This is exactly what we, the contenders of the speedboatrace, owed to the spectators: Explaining and teaching for a better understanding! You did an amazing job here. I encountered so many people being curious about how to improve their 3D printing skills by using modern methods, slicing tricks and clever designs for their machines. Thank you! It encourages a lot to go on and to improve the presentation of our work!
This is hands down the best 3D printing video i've ever seen. Period. The amount of knowledge packed into these 21 minutes is just beyond compare. THANK YOU for taking the time to explain all this. It will make everyone's prints better and faster. You're amazing
Michael, being involved in the making of V-Core 3 and fathering EVA I would like to express my gratitude. Having this series not sponsored by anyone was important with the additional agreement they signed allows me to believe that you actually like all the hard work put into the 2 projects. Thank you for the time put into it and thank you for doing a brilliant job with describing all that is important about EVA. EVA makes no money outside of donations so when I'm talking about promotion it's all about bringing smart people's attention to it which allows it to iterate and get better and better - your series made a huge dent already. Thank you!
Pawel, your design work is top notch. As I reflect on this series I see so many things the community wants in the Rat Rig/EVA. It's a blueprint for how things should be done. As I said as a guest on the livestream, working on this printer was always exciting and something I looked forward to.
Excellent description of the journey to a speed benchy! Each problem resolved will reveal another one, tackling them one by one is the way to go! Thank you for this fantastic video and thanks for the mention :)
@@miklschmidt Thank you man! We are all in this together. My dream is now to see material manufacurers to jump in and provide us with speed dedicated plastics.
Michael, I can't tell you how happy it makes me that you've not only participated but went above and beyond to show why the #SpeedBoatRace exists and how it can be helpful. I never thought someone as popular and respected as you would participate and I wanna say thank you!
Word! Also thank you for the initiative Keith and who ever else involved. This has been a remarkable learning experience for myself as well as many others!
I have recently jumped into printing and I have to say that once my wife understood what I enjoyed about it, I have been more free to explore the hobby. Currently starting out with my Ender 3 V2 and heavy modding to get the most out of it but want to eventually get to something like the Rat Rig or the Voron because these machines are meant for quality as well as speed. Watching this series has just reinforced what is possible with 3d printing and I thank you so much for all of the wonderful work you do to move the hobby forward and showcase the enjoyable aspect of it.
This explanation is gold! I always thought that the speedboatrace was useless because the printquality was crap. But now I understand that the road to a fast benchy is the road to better understanding the printer and the quality/speed settings. I just got my first printer, but I've been lurking around this and other channels for nearly a year now (couldn't make up my mind). I'm far away from trying these things, but still, they're a valuable lesson. Thanks for taking the time to explain it to us. It's most helpful!
That you have built and tuned a printer so that it can print a high-quality Benchy in 23m40s is an amazing achievement. Congratulations on reaching this goal, and thank you for sharing how you did it. You and all the others that have participated in this challenge have made a great contribution to the 3D printing community.
19:44 IS the point. F1 race cars are science experiments where innovation eventually trickles down to cars we all drive. Pushing limits with set rules is a great system for pushing progress. Well done. Congrats.
Wow. This whole series was great. I am someone who doesn't even have a FDM printer and only have a Saturn right now. But this was very interesting, enjoyable, and educational. I'd love to be able to buy a printer like this, ready to rock and print right out of the box. I have the skills to build it, but tuning it is well beyond my skillset. But I really enjoyed watching your journey through it and the results you got. So thank you.
The speedboatrace event is what got me to finally try klipper and fine-tune my Ender3 for reliable speeds. I've always been very hesitant about tampering with speeds and acceleration, but after a lot of trial and error (and help from an ADXL345 accelerometer and the ever-helpful TT calibration and tuning guide ;) ) I've found that I can reliably work at around 5k accel and 150mm/s, which is leagues above my usual comfort level of 500//60mm. No 10min benchy for me, but I've at least gone down from 1h45 stock to now less than 30mins with a negligible loss in quality (in particular, top surfaces seem to curl up around the extrusion lines a bit; I assume that's either still a cooling issue or just an artifact of the faster moving nozzle pulling on the laid-down filament more).
YES - Michael you are a teacher at core and with ❤️. Before #speedboatrace was a "so what" (maybe with a bit of awe). YOU clearly made it useful. THANK YOU!
I think you should blast the model with a desk fan. This is what a lot of the people do. If you hit it from all sides you would be able to increase nozzle temp and flowrate. You are the gold standard in 3d printing hardware and tutorials.
Although this is an option, would you want to do this for every print in the future? Just a thought :) might be better to just try and increase the efficiency of your toolhead cooling like for example the Annex k1/k2 printers, with the 5020 fans or the toolhead Josh is using on his v2.4 or the insane toolhead Miklschmidt is using on his ratrig or go the berd air/hevacs route Mirage and Vez are taking.
@@someguy2741 the topic of the video is the speedboat challenge indeed :) the idea of the speedboat challenge is to improve day to day printing speeds while maintaining quality :) like I said, some people have been doing it already and it has sort of worked, but it's not a permanent solution to increase your day to day printing speed.
Fantastic and loved you explanation of what doing the challenge has helped you achieve. Now all we need is an out of the box printer that incorporates what you have learned and I’ll buy it
Cool study! Can't agree with the "no loss in quality" though. I don't know if it has anything to do with the filament, but 20:04 surface looks very uneven. Would be great to see a similar study for some of your other printers (same firmware and settings with an acceleration that works across them obviously) to show difference in quality for different architectures.
So I just realized for the most part out printers are the exact same setup, skr pro 1.2, mosquito magnum, orbiter extruder, your klipper files just made my life so much easier! Thank you
A video of how to print 200mm/s with minimal upgrades (less money spent upgrading) required on a ender 3/pro/5 with all the slicer tunings explained would be a great video! Ender 3 sells for as less 175$AUD and I got a ender 5 for 220$AUD last week EOF sale. Just need to look in right places at right time.
Tbh reliably printing 200mm/s on ender 3 requires changing many components. You will need to change hotend, cooling, motors which will cost you quite a bit and still it will be terriable ghostingways. Im talking about at least 2k acceleration since lower you won't really reach that speed. Ender 3 just Isnt a printer for fast quality printing. Ender 5 though you could do a corexy conversion and change earlier mentioned components. Though upgrades will cost another Ender 5
Hello Michael, did you use G2/G3 in your G-Code? That will improve quality and time too.... That can cut your G-Code file to 50% changing hundreds of stupid little G1 lines to one G2 or G3 arc. This will keep up your speed caused by longer distances and your machine is running much smoother. Best regards from Germany Martin
Would be great if Rat Rig V-Core 4 included all the improvements you suggested by default and every part came in an ultimate version kit within a $1500 price point.
Respect to your results! I don't mind for the time, I am more interested in ultra-quiet printers. I don't print that often, but when I do, I'd like to have it not disturbing me while working :)
I think there’s some parallels between high speed and low noise. Frame rigidity is the most vital thing, optimizing fan ducts so you can run minimal fan, fine tuning the motion system. My printer is barely audible when printing 150mm/s. I sit most days about 4 feet away from it.
Time is not my top priority, BUT, when you print large parts, you not only saving a lot of time, but reducing risks as well, my longest print yet took 20 hours, but this was a tuned and optimized setting i worked on, the Default settings would take 36 hours. Few months after that, i discovered i had problems with cooling and the extruder, few weeks ago i changed the parts and fixed the problems, now i can use a bigger nozzle and print the same part in less than 12 hours, and No, im not cranking speed and acceleration, because i cant, more than 120mm/s or 6mm/s jerk causes layer shift on my printer. learning how to tune your slicer for your printer, helps A LOT in time AND quality.
@@originaltrilogy1 - Gradual Infill, big nozzle, faster infill, are some of the biggests time savings. Also, if you have a bowden printer, make sure you have a descent extruder, i had a bad one and needed to work with 10mm retractions, changed a few weeks ago, and now i need only 4mm with less stringing. I keep hearing people saying 1/2 width in layer height, but my best quality is 1/3, i can print with 0.6mm nozzle and 0.2mm layer height with better quality and less walls, sometimes i use 0.8mm nozzle and 1mm wall thickness on big parts. I use Ultimaker Cura for slicing, and the software have tons of settings to tune, its a bit hard and time consuming to master the software, but its very rewarding.
You've put alot of time and resources into helping the community. Thank you for everything you've done to help every random person picking up 3d printing to have the most success they can.
Get a board that can handle higher driver currents, can perform calculations faster to turn arcs at high speeds, get a lighter bed, and then dont use the Ender 3 profile in Cura. Nobody talks about that last part, but Cura assumes the limits of the Ender 3. You gotta push past the 500mm/s2 accel cap on the stock firmware.
So the nozzle is drifting rally style to get around the corners faster? Love it! Excellent video. There are enough tips in this video alone that it should be a reference source.
This is one of the best examples of diminishing returns - a printer that costs 10 times as much but only prints twice as fast. And yet, it is also one of the best examples in a different way - as an example, it's something that can be learned from to boost performance on existing printers that people have already purchased, as well as used as a guide for optimizing part selections for those looking to build their own custom machines, or even those looking to buy premade ones and simply wanting to more easily filter out the units that won't be as useful despite having "the same specs" etc Truly a fantastic video as always and even though I have no interest in building this kind of printer any time soon (due to budget and space constraints) it's still been an incredible series packed full of great information :)
I doubt many $200 printers can run past 80mm/s reliably... but that said, diminishing returns are real. (ender 3 pla speeds are 60mm/s for decent quality)
@@eldricliew6223 yeah, I’ve been experimenting with the CR-6 which is just a fancier version of the ender 3 and it’s definitely possible to do 100-120 on it but then again that’s comparing a completely prebuilt machine (which is always going to be overpriced) against a built-from-scratch rat rig so my 10x price estimate was what one of Michael’s machines would cost if you bought it ready to print out of the box (designed, manufactured, assembled, calibrated, etc) like the CR6 is
@@eldricliew6223 you might be surprised if you take a look at some of the other #speedboatchallenge entries docs.google.com/spreadsheets/d/1lFiJi-X3Xm3hh3I9Ty9dfACMiBWxHFAOUeiU1km9m6I/edit?usp=sharing
@@C1Rob1990 Honestly I'm surprised by ruclips.net/video/6xFSiSomz9k/видео.html. Seen quite a few stock printers which don't run much faster than 60-80mm/s.
I just wanted to say thanks for the rat rig v-core 3 series, I actually got one as a result of your videos and have been using this series as a guide. It's a pity there wasn't an affiliate link since I never even heard about rat-rig until your first video.
Back during the first month of the pandemic, the local 3d printing community in my country banded together to mass produce face shields. We started with a common design that would print in 3hrs using standard settings and parts. By the end of our efforts, we had some members reach 10min
Pushing the hardware is how you find problems, bottlenecks, and waste. This is how you move a technology forward in leaps and bounds instead of crawling along. I love it!
This and your ratrig v-core build series convinced me. Ordered a v-core 3 with the same components as in your build. It will replace my 13 year old makerbot replicator 1 😄
Question: Have you ever considered moving your cooling fan completely off the rig altogether? I'm thinking you run a bowden style tube (probably 2 or 3 of them really) and point those at the part. The fans you are using probably move around 35CFM without the ducts, and likely much less than that when there are ducts. A decent air compressor can move around 5scfm, but that's at 90PSI, and I think it might do much much higher at lower pressures. If you ran a bunch of tubing through a big cooler full of ice water, you could both lighten up the print head, and push some SERIOUS cold air over the part. I know nobody would run such a cooling setup on a regular printer, but in the name of science and friendly competition I think you should take a crack at it.
I run 60 psi on small metal pipes with qty5 1mm holes on them. you can see a reference to my printer at 19:24 or here: ruclips.net/video/65FVQ1jArME/видео.html
@@MirageC Yeah... funny, I made that comment before that section of the video that talked about compressed air. That's exactly what I was thinking. Do you do anything to actively cool the air? What CFM does your compressor do? Nice job!
@@LabRatJason I have a shop air compressor hooked up to the printer. Only one piston and it barely keeps up. I am planning to address that and add better drying of air. Compressed air will cool on its own when pressure decreases at nozzle exit. Compressor tank and hose length are offering enough surface area to dissipate the initial heat increase cause by the gas compression.
The speedboat challenge has a lot of parallels with the 1930 airplane races. Both were new technologies that pushed the limits of their machines. It feels great and exciting.
I've been shy about modifying slicer settings, but seeing just how much improvement is possible just pushed me to start modifying everything! Thanks for the great vid!
I got my first printer 5 months ago. I wish I got in earlier, but I think I got in at a great time. I got an Ender 3 S1 Pro for $360USD and the Sonic Pad for $140. So 500 in and I went from taking an hour for a Benchy without the pad, to getting them down to 30 minutes while having them the same quality. I didn't even have to do any python or anything. Next stop, I'm getting 2 Bamboo Labs P1Ps, each for the same price as my Ender and Pad. Enders are great as a first printer, if you're good at modding a tweaking all the weird things Creality does. But I do think these days, getting a P1P if you just wanna print and not worry about open source modding, is the way to go. Hindsight is 20/20. Now I just need to know what's the best consumer CNC mill. A 5 axis would be great.
I'm excited about that one too. I've wanted to change the board for a while so I could run linear advance. The deciphering of the ribbon cable and interfacing with the touch screen were a burden, but I think with Klipper I don't have to change any wiring and the TFT I have looks to be a size match for the Creality screen.
I think this is cool and I've loved watching you do this, but I am far more interested in seeing what the best results are you can get out of a totally stock printer without upgrading any parts, other than the nozzle of course.
@@gingerflow5594 but you can adjust these things with slicer tuning, smaller diameter nozzles, and filament dry boxes. It just seams like a more cost friendly competition, that’s all.
PID is not the best solution for controlling the hot end. It is however an easy solution which is why it is used. PID auto tune does not do a good job of setting the parameters which is probably part of the reason for the under extrusion at these speeds. The big problem with PID is that it is always reacting to what has already happened. It is always trying to play catchup. You extrude a bunch of material and the PID loop doesn't know about it until the thermistor (or thermocouple) sees the drop in temperature. The heater cartridge power is then turned up and hopefully it will eventually catch up. There is both a delay in sensing the temp and a delay in adding energy to the heater block. The temp sense delay is bad but the heater block delay helps keep PID from overshooting and burning the plastic. Hand tuning the PID values will help a lot as the auto tune doesn't generate values that work will near the limits. People will often turn up the extrusion temperature in an attempt to increase print speed but in the situation where you are already extruding at near the theoretical maximum this will work at most for a few seconds. And greatly increasing the acceleration and print speed does this. We know that it takes a given amount of energy to raise the temperature of the plastic. This value will vary slightly with the type of plastic. For optimum control we need to know: 1) The amount of energy it takes to change the temperature of the plastic. 2) The amount of material being extruded. 3) The temperature of the filament. Knowing the room temperature would be good enough. 4) The desired temperature of the molten plastic. 5) The delay in sensing the nozzle temp. 6) The delay in changing the temp of the hot end. 7) The amount of energy lost to the environment. This is what is lost when heating the air around the hotend and a large part is due to part cooling fans. 1, 5, 6, and 7 could be figured out during the warm up and nozzle prime making it self tuning. It can also tell if the nozzle is blocked or extruder is skipping when temps don't change as expected. One of these years someone will implement this in Marlin, Klipper, Smoothie, or some other software. Hopefully nobody has patented this and that would make this posting representative of prior art invalidating any future claims.
WOW I used to drag race my old 1964 Corvette so i know the problems with going fast. Boy was this a fun video. Burning rubber and blowing engine not really but blow the sides out a benchy because your cooling was too low relates. I can not wait to see what my pig of a bed slinger will get to in that contest they should have catagories, modified stock class you know. Take care and thank you again.
Hey man, love your videos! Could you make a strength test comparison of fiberglass reinforced resin vs PLA vs ABS and compared to other filaments in a strength test?
Thanks for revealing the origin of the rattle. I thought it was the hotend. P.S.: Interesting video. Compared to CNCKitchen it seemed to go very smooth using the Ratrig.
From memory he had issues with his machine distorting from the filament used to make it. Once Stefan fixes this, the smaller Voron 0 should really fly. I also greatly benefitted from the premade upgrade path of the EVA carriage. It was very easy to change cooling and hot ends.
Hi Michael, I think I am about to embark on this journey to build this printer. I really want an enclosed 500mm^3 CoreXY printer. I currently have an Ender 5 Pro with a decent amount of mods so I feel confident in taking this on. I have been eyeing a Voron 2.4, V-Core 3 or Exocube (I have Exoslides on my E5Pro) but I think I have settled on the V-Core, I just love the 3 point kinematic bed. It's brilliant. Although I am very curious about the latest Exocube XYZ kit that uses 4 individual Z-axis motors with belts. The gantry moves in all 3 orthogonal directions while the bed is planted to the bottom of the frame. I have not seen many people post any reviews about this setup and am hoping the designer updates the blog section with this latest design. His last blog post was a version that used 4 Z-lead screws rather than belts. Would love to see you build the belt version and test it out! Was hoping you (or anyone with experience/knowledge) could offer some suggestions or tips about the V-core. I see you have updated your hotend from your original choice but kept the same extruder. I see RatRig offers the Orbiter V1.5 or Bondtech LGX. Do you still find the Orbiter to be more than enough for your build? I like the idea of the lightweight design but the LGX just seems much more robust and I like the pre-tension lever design. Another question is whether there are any major concerns to be weary of when building a large build volume FDM printer. Am I better off using linear rails for this setup or are V-wheels/Exoslides better for that large of extrusion bars? One other thing I've considered is buying something pre-built like a Tronxy X5SA-500 Pro and then gutting out the cheap parts with the upgraded mods that I want. But I am not sure I want to take a gamble on material quality of everything that I do decide to reuse. Not sure if you will respond but I really appreciate your time and help. You have been an incredible resource for my modded E5Pro and I am sure many other people feel the same! Thanks again!
Another thing to increase speed is the stl mesh angle. While smaller angles make smoother finishes, they cause a lot of extra gcode in the final product.
That is a sensible challenge. I tweaked a lot of my setting before to get faster prints. Learning to know what the machine is capable of is so usefull, like learning to know when it starts to affect quality. I did not go that extreame though, due to wanting to get fast prototype prints but yet reasonable quality. You never know what your machine can/cannot do unless you try it out. A limiter for me was at some point default hard coded accelleration limits in the firmware. It took a while to realize those.
I love the parallels to this and the automotive industry! I'm new here but based on your outfits I feel you see the parallels as well! If you're not into cars, I suggest you don't start, it's a far more expensive hobby with far greater disappointments hahah! You're very much aligned with the guys in the early 20th that were pushing race cars to their limits imho! Thanks for sharing things like this with us. I'm off to print something at 60mm/sec a lesser quality than your 250mm/sec print now :D
Hey Michael. Not sure if you'll see this, but if you do, have you thought of making a video comparing the two corexy printer kits you've put together: the SecKit SK-Go and the RatRig VCore 3. I'm thinking of trying to save up for a printer kit now, and I'm leaning towards the RatRig, but I'm really interested in seeing how it compares to another kit. Especially going over things such as ease of assembly, documentation, calibration, and upgradability
Michael - will you please give us your “summary” opinion of the RatRig now that you’ve taken us this far? If you can afford it - is this the best printer to buy for under 2K? Thanks for all you do - excellent and very helpful!
Baby steps! ~10 years ago on a Prusa MK1 i was already printing sub 20min benchies on that POS frame. Tho 5mins is starting to get nice. I might get back into this racing -- tho for practical prints i've always thought 1.75mm filament is just too thin, but then again you want to melt it fast and ~10years back you just cranked the temp up ande hoped for minimal degradation. Oh well, give and take :) :) Glad to see people finally getting interested on print speeds. Of to order my Voron or Rat Rig once i sell a few other printers off
You can get rid of the bits of infill in S3D by switching to concentric external infill patterns and change the external wall type to perimeters only and as well the internal wall to "allow for single extrusion fill" it cuts a lot of waste and ends up printing similarly to vase mode specially if your reduce your over lap perimeter.
Brilliant video. One beef: Many “improved” cooler designs completely ignore the physics of fluid mechanics. By blasting air from more than one direction, all you change is the stagnation point, and for “opposed” air injection (including “all around”), the nozzle is right in the middle of the stagnation point, ergo, no cooling at all! I therefore worry about “improved” shrouds that blow from three sides. I don’t think they would be any better that a shroud that blows from one side only. It’s a tricky problem that requires a solution backed up by physics.
One thing I would like to see attempted is to modify the Y-axis to remove the 30mm extrusion. I've seen printers forgo this support in favor of the tool steel linear rail which 𝘴𝘩𝘰𝘶𝘭𝘥 have enough rigidity to hold up a lightweight hot end. At the speeds you're attempting, every gram matters. Inertia is a bitch when acceleration is involved
i guess the best cooling wound be engineering an air conditioner unit on the side and add a new z rod with a blower to follow the nozzle, if i had the time and money i would so do that. liquid nitrogen cooling system wound work in theory but that's a little too risky
I really want to see an expert talking about and testing BerdAir part cooling. I believe it has great value for getting as much air as possible to the work without putting much weight on the tool head. Could you give your input on it?
Just an idea, I got while watching. How about using a compressed air for cooling? It could be supplied by a tube/hose, it would remove the mass of the fan. As a bonus, the expanding air is colder. Compressors with reasonably big tanks are not very expensive and can be used for many other tasks.
That's what MirageC is using on his HeVORT. In terms of noise and room taken up there are some things to overcome but in terms of cooling power and being compact on the printer it looks amazing.
@@cybkeats SKR RRF, micro Swiss all metal hotend, BMG clone direct drive, dual 5020 blowers, linear rails on all axis, dual Z with integrated lead screws. Upgraded the Y stepper to a bigger motor with a smaller pulley, so lots more torque. Adjustable metal belt tensioners. Sitting on a concrete paver on top of foam. Plus a million other small little things.
I own an anet a8 running klipper and my god let me tell you😲 300mm/s 2K accels on the daily with the same quality as an ender 3, and with a 1.3kg bed!!! klipper is a miracle of engineering, and we should all be pushing at least 200mm/s
people called me a madman for pushing 35m benches with 3 walls back in the day, now everyone is pushing faster and faster! this is the true spirit of speed, to push the limits far beyond reasonable, and end up with a stable, high quality, fast printer for daily driving
I haven't done any research to confirm if this has already been implemented, but as you can tilt the bed of your RatRig, would it be possible to program it to "lean in" against the direction the nozzle is travelling while printing, which would effectively reduce the distance the nozzle has to travel for any straight line? It would be sort of like adding a delta motion system to the bed to complement the core-xy of the 'main' printer. I'm not sure how much of an effect this would have on print times for the complexity it would add, but I think it could be an avenue worth exploring.
1. J-Dampers + Tuned Mass Damper 2. Friction Springs 3. Polynorbornene Fittings 4. Inline Extrusion Diameter Sensor 5. Print directly out of drybox 6. Before actually printing your Benchy, try using Purge Filament and Cleaning Filament as well. Also consider switching out your infill pattern for something much faster. (I wonder if Gyroid would be faster than Rectilinear...)
Heck, i thought it was just a race. But that does make a excellent point. If you can print decent extremely fast you can print good or great at a quick pace. Certainly food for thought.
Do you know the reasoning for requiring the use of stepper motors in the speedboatrace rules? Are there other motor types that would allow you to get much faster prints?
The stepper rule was implemented because currently servo motors are still very expensive (compared to their stepper counter parts) and although some have started using them (awesome work Vez and Mirage!) there still isn't native support for them firmware wise and before Vez made his video on the subject, no real info on how to use them. It was done as a means to somewhat even the field (of course a 2k dollar printer and a 200 dollar printer are hard to match up, but this makes it atleast somewhat fair) and looking at some of the top entries the stepper printer times are really not that far off from the servo printer times... (nitram is actually at the top with 48v steppers outright beating the servo times all together, and the Annex Tycho is high up there still on 24v normal steppers!). There have been talks of doing a second contest/class in the future where servos would be allowed. For now the point was to try and push the printer to the limit with the current technology as far as I understood it :)
I would love one of these in the 500mm size over a new Prusa XL which i recently cancelled my day 1 pre order as i am simply sick of waiting. It seems RatRig do not ship to Australia these days and the cost of the printer if you can get it to Australia is over $3000 now. I would buy a Voron if i could get a 500mm size kit.
Awesome video! I understand that going this fast in the speedboattest will increase noise, but i am wondering with the improvements you did to bring down the print time with 59%, how lowed is the printer than? From stock and possibly compared to other devices? Super interested in swapping out my Ender 5 but noice is a big "issue" at home so :) Thanks
In the slicer, did you play around with infill every other layer? The rules allow for it explicitly, but I don’t see anyone talking about it. I’d assume it would only save time, but also assuming the hotend could handle the mm3/s of filament by doing double thick infill? Part cooling wouldn’t be an issue since it’s only infill…
I tried to that right at the end. Prusa Slicer said it would shave off a minute (off it's inaccurate 28 min prediction). I changed other things at the same time so I don't know exactly how much it saved in real life.
This is exactly what we, the contenders of the speedboatrace, owed to the spectators: Explaining and teaching for a better understanding! You did an amazing job here.
I encountered so many people being curious about how to improve their 3D printing skills by using modern methods, slicing tricks and clever designs for their machines. Thank you!
It encourages a lot to go on and to improve the presentation of our work!
Thanks very much, glad you enjoyed. And well done on your fast benchys.
This is hands down the best 3D printing video i've ever seen. Period. The amount of knowledge packed into these 21 minutes is just beyond compare. THANK YOU for taking the time to explain all this. It will make everyone's prints better and faster. You're amazing
Thanks very much for your guidance in this challenge. As the video said, your post on the last one ended up being my blueprint.
@@TeachingTech my pleasure! Congrats on the impressive result!
He is the man. Was your first printer the SD2, like me?
Michael, being involved in the making of V-Core 3 and fathering EVA I would like to express my gratitude. Having this series not sponsored by anyone was important with the additional agreement they signed allows me to believe that you actually like all the hard work put into the 2 projects. Thank you for the time put into it and thank you for doing a brilliant job with describing all that is important about EVA. EVA makes no money outside of donations so when I'm talking about promotion it's all about bringing smart people's attention to it which allows it to iterate and get better and better - your series made a huge dent already.
Thank you!
Pawel, your design work is top notch. As I reflect on this series I see so many things the community wants in the Rat Rig/EVA. It's a blueprint for how things should be done. As I said as a guest on the livestream, working on this printer was always exciting and something I looked forward to.
Eva ftw!!
Excellent description of the journey to a speed benchy! Each problem resolved will reveal another one, tackling them one by one is the way to go! Thank you for this fantastic video and thanks for the mention :)
You're one of the greats my man. Thank you for teaching and inspiring me and so many others with your advancements in this field!
@@miklschmidt Thank you man! We are all in this together. My dream is now to see material manufacurers to jump in and provide us with speed dedicated plastics.
@@MirageC that would be a dream! I would push someone if I had the influence haha
@@KeithSachs as a group we do have a lot of influence ;)
MirageC it was your video that got me hooked. My patrons posted it and we were all gobsmacked by what you were doing.
Michael, I can't tell you how happy it makes me that you've not only participated but went above and beyond to show why the #SpeedBoatRace exists and how it can be helpful. I never thought someone as popular and respected as you would participate and I wanna say thank you!
This 100%
Word! Also thank you for the initiative Keith and who ever else involved. This has been a remarkable learning experience for myself as well as many others!
Thanks very much and congrats on your very impressive speed benchys!
I have recently jumped into printing and I have to say that once my wife understood what I enjoyed about it, I have been more free to explore the hobby. Currently starting out with my Ender 3 V2 and heavy modding to get the most out of it but want to eventually get to something like the Rat Rig or the Voron because these machines are meant for quality as well as speed. Watching this series has just reinforced what is possible with 3d printing and I thank you so much for all of the wonderful work you do to move the hobby forward and showcase the enjoyable aspect of it.
Good work there Micheal!
This explanation is gold! I always thought that the speedboatrace was useless because the printquality was crap. But now I understand that the road to a fast benchy is the road to better understanding the printer and the quality/speed settings. I just got my first printer, but I've been lurking around this and other channels for nearly a year now (couldn't make up my mind). I'm far away from trying these things, but still, they're a valuable lesson. Thanks for taking the time to explain it to us. It's most helpful!
That you have built and tuned a printer so that it can print a high-quality Benchy in 23m40s is an amazing achievement. Congratulations on reaching this goal, and thank you for sharing how you did it. You and all the others that have participated in this challenge have made a great contribution to the 3D printing community.
19:44 IS the point. F1 race cars are science experiments where innovation eventually trickles down to cars we all drive. Pushing limits with set rules is a great system for pushing progress. Well done. Congrats.
I was going to say pretty much the same thing mate. Nice one!
I appreciate your systematic analysis of the bottlenecks and experiments to minimize them. Well done!
Learning how to make something work better and faster is never a waste of time or money.
Wow. This whole series was great. I am someone who doesn't even have a FDM printer and only have a Saturn right now. But this was very interesting, enjoyable, and educational. I'd love to be able to buy a printer like this, ready to rock and print right out of the box. I have the skills to build it, but tuning it is well beyond my skillset. But I really enjoyed watching your journey through it and the results you got. So thank you.
Thanks very much, I'm glad you enjoyed it.
The speedboatrace event is what got me to finally try klipper and fine-tune my Ender3 for reliable speeds. I've always been very hesitant about tampering with speeds and acceleration, but after a lot of trial and error (and help from an ADXL345 accelerometer and the ever-helpful TT calibration and tuning guide ;) ) I've found that I can reliably work at around 5k accel and 150mm/s, which is leagues above my usual comfort level of 500//60mm. No 10min benchy for me, but I've at least gone down from 1h45 stock to now less than 30mins with a negligible loss in quality (in particular, top surfaces seem to curl up around the extrusion lines a bit; I assume that's either still a cooling issue or just an artifact of the faster moving nozzle pulling on the laid-down filament more).
YES - Michael you are a teacher at core and with ❤️. Before #speedboatrace was a "so what" (maybe with a bit of awe). YOU clearly made it useful. THANK YOU!
This was the entire point all along but most people apparently did not understand that 😅
@@C1Rob1990 So very true. Hopefully this video will enlighten the haters :)
I think you should blast the model with a desk fan. This is what a lot of the people do. If you hit it from all sides you would be able to increase nozzle temp and flowrate.
You are the gold standard in 3d printing hardware and tutorials.
Although this is an option, would you want to do this for every print in the future? Just a thought :) might be better to just try and increase the efficiency of your toolhead cooling like for example the Annex k1/k2 printers, with the 5020 fans or the toolhead Josh is using on his v2.4 or the insane toolhead Miklschmidt is using on his ratrig or go the berd air/hevacs route Mirage and Vez are taking.
@@C1Rob1990 The topic of the video is the speedboat challenge so this suggestion is for that challenge only.
@@someguy2741 the topic of the video is the speedboat challenge indeed :) the idea of the speedboat challenge is to improve day to day printing speeds while maintaining quality :) like I said, some people have been doing it already and it has sort of worked, but it's not a permanent solution to increase your day to day printing speed.
Those concepts are very useful for the people that wants to improve their prints! Thank you very much for share your knowledge!
Fantastic and loved you explanation of what doing the challenge has helped you achieve.
Now all we need is an out of the box printer that incorporates what you have learned and I’ll buy it
Cool study! Can't agree with the "no loss in quality" though. I don't know if it has anything to do with the filament, but 20:04 surface looks very uneven. Would be great to see a similar study for some of your other printers (same firmware and settings with an acceleration that works across them obviously) to show difference in quality for different architectures.
I noticed that too. After all of this hard work the printer is probably due for a bit of maintenance. Possibly just something very slightly loose.
@@TeachingTech oh, that's promising news. Looking forward to yet another update. I love your work btw! 🤘
So I just realized for the most part out printers are the exact same setup, skr pro 1.2, mosquito magnum, orbiter extruder, your klipper files just made my life so much easier! Thank you
It was a wonderful journey for me. i got to 18 min, but learned so much in a short period.
Great video Michael. You've done some really good work, and fantastic service to the community by explaining each step.
I use this video constantly as a reference for I dont even know how many parameters. Can't thank you enough for the time you put in.
A video of how to print 200mm/s with minimal upgrades (less money spent upgrading) required on a ender 3/pro/5 with all the slicer tunings explained would be a great video!
Ender 3 sells for as less 175$AUD and I got a ender 5 for 220$AUD last week EOF sale. Just need to look in right places at right time.
Tbh reliably printing 200mm/s on ender 3 requires changing many components. You will need to change hotend, cooling, motors which will cost you quite a bit and still it will be terriable ghostingways. Im talking about at least 2k acceleration since lower you won't really reach that speed. Ender 3 just Isnt a printer for fast quality printing. Ender 5 though you could do a corexy conversion and change earlier mentioned components. Though upgrades will cost another Ender 5
These have been really interesting mate. Great to be able to learn more about printing and all the parameters.
Hello Michael,
did you use G2/G3 in your G-Code?
That will improve quality and time too....
That can cut your G-Code file to 50% changing hundreds of stupid little G1 lines to one G2 or G3 arc.
This will keep up your speed caused by longer distances and your machine is running much smoother.
Best regards from Germany
Martin
Thank you for an informative and easily understood documentation of the Rat Rig setup and your mods. Appreciated.
Would be great if Rat Rig V-Core 4 included all the improvements you suggested by default and every part came in an ultimate version kit within a $1500 price point.
Respect to your results! I don't mind for the time, I am more interested in ultra-quiet printers. I don't print that often, but when I do, I'd like to have it not disturbing me while working :)
I think there’s some parallels between high speed and low noise. Frame rigidity is the most vital thing, optimizing fan ducts so you can run minimal fan, fine tuning the motion system. My printer is barely audible when printing 150mm/s. I sit most days about 4 feet away from it.
Time is not my top priority, BUT, when you print large parts, you not only saving a lot of time, but reducing risks as well, my longest print yet took 20 hours, but this was a tuned and optimized setting i worked on, the Default settings would take 36 hours.
Few months after that, i discovered i had problems with cooling and the extruder, few weeks ago i changed the parts and fixed the problems, now i can use a bigger nozzle and print the same part in less than 12 hours, and No, im not cranking speed and acceleration, because i cant, more than 120mm/s or 6mm/s jerk causes layer shift on my printer.
learning how to tune your slicer for your printer, helps A LOT in time AND quality.
@@brianfhunter I have parts that can take 10 *days* to print. Improving that by 40% would be a game changer.
@@originaltrilogy1 - Gradual Infill, big nozzle, faster infill, are some of the biggests time savings.
Also, if you have a bowden printer, make sure you have a descent extruder, i had a bad one and needed to work with 10mm retractions, changed a few weeks ago, and now i need only 4mm with less stringing.
I keep hearing people saying 1/2 width in layer height, but my best quality is 1/3, i can print with 0.6mm nozzle and 0.2mm layer height with better quality and less walls, sometimes i use 0.8mm nozzle and 1mm wall thickness on big parts.
I use Ultimaker Cura for slicing, and the software have tons of settings to tune, its a bit hard and time consuming to master the software, but its very rewarding.
Thanks for adding the "why" at the end. Makes a lot of sense actually.
Love this guide. Your RatRig makes my SW X1 look like a dot matrix printer.
My X1 was previously my go to fast printer. I still really like it though and will happily use it plenty more in future.
You've put alot of time and resources into helping the community. Thank you for everything you've done to help every random person picking up 3d printing to have the most success they can.
Loves the Series! You are a great inspiration, and I enjoyed every second of the series. I hope to try all this myself sometime soon.
Great video, and I believe this exactly exemplifies what the purpose of the challenge is. Thanks!
Good job of explaining... Next video series could be... How to print faster on a crappy standard i3 machine...
I would love a video about bringing the ender 3 to it’s limit 🤣
me too !
And Prusa Mini too :)
Get a board that can handle higher driver currents, can perform calculations faster to turn arcs at high speeds, get a lighter bed, and then dont use the Ender 3 profile in Cura. Nobody talks about that last part, but Cura assumes the limits of the Ender 3. You gotta push past the 500mm/s2 accel cap on the stock firmware.
lol at the dragster asterix. Car guys are the same as horse girls. Great tips sir, thank you!
So the nozzle is drifting rally style to get around the corners faster? Love it! Excellent video. There are enough tips in this video alone that it should be a reference source.
This is one of the best examples of diminishing returns - a printer that costs 10 times as much but only prints twice as fast.
And yet, it is also one of the best examples in a different way - as an example, it's something that can be learned from to boost performance on existing printers that people have already purchased, as well as used as a guide for optimizing part selections for those looking to build their own custom machines, or even those looking to buy premade ones and simply wanting to more easily filter out the units that won't be as useful despite having "the same specs" etc
Truly a fantastic video as always and even though I have no interest in building this kind of printer any time soon (due to budget and space constraints) it's still been an incredible series packed full of great information :)
I doubt many $200 printers can run past 80mm/s reliably...
but that said, diminishing returns are real.
(ender 3 pla speeds are 60mm/s for decent quality)
@@eldricliew6223 yeah, I’ve been experimenting with the CR-6 which is just a fancier version of the ender 3 and it’s definitely possible to do 100-120 on it but then again that’s comparing a completely prebuilt machine (which is always going to be overpriced) against a built-from-scratch rat rig so my 10x price estimate was what one of Michael’s machines would cost if you bought it ready to print out of the box (designed, manufactured, assembled, calibrated, etc) like the CR6 is
@@eldricliew6223 you might be surprised if you take a look at some of the other #speedboatchallenge entries docs.google.com/spreadsheets/d/1lFiJi-X3Xm3hh3I9Ty9dfACMiBWxHFAOUeiU1km9m6I/edit?usp=sharing
@@C1Rob1990 Honestly I'm surprised by ruclips.net/video/6xFSiSomz9k/видео.html. Seen quite a few stock printers which don't run much faster than 60-80mm/s.
I just wanted to say thanks for the rat rig v-core 3 series, I actually got one as a result of your videos and have been using this series as a guide. It's a pity there wasn't an affiliate link since I never even heard about rat-rig until your first video.
Back during the first month of the pandemic, the local 3d printing community in my country banded together to mass produce face shields. We started with a common design that would print in 3hrs using standard settings and parts. By the end of our efforts, we had some members reach 10min
Great video thanks Michael for all you hard work
Pushing the hardware is how you find problems, bottlenecks, and waste. This is how you move a technology forward in leaps and bounds instead of crawling along. I love it!
Very well covered, first time anyone has talked about the benefits used for quality prints.
You are one of the harbingers to an open source future.
This and your ratrig v-core build series convinced me. Ordered a v-core 3 with the same components as in your build.
It will replace my 13 year old makerbot replicator 1 😄
My V-Core 3.1 400 is on a plane on it' way here right now, so excited to get it built and going !!!!
wow! great work Michael! well done!
Extremely interesting and definitely a really worth-while project!
Question: Have you ever considered moving your cooling fan completely off the rig altogether? I'm thinking you run a bowden style tube (probably 2 or 3 of them really) and point those at the part. The fans you are using probably move around 35CFM without the ducts, and likely much less than that when there are ducts. A decent air compressor can move around 5scfm, but that's at 90PSI, and I think it might do much much higher at lower pressures. If you ran a bunch of tubing through a big cooler full of ice water, you could both lighten up the print head, and push some SERIOUS cold air over the part. I know nobody would run such a cooling setup on a regular printer, but in the name of science and friendly competition I think you should take a crack at it.
Liquid nitrogen?
Interesting, basically like the air-assist on a laser cutter?
I run 60 psi on small metal pipes with qty5 1mm holes on them. you can see a reference to my printer at 19:24 or here: ruclips.net/video/65FVQ1jArME/видео.html
@@MirageC Yeah... funny, I made that comment before that section of the video that talked about compressed air. That's exactly what I was thinking. Do you do anything to actively cool the air? What CFM does your compressor do? Nice job!
@@LabRatJason I have a shop air compressor hooked up to the printer. Only one piston and it barely keeps up. I am planning to address that and add better drying of air. Compressed air will cool on its own when pressure decreases at nozzle exit. Compressor tank and hose length are offering enough surface area to dissipate the initial heat increase cause by the gas compression.
Great stuff, decided to go for a ratrig, very very valuable information, top notch👍
It’s a great improvement! Nice job
The speedboat challenge has a lot of parallels with the 1930 airplane races. Both were new technologies that pushed the limits of their machines. It feels great and exciting.
I am finding your videos to be some of the best aand informative ones around. Thanks :)
I've been shy about modifying slicer settings, but seeing just how much improvement is possible just pushed me to start modifying everything!
Thanks for the great vid!
With the slicer you can always fall back on your original profile so no harm done and well worth experimenting.
I got my first printer 5 months ago. I wish I got in earlier, but I think I got in at a great time.
I got an Ender 3 S1 Pro for $360USD and the Sonic Pad for $140. So 500 in and I went from taking an hour for a Benchy without the pad, to getting them down to 30 minutes while having them the same quality. I didn't even have to do any python or anything.
Next stop, I'm getting 2 Bamboo Labs P1Ps, each for the same price as my Ender and Pad. Enders are great as a first printer, if you're good at modding a tweaking all the weird things Creality does. But I do think these days, getting a P1P if you just wanna print and not worry about open source modding, is the way to go. Hindsight is 20/20.
Now I just need to know what's the best consumer CNC mill. A 5 axis would be great.
I'm looking forward to seeing your series on the cr-10 max.
I'm excited about that one too. I've wanted to change the board for a while so I could run linear advance. The deciphering of the ribbon cable and interfacing with the touch screen were a burden, but I think with Klipper I don't have to change any wiring and the TFT I have looks to be a size match for the Creality screen.
I think this is cool and I've loved watching you do this, but I am far more interested in seeing what the best results are you can get out of a totally stock printer without upgrading any parts, other than the nozzle of course.
I love how happy you look in that thumbnail.
great job! I would love an extremely precise printing challenge!
Absolutely. How about flat vertical walls. That is a challenge for my printer.
But for precisin in fdm printing is so much more involved that u cant adjust , that it would make No scence
@@gingerflow5594 I’m not sure I follow. What do you mean by, so much more involvement that you can’t adjust?
@@orenaverbuck3845 Filament Diameter moisture Temperatur inconsistencis If its Spelled Like this 😂 and many more
@@gingerflow5594 but you can adjust these things with slicer tuning, smaller diameter nozzles, and filament dry boxes. It just seams like a more cost friendly competition, that’s all.
PID is not the best solution for controlling the hot end. It is however an easy solution which is why it is used. PID auto tune does not do a good job of setting the parameters which is probably part of the reason for the under extrusion at these speeds. The big problem with PID is that it is always reacting to what has already happened. It is always trying to play catchup. You extrude a bunch of material and the PID loop doesn't know about it until the thermistor (or thermocouple) sees the drop in temperature. The heater cartridge power is then turned up and hopefully it will eventually catch up. There is both a delay in sensing the temp and a delay in adding energy to the heater block. The temp sense delay is bad but the heater block delay helps keep PID from overshooting and burning the plastic. Hand tuning the PID values will help a lot as the auto tune doesn't generate values that work will near the limits. People will often turn up the extrusion temperature in an attempt to increase print speed but in the situation where you are already extruding at near the theoretical maximum this will work at most for a few seconds. And greatly increasing the acceleration and print speed does this.
We know that it takes a given amount of energy to raise the temperature of the plastic. This value will vary slightly with the type of plastic. For optimum control we need to know:
1) The amount of energy it takes to change the temperature of the plastic.
2) The amount of material being extruded.
3) The temperature of the filament. Knowing the room temperature would be good enough.
4) The desired temperature of the molten plastic.
5) The delay in sensing the nozzle temp.
6) The delay in changing the temp of the hot end.
7) The amount of energy lost to the environment. This is what is lost when heating the air around the hotend and a large part is due to part cooling fans.
1, 5, 6, and 7 could be figured out during the warm up and nozzle prime making it self tuning. It can also tell if the nozzle is blocked or extruder is skipping when temps don't change as expected.
One of these years someone will implement this in Marlin, Klipper, Smoothie, or some other software. Hopefully nobody has patented this and that would make this posting representative of prior art invalidating any future claims.
WOW I used to drag race my old 1964 Corvette so i know the problems with going fast. Boy was this a fun video. Burning rubber and blowing engine not really but blow the sides out a benchy because your cooling was too low relates. I can not wait to see what my pig of a bed slinger will get to in that contest they should have catagories, modified stock class you know. Take care and thank you again.
Hey man, love your videos! Could you make a strength test comparison of fiberglass reinforced resin vs PLA vs ABS and compared to other filaments in a strength test?
Nice video and explaining. Those core xy printers can go really fast.
Thanks for revealing the origin of the rattle. I thought it was the hotend.
P.S.: Interesting video. Compared to CNCKitchen it seemed to go very smooth using the Ratrig.
From memory he had issues with his machine distorting from the filament used to make it. Once Stefan fixes this, the smaller Voron 0 should really fly. I also greatly benefitted from the premade upgrade path of the EVA carriage. It was very easy to change cooling and hot ends.
Great video, really enjoyed the detailed breakdown you gave 👍🏻
Very very impressive
Hi Michael, I think I am about to embark on this journey to build this printer. I really want an enclosed 500mm^3 CoreXY printer. I currently have an Ender 5 Pro with a decent amount of mods so I feel confident in taking this on. I have been eyeing a Voron 2.4, V-Core 3 or Exocube (I have Exoslides on my E5Pro) but I think I have settled on the V-Core, I just love the 3 point kinematic bed. It's brilliant. Although I am very curious about the latest Exocube XYZ kit that uses 4 individual Z-axis motors with belts. The gantry moves in all 3 orthogonal directions while the bed is planted to the bottom of the frame. I have not seen many people post any reviews about this setup and am hoping the designer updates the blog section with this latest design. His last blog post was a version that used 4 Z-lead screws rather than belts. Would love to see you build the belt version and test it out!
Was hoping you (or anyone with experience/knowledge) could offer some suggestions or tips about the V-core. I see you have updated your hotend from your original choice but kept the same extruder. I see RatRig offers the Orbiter V1.5 or Bondtech LGX. Do you still find the Orbiter to be more than enough for your build? I like the idea of the lightweight design but the LGX just seems much more robust and I like the pre-tension lever design.
Another question is whether there are any major concerns to be weary of when building a large build volume FDM printer. Am I better off using linear rails for this setup or are V-wheels/Exoslides better for that large of extrusion bars? One other thing I've considered is buying something pre-built like a Tronxy X5SA-500 Pro and then gutting out the cheap parts with the upgraded mods that I want. But I am not sure I want to take a gamble on material quality of everything that I do decide to reuse.
Not sure if you will respond but I really appreciate your time and help. You have been an incredible resource for my modded E5Pro and I am sure many other people feel the same! Thanks again!
Another thing to increase speed is the stl mesh angle. While smaller angles make smoother finishes, they cause a lot of extra gcode in the final product.
That is a sensible challenge. I tweaked a lot of my setting before to get faster prints. Learning to know what the machine is capable of is so usefull, like learning to know when it starts to affect quality. I did not go that extreame though, due to wanting to get fast prototype prints but yet reasonable quality. You never know what your machine can/cannot do unless you try it out. A limiter for me was at some point default hard coded accelleration limits in the firmware. It took a while to realize those.
"You never know what your machine can/cannot do unless you try it out." Very well said!
I love the parallels to this and the automotive industry! I'm new here but based on your outfits I feel you see the parallels as well! If you're not into cars, I suggest you don't start, it's a far more expensive hobby with far greater disappointments hahah! You're very much aligned with the guys in the early 20th that were pushing race cars to their limits imho! Thanks for sharing things like this with us. I'm off to print something at 60mm/sec a lesser quality than your 250mm/sec print now :D
I'VE BEEN WAITING FOR THIS VIDEO!
You are the best! Awesome video!
Hey Michael. Not sure if you'll see this, but if you do, have you thought of making a video comparing the two corexy printer kits you've put together: the SecKit SK-Go and the RatRig VCore 3. I'm thinking of trying to save up for a printer kit now, and I'm leaning towards the RatRig, but I'm really interested in seeing how it compares to another kit. Especially going over things such as ease of assembly, documentation, calibration, and upgradability
excited hopefully to see if there is non planar 3d printing support for it
Michael - will you please give us your “summary” opinion of the RatRig now that you’ve taken us this far? If you can afford it - is this the best printer to buy for under 2K? Thanks for all you do - excellent and very helpful!
Baby steps!
~10 years ago on a Prusa MK1 i was already printing sub 20min benchies on that POS frame.
Tho 5mins is starting to get nice. I might get back into this racing -- tho for practical prints i've always thought 1.75mm filament is just too thin, but then again you want to melt it fast and ~10years back you just cranked the temp up ande hoped for minimal degradation.
Oh well, give and take :) :) Glad to see people finally getting interested on print speeds. Of to order my Voron or Rat Rig once i sell a few other printers off
Liar.
You can get rid of the bits of infill in S3D by switching to concentric external infill patterns and change the external wall type to perimeters only and as well the internal wall to "allow for single extrusion fill" it cuts a lot of waste and ends up printing similarly to vase mode specially if your reduce your over lap perimeter.
I had a great suggestion from my patron Patrick to run multiple processes and turn off solid layers for that portion.
Brilliant video. One beef: Many “improved” cooler designs completely ignore the physics of fluid mechanics. By blasting air from more than one direction, all you change is the stagnation point, and for “opposed” air injection (including “all around”), the nozzle is right in the middle of the stagnation point, ergo, no cooling at all! I therefore worry about “improved” shrouds that blow from three sides. I don’t think they would be any better that a shroud that blows from one side only. It’s a tricky problem that requires a solution backed up by physics.
One thing I would like to see attempted is to modify the Y-axis to remove the 30mm extrusion. I've seen printers forgo this support in favor of the tool steel linear rail which 𝘴𝘩𝘰𝘶𝘭𝘥 have enough rigidity to hold up a lightweight hot end. At the speeds you're attempting, every gram matters. Inertia is a bitch when acceleration is involved
i guess the best cooling wound be engineering an air conditioner unit on the side and add a new z rod with a blower to follow the nozzle, if i had the time and money i would so do that. liquid nitrogen cooling system wound work in theory but that's a little too risky
I really want to see an expert talking about and testing BerdAir part cooling. I believe it has great value for getting as much air as possible to the work without putting much weight on the tool head. Could you give your input on it?
VEZ3D, on youtube, is testing out something like this. He is also one of the faster speed boaters.
I'll add MirageC to that too. I'd like to try it at some stage. I have some 24V air solenoids ready and everything.
Great job on this video :) clear explanations and a nice view of the steps you took to reach your goal!!
Just an idea, I got while watching. How about using a compressed air for cooling?
It could be supplied by a tube/hose, it would remove the mass of the fan. As a bonus, the expanding air is colder. Compressors with reasonably big tanks are not very expensive and can be used for many other tasks.
That's what MirageC is using on his HeVORT. In terms of noise and room taken up there are some things to overcome but in terms of cooling power and being compact on the printer it looks amazing.
I would be interested in an Ender 3 or other cheap printers race. The community seems to have mastered print quality, how far can we push them
I got my modified Ender 3 to 16 minutes. I can barely get above 5k acceleration, but it’s still crazy fast to watch.
@@Crazyates11 what kind of mods are we talking about. I am genuinely interested.
@@cybkeats SKR RRF, micro Swiss all metal hotend, BMG clone direct drive, dual 5020 blowers, linear rails on all axis, dual Z with integrated lead screws. Upgraded the Y stepper to a bigger motor with a smaller pulley, so lots more torque. Adjustable metal belt tensioners. Sitting on a concrete paver on top of foam. Plus a million other small little things.
I own an anet a8 running klipper and my god let me tell you😲 300mm/s 2K accels on the daily with the same quality as an ender 3, and with a 1.3kg bed!!! klipper is a miracle of engineering, and we should all be pushing at least 200mm/s
people called me a madman for pushing 35m benches with 3 walls back in the day, now everyone is pushing faster and faster! this is the true spirit of speed, to push the limits far beyond reasonable, and end up with a stable, high quality, fast printer for daily driving
Like in squere or even in a cube. Great series. Cost is bating but the performance is insane.
4:14 turning the heat sinks 90* would imho improve the airflow through them
I haven't done any research to confirm if this has already been implemented, but as you can tilt the bed of your RatRig, would it be possible to program it to "lean in" against the direction the nozzle is travelling while printing, which would effectively reduce the distance the nozzle has to travel for any straight line? It would be sort of like adding a delta motion system to the bed to complement the core-xy of the 'main' printer.
I'm not sure how much of an effect this would have on print times for the complexity it would add, but I think it could be an avenue worth exploring.
great information. thanks for sharing.
1. J-Dampers + Tuned Mass Damper
2. Friction Springs
3. Polynorbornene Fittings
4. Inline Extrusion Diameter Sensor
5. Print directly out of drybox
6. Before actually printing your Benchy, try using Purge Filament and Cleaning Filament as well.
Also consider switching out your infill pattern for something much faster. (I wonder if Gyroid would be faster than Rectilinear...)
Me: oh nice and fast
My Ender 3 V2: don't go that rabbit hole...
In your last video you asked what people want - i didn't get the feeling that fast was it from the comments .
I'm not pretending fast was the only thing people asked for, but trust me, a lot of people wanted faster printing without compromising quality.
Heck, i thought it was just a race. But that does make a excellent point. If you can print decent extremely fast you can print good or great at a quick pace. Certainly food for thought.
Great job!
Do you know the reasoning for requiring the use of stepper motors in the speedboatrace rules? Are there other motor types that would allow you to get much faster prints?
The stepper rule was implemented because currently servo motors are still very expensive (compared to their stepper counter parts) and although some have started using them (awesome work Vez and Mirage!) there still isn't native support for them firmware wise and before Vez made his video on the subject, no real info on how to use them. It was done as a means to somewhat even the field (of course a 2k dollar printer and a 200 dollar printer are hard to match up, but this makes it atleast somewhat fair) and looking at some of the top entries the stepper printer times are really not that far off from the servo printer times... (nitram is actually at the top with 48v steppers outright beating the servo times all together, and the Annex Tycho is high up there still on 24v normal steppers!). There have been talks of doing a second contest/class in the future where servos would be allowed. For now the point was to try and push the printer to the limit with the current technology as far as I understood it :)
I would love one of these in the 500mm size over a new Prusa XL which i recently cancelled my day 1 pre order as i am simply sick of waiting. It seems RatRig do not ship to Australia these days and the cost of the printer if you can get it to Australia is over $3000 now. I would buy a Voron if i could get a 500mm size kit.
@20:20 fast, beautiful prints at 200mm/s...that's the best takeaway from this challenge!
Awesome video!
I understand that going this fast in the speedboattest will increase noise, but i am wondering with the improvements you did to bring down the print time with 59%, how lowed is the printer than? From stock and possibly compared to other devices?
Super interested in swapping out my Ender 5 but noice is a big "issue" at home so :)
Thanks
Nice one! Great video again. Thanks ;)
In the slicer, did you play around with infill every other layer? The rules allow for it explicitly, but I don’t see anyone talking about it. I’d assume it would only save time, but also assuming the hotend could handle the mm3/s of filament by doing double thick infill? Part cooling wouldn’t be an issue since it’s only infill…
I tried to that right at the end. Prusa Slicer said it would shave off a minute (off it's inaccurate 28 min prediction). I changed other things at the same time so I don't know exactly how much it saved in real life.
Great video!