working on, its not just a ceramic heater that you see currently on the market
Exciting to finally see a leap in hotend tech. In this case breaking backwards compatibility is worth it. For this magnitude of improvements they had to change the design fundamentally. If you think about it there is a lot of improvement here.
Revo have small heatbrake , that's the reason it is quick. But is hasn't thermal capacitance that means it's vulnerable to big temperature changes from part cooling or big volume extrusions. Perhaps PID can handle that perhaps not.
It can and does, its the most thermally stable hotend ive run so far. Its ability to quickly adjust its output to fans and filament flow keeps temps stable
@@CanuckCreator did you have same version of klipper on both machines? it's actually changed default pid response time for hotend from 2 seconds to 1 second(which can be even smaller)
@@CanuckCreator Yes i had max 0.3°C deviation when the part cooling fan kicked in.
Meanwhile Slice Engineering is getting ready to patent this...
And their socks
I like the vids but that nose whistle though...
I don’t see why this is important. The faster a heater block heats up, the faster it can lose that heat. Thermal mass is needed to maintain heat. Less thermal mass the harder the heater has to work to keep up with the needs of the printer.
A chunk of copper is always going to be better, more stable and more consistent.
The revo won’t be printing fast anytime soon.
It can hold a stable temp better than a larger block because it can actively adjust the heater output quicker and respond faster while using less power.
So it wont overheat filament due to a large latent thermal mass when going form a High flow print section like infill to a small slow detailed area, and wont lag on heatup when doing the opposite.
The machine I Have the revo in previously had a mosquito. Temps are now more stable, and was able to push flow rates a tad higher.
200 mm/sec isn't fast for you? My Revo beta Is more consistent than anything I've tried before
@@CanuckCreator Don't see how maintaining a temp (here 240) would "overheat filament"... 240 is still 240 even if it weights 1 pound
Second, the only extra power it will use, is the power to heat the extra mass. After that it will be same for both.
But granted if it can maintain stable temp with less thermal mass, translating in faster thermal changes, that's all that is needed
Edit: the limiting factor will be the temp sensor and how quick can it register the change. ..
I’d rather have something I can get at a reasonable cost.
You aren't who they are selling to then. This is a brand new product, it isn't going to be sold at a cheap price.
I personally believe it is fairly reasonable for a genuine product.
120 USD for the whole hotend kit and 4 nozzles, isnt bad compared to other "domestic" hotends
@@CanuckCreator I absolutely expect TriangleLab and Mellow to come up with clones. $120 is pretty decent already and they will probably make it as cheap as a Dragon
@@jonathanlarsen4177 absolutely. Everything can be cloned. Then E3D will start behaving like Slice. Really bizarre actually that these Voron guys are pushing this proprietary hotend in contrast to how they have reacted to the whole Slice thing. Feels like there is something in it for Voron
I like how people are super offended that a clearly better design might work better. I still don't think we're quite there yet - we shouldn't be in a position where we're indirectly heating the filament path, we should be inducing heat inside the nozzle, precisely where we need it at a gradient we need it using electromagnetic induction, where that gradient can be tuned and shaped for the best outcome (i.e. performance and print quality). Also, if anybody who makes 3d printer things uses that you can have that one for free, but if you start throwing patents around I will show up with prior art.
com'n man, this is highly questionable what you're doing and your rhetoric, this is marginal difference, who cares if the diff is 30 sec. My SF takes 1.20 min to 250C
in real life the difference is night and day
@UCk9o8XUgm6fv8l8I8msA7rQ sounds like someone is mad that slice got the boot rather then E3d. Slice took an already existing design from the heater part of the community and made it proprietary. Then decided to overreach all makers by enforcing their copyright on them even when they are helping sell their product by making it more accessible. Trying to bash a product you haven’t used versus him saying it’s a decent hotend and will only get better I only have a beta. You’re blind to your own emotions on your preferences. He is testing a beta unit. Not only that if you watched the whole stream. He states that he likes the best thing out at the time and what works. He used slice at one point now he doesn’t. Nothing wrong or biased about it. Denying something you don’t even have experience with is just dumb. If you don’t want to spend your money then don’t. Everyone in this community forgets we’re all early testers of 3d printing and the more people who change and improve things the better these machines will get. Right now everything is changing and getting better fast if you want to stick to what you know then keep printing on your prusa with your slice hotend because it just works…
It's also got huge quality-of-life and reliability advantages, like being able to swap nozzles so easily. And swapping the nozzle and heatbreak at the same time means that you can clear more types of clogs with the swap. You can keep printing while you unclog the nozzlebreak. With a traditional hotend a clog in the heatbreak is a NIGHTMARE, as you have to disassemble the toolhead and hotend just to get at it.
There's also a design flaw in existing designs, in that nozzle changes can break the heatbreak. This is true, even for supported heater blocks like the Mosquito or Dragon. I broke the brazing on my heatbreak when changing a nozzle. The whole thing ended up fused together and I had to replace the entire hotend. I'm not the only person who's had this problem.
@@fwiffo with revo you might just strip thread or rip steel tube of heatbreak apart since you definitely prone to apply excessive amount of force, there's no "threated desings" which can actually address such issue
@@SergeiSugaroverdoseShuykov I very much doubt that excessive force will be an issue when you're finger-tightening it.