DIY 3D Printer Build Experience (II) // Mechanics & EVA Printhead // RatRig V-Core-3
ฝัง
- เผยแพร่เมื่อ 29 พ.ค. 2024
- Learn about one of the most advanced 3D printers that can only be had by spending 40 hours building it yourself! The RatRig V-Core-3.1. Along the way I'll teach you about state-of-the-art 3D printer technology, and you can see the DIY printer experience for yourself!
Link to the entire build series playlist:
• DIY 3D Printer Build S...
RatRig V-Core-3 Home: v-core.ratrig.com/
==TIMELINE==
00:00 Introduction
00:59 Bed Arm Assemblies
01:55 RatRig Build Inspiration
03:03 Print Bed Installation
05:00 The Amazing EVA
07:10 Problems in EVA Town
09:24 Boss Level: EVA Upgrade
13:01 Mechical Assembly Tour
17:30 Closing Thoughts
#3dprinter #3dprinting #ratrig #vcore3 - วิทยาศาสตร์และเทคโนโลยี
The initial fan duct you had (the one with 40x40 mount) was intended for 4028 server fans not 4010 that you initially had. The 4028 offers a lot of power if needed, It's lighter than 2x5015 blowers and quiet as hell for "regular" printing (30% max at around 200mm/s, 7k accel for PLA). Might want to revisit this idea in the future. If so the general consensus is a Delta fan, model FFB0412UHN-SM36.
Yes! I had printed off the default EVA parts from the RR repo and that tiny 4010 fan was all I had that fit. That's why I ultimately changed it out. I am mostly printing ABS and other high temp materials (Nylon & PC) which require minimal cooling. I considered the dual 5015 fan adapter, but this has been more than sufficient so far. Even for a 8:35 benchy in ASA!
@@ObsessiveEngineering yep, not much cooling needed for ASA and PC :)
I like the blue mat you put the screws and part on, where can I find that?
Yeah that mat is awesome. I got it from (US) Amazon: amzn.to/3WBgIH9
Did you end up adding the second extruder ?
How did that work out, second extruder for soluble material support definitely worth adding.
I never did. This printer ended up being my workhorse, and I never wanted to put it out of commission long enough to try that. It also was kind of an experimental upgrade in the community.
At this point, if I'll probably just build a V-core-4 which has a native IDEX option. I plan to use it primarily for support materials, but also for duplicating/mirroring to print multiple parts at once.
Just a tip from what I've read. Is you should use abs asla
Yeah, they recommend PETG if you aren't planning to enclose the printer. Or ABS or ASA if you are. In this build everything except for the EVA printhead is printed in ASA (the teal/blue). Only the EVA parts are printed in the Nylon12(+CF?) that the Fuse 1 uses. I heard that nylon has creep issues especially in a heated enclosure, but the printhead is pretty low stress compared to the parts holding the bed or tensioned belts. I wasn't 100% confident at the time, but I've printed 250 hours on this thing so far and it's holding up great (and much of that has been at 60C/140F chamber temp!)
@@ObsessiveEngineering I understand, but if you look at alot of people who have done the setups, they say abs is the best by far, the other plastics tend to not be as good. And good luck with your build I hope to do a custom set up soon.
Yes, ABS/ASA are the universally recommended material if you can print it. It is rigid enough, has almost 100C heat tolerance, and super low creep even at elevated temperatures. (ABS and ASA are nearly identical in terms of print settings and mechanical properties). But you probably won't be able to print the RatRig parts in ABS/ASA without an enclosure.
If you don't plan to enclose your printer you can instead use PETG. In fact, if you buy the printed parts from RatRig instead of printing them yourself, they will send you parts printed in PETG.
You can buy ABS/ASA parts from community members in the unofficial Discord Server
@@heisenberg19x66 yes I know
You mentioned in this video and the EVA video that the paid parts had to be shrunk, but how did you figure out by how much?
Based on your question, I just had an extended conversation with folks in the RatRig Discord about this, and it seems that my conclusion may have been wrong. Or it was right but we don't know why. They claim the parts are designed for PETG but that PETG has virtually no shrinkage, so the model files are not actually oversized...?
My guess is that the absolute scaling of the parts is correct but that the holes have large tolerances to accommodate FDM print artifacts (like blobs at the start&stop of tool paths which can make the holes effectively smaller, which SLS and SLA printing wouldn't have).
My recommendation is just print the heatset insert EVA parts and enjoy your life, haha.
P.S. - For completeness: RR says the parts are designed to be printed in PETG, and that you should scale the models to 100.5% if you plan to print in ABS/ASA since it shrinks a little bit. I printed all the non-EVA parts in ASA and did the 100.5% thing and everything there was pretty smooth. There are a few captive nuts in those parts but they worked well enough.
Hey, this is a really interesting build!
Is it possible to add a heated bed to this? Cheers.
It has a heated bed!
It's funny I went over all the most advanced features of the printer and didn't cover the basics, whoops! Yes it has a heated bed, you can see me applying the 1,000W heater pad to the bed at 04:45. I also mention it in the first video in this RR build series playlist.
th-cam.com/play/PLrTmSWaWJwqLDVR0vk3-6i47G8yUVdCsp.html
Thanks for watching!
@@ObsessiveEngineering ohhh gotcha! Thanks.
I printed a set of ratrig parts in that same color. Then i decided on a different green. Is that overture teal asa?
It's Polymaker teal ASA. Though I've seen other brands with a very similar color.
@@ObsessiveEngineering my bad, you're right. That's the exact stuff i meant.
Actually CoreXY isn't the only way to have fixed motors (maybe even outside the enclosure)and a light XY-gantry. Ultimaker actualy uses the crossing rods design for a long time now. With crossing rods you use classic cartesian kinematic, but have two linear rods or rails left and right with belt driven cariages and a x-rod between them. Same front and back, with the y-rod between them. Imho that design have some advantages if you want a big size printer, as you have much less belt lenght, you might use (big) carbon tubes instead of rods and you can build the system staticly constrained (instead overconstrained). The only real disadvantage of a crossing rod design is you can not add a toolchanger in an easy way, as you have linear rods/rails on all four sides and the crossing rods - so it would conflict with toolchanging. But on the other hand crossing rods is muche more easy if it comes to IDEX, so if you only want a second toolhead for easy to remove supports, crossing rods in my opinion is the better design. But if you want more printheads (for example not only change material for supports, but also nozzle size), definitly the V-Core-3 or the Voron Trident etc. is the way to go. Of course, for pure multi-colour there is always the ERFC. But for high-flow hotends the filament tip after unloading is anything but perfect, even with finetuning ramping on unload. Actually an EVA with integrated filament cutter would be nice - the reliability of bambulabs AMS actually is caused mainly by the integrated filament cutter in the printhead, so the old filament in heatbreak and nozzle is kept there and only the unmolten filament with a fresh cut is unloaded. The old filament in heatbreak and nozzle is the reason you get so much poo on filament change on a bambu - but much less failed multicolour-prints (and it even works with materials like PETG, wich does a lot of stringing if unloaded without cutting).
For sizes up to 325by325mm² the Qidi X-3 series is definitly an ready2use alternative for the V-Core-3 (released after your video). Still, while I consider the Voron 2.4 obsolete now (exept for the fun of building it and the pride feeling if everythink works), the V-Core 3.1 definititly is the choise to make not only if to plan to add a toolchanger, but also if it comes to bigger size. Actually a ready2use Qidi X-Max not only weights 32,5kg, but the box barely fits through the door. So I belive there will never be a ready2use printer with 500by500mm² bed size for use at home, as it will be mostly impossible to get it inside the home...
For me my choice was clear: I ordered a X-Max 3. But I will build my very own printer. As I already have a Voron 2.4-inspired kit with >600by600 bed and mashined parts, It wouldnt be a full V-core 3.1, but I will redesign the kit to become a mix of Voron 2.4, Voron Trident and V-core. I definitly want the bed kinematic of the V-core 3.1 (as the fixed bed of a voron 2.4 wouldn't be a good idea if using a toolchanger). The new Qidi printers with activly heated chamber actually seem optimal to print strong parts for the DIY project (I do not have access to a SLS printer).
Actually esp. with bed sizes 400by400 or beyond, I would call a toolchanger a must2have. Using that area results in printing times to be measured in days and weeks, not in hours, if using a standard 0.4 mm nozzle. Still, you might want details you can not use a 0.8 or bigger nozzle for.
So changing nozzle size mid print can be a good idea (for example: printing a front plate for a 19 inch rack, you might want the first 3 layers with 0.4mm nozzle or even smaler and maybe even do multi colour to embed text and logos - but the remaining layers do not need a lot of details, so you might go to 0.8 or bigger to finish the print in acceptable time.
While I do not have an automatic toolchanger now, I did test this with a Revo-Nozzle, changing to a bigger one after 3 layers. Its actually a quite nice trick.)
Yeah I've looked closely at Ultimaker's design. I believe it's called "CrossXY". I never realized that it makes toolchanger designs difficult. I guess I don't think much about toolchangers, I assumed I'd do some kind of IDEX thing if I was going to do anything.
I've seen Qidi on Amazon, and a couple years ago I couldn't tell if it was legit. I've been gradually seeing more and more of it, so I assume it is now. The price point seems right and the actively heated chamber is a nice-to-have for engineering filaments.
@@ObsessiveEngineering Yes, the Qidi printers are quite interesting for years now. The quality is very convincing (esp. the bed - no "banana bed" like a certain other popular brand sometimes have). Even the reaction to the failiure with the x-3 series was very good - they stoped delivery and reworked the printer. Customers could choose between 50% back and keeping the printer (use it without active heating or tinker themself) or return it.
The new version of the x-series fixed the problem. My X-Max 3 will be delivered next week. Can't wait to get it.
Idex with Crossing Rods shouldn't be a problem. I am thinking to convert my Creality CR-200b just for fun (I barely use it, so it might be fun to tinker on it. But too many projects at the moment).
I thought a bit: Crossing Rods would make toolchanger more complicated, but not impossible. Actually there are also "Double crossing rod"-Designs, using two parallel rods in x and y. So the printhead actually is inside a sqare, bearings left and right and front and rear. So if constructing a toolchanger, it could be placed above the gantry, simply lifting the toolhead out of the sqare (like a crane) and inserting another tool. The locking mechanism seem even more simple (and light) - as gravity holds the tool in place (unless you want to use the printer on the ISS).
The reason I want a toolchanger simply is I am about to construct a big printer (600 by 600 by 600 mm²). I mostly want to print parts for retro computing (like authentic looking Amiga cases for emulators) and also some frontpanels for 19 inch racks. With embeding stuff like text and logos in the bottom layers, so a small (0.4 or less) nozzle size is needed for the details. On the other hand, after the first two or three layers, I want to change to a bigger nozzle to finish the print in reasonable time...
On such big printers CoreXY can be quite tricky. The different materials (aluminium, steel, belts) have different thermal expension. Any slight disalignment because of different expansion can cause layershifts. Also on CoreXY the belts get insanly long.
Actually crossing rods can solve some of the problems, as you can build it more staticly constrained. Simply clamp the crossing rods only on one side to the slider, but use a bearing on the other side, so the crossing rods can expand and contract freely. Also you could place motors for x and y and linear rails outside the enclosure to minimize thermal effects there (of course the enclosure needs openings for the crossing rods themself, so some heat escapes, but maybe theese parts you see on the bottom of doors sometimes can reduce that...
So maybe I give it a try and build my big printer with double crosing rods...
Some insulation under the bed?? Just a silicone pad alone?
I have thought about adding some insulation under the bed. It would reduce heating time for the bed, but would actually increase heating time and max temp of the chamber, since the bed would shed less heat into the air. It's kind of a weird decision/trade-off, but my chamber heater is not great, and recently I haven't been using it at all and just letting the bed heat the chamber. That said, the chamber still takes a lot longer to reach equilibrium than the bed. But for a long ABS/ASA/Nylon print at 110C bed, the chamber will get up to 60C within 30-45min which seems to work out well even if it's not at full temp when the print starts.
If you have a separate chamber heater or your printer just isn't enclosed, then insulation underneath the bed should really improve speed and efficiency
Disculpe mi ignorancia, ¿que caracteristicas tiene esta impresora?,¿qué materiales puede imprimir?,¿cual es el area maxima de impresión?.gracias.
Hi! You should check out the first video in this series where I explain all of that:
th-cam.com/play/PLrTmSWaWJwqLDVR0vk3-6i47G8yUVdCsp.html
The build volume is 400x400x400mm, and it can print any consumer 3D printing material, as long as you have the enclosure: PLA, PETG and TPU do not require an enclosure, but ABS, ASA, nylon, polycarbonate, polypropylene all do, even better if it's actively heated. If you want to print carbon-fiber or glass-fiber filaments, just make sure you buy a hardened nozzle.
@@ObsessiveEngineering muchas gracias, que Dios te bendiga.
A few questions:
* Is that a 500mm RatRig?
* Why would the X axis have such a large resonance compared to Y? There's a lot more mass moving in Y, I'd expect a lower frequency but large resonance there.
If the accelerometer is a single-axis accelerometer, that might explain why it does not see much resonance in X when Y moves -- it would just be measuring cross-coupling.
Final note, any chance you'd be willing to put that accelerometer on the frame itself (top side, where the X axis meets the slider when Y is centered), and rerun the test? I suspect your frame is resonating as well, if that's a 500mm RatRig.
If it is, you could brace the frame with 8020 diagonals, which should make it much stiffer and drive the resonance up past 200 Hz where the amplitudes are tiny.
It's a 400mm V-Core-3. And you just touched on a detail I left out of the video -- the clip and the chart are only for X-axis shaking test. The chart shows the output of all 3 axes of the accelerometer, but the print head was only moving in X, thus you see mostly X-vibrations.
If I had continued the clip you would see the Y-axis shaking test next, and a separate chart with the results from that showing vibrations mostly in the Y-axis (actually, labeled Z because of the accelerometer orientation). And yes, Y-axis is much more mass so it has a lower frequency. Resonance for my X-axis is around 72 Hz, Y-axis is around 28 Hz. Magnitudes are around the same order of magnitude.
I recently attempted a speed benchy with input shaping disabled, and I was astonished how terrible my machine sounded. With input shaping re-added it's not a quiet machine, but it sounded reasonable at high speed, not like it was going to shake itself apart. I did eventually succeed at an 8:35 benchy, and most prints have minimal ringing, so I don't plan any more updates to the printer (for now)
P.S. - Sadly something has gone wrong with the accelerometer (or the wire/connector), and I haven't been able to run it recently. However, I noticed that RatRig posted default resonances to use for each machine size for people that don't have the accelerometer. Apparently the frequencies are pretty similar between builds.
@@ObsessiveEngineering I'm thinking of building a 500mm RatRig to print some molds I need. I was planning to buy ABS/ASA printed parts off Etsy, and enclose the machine, mostly to keep the temperatures and thus shrinkage consistent rather than to print nylon or other high-temp material. Any warnings you'd like to give me?
@@IainMcClatchie If you need accuracy, Nylon & PC are very difficult to print unless you have a crazy hot enclosure, especially larger parts. Getting the chamber hot enough for that could require different rails, belts, bearings, etc. However, if you are only planning ABS/ASA, then you should be good with the default enclosed V-core-3. I have 3mm polycarbonate panels for most of the machine, and it gets up to ~55C (eventually) from pre-heating the bed to 110C (chamber to 60C with insulating foam on top). This is good enough to do even large ABS and ASA parts very reliable. I added a 200W heater to help speed up chamber heating, but it only made it like 15 min instead of 30 min. Beneficial but may not be worth the trouble:
- www.printables.com/model/286458-ratrig-chamber-temp-controller-and-rocker-switch-m
The old style enclosure is probably not an option anymore, but be sure you're doing Enclosure 2.0 based on the original "open-front mod". The old-style enclosure before late 2022 (with the hinged lid) technically works but is quirky and obnoxious.
The only other thing to be aware of is that doing the enclosure is not cheap. You'll have to either pay a company to source and cut the panels for you (which can be like $500), or buy the polycarbonate panels yourself and get access to a CNC router to get them cut. It was an excuse for me to learn to use a CNC router, but not everyone is that committed 😅. If you want to go that route and are in the US, you can keep an eye on Piedmont Plastics clearance page, which is where I got them like 70% off, though they weren't all the right size and I had to do some panels in halves.
Let me know if you have any other questions.
@@ObsessiveEngineering If I'm not concerned about looks, why wouldn't I cut polycarbonate side panels from 8mm dual-wall polycarbonate (typically used in greenhouses) using a handheld jigsaw, and then hand drill the mount holes? Does accuracy matter here? I don't even care if it bows a bit when it heats up, right?
I want to print whatever material gives best accuracy. There will not be much mechanical stress on the mold, and I plan to epoxy fill and polish the surface, so I'll be using the largest nozzle, flow rates, and layer heights that work.
The stiffness requirements on these rigs, especially on the belts from motion in the Y dimension, are large. I've seen acceleration numbers like 50,000 mm/s^2 thrown around. That's 5 G acceleration: +/-2.6mm of amplitude at 22 Hz. That Y dimension appears to be at least one kg, and if it were just one kg 50 m/s^2 acceleration is 50 N of force. The belts on a 500 mm machine are supposedly 2.9 m long loops (I think a little smaller). The short legs would take about 20 N of that force each, across a belt somewhere around 600 to 900 mm long.
I looked on the Gates website for a stiffness spec, and did not find it. I did find this paper: Nominal Stiffness of GT-2 Rubber-Fiberglass Timing Belts for Dynamic System Modeling and Design (www.mdpi.com/2218-6581/7/4/75). With a little extrapolation, I think it suggests a 900 mm long, 9 mm wide fiberglass GT2 belt with 20 N of force change would have a length change of 1.6 mm.
That is to say, the drive pulleys would have to oscillate 2.6+1.6=4.2 mm at the pitch line to get the Y axis to oscillate 2.6 mm. That's a very large amount of error. I can see why folks would tend to stay away from 50,000 mm/s^2 accelerations.
Kevlar belts should be about 3x stiffer than fiberglass belts. Do the RatRig kits come with kevlar belts already? Is there any chance the belts could be made any wider? I understand the RatRigs are using custom pulleys, which suggests no.
If you have any spare 9mm wide GT2 belt lying around, I'd be quite curious if adding 5 kg of weight to one causes it to stretch 0.18%. That seems like a pretty easy thing to make into a video: compare Gates belts to whatever RatRig ships, measure if the belt is stiffer, plop it in the rig to see if the X and Y axes get stiffer, see if you can accelerate faster with the same ringing artifacts.
@@IainMcClatchie So I can't respond to everything here in a TH-cam comment, but I have a few high-value responses:
1. 50k accel is absurd, and may not be realistic with a 500mm V-Core-3. At 20k accel I did an 8m35s benchy, and was worried about premature wear on all the components. Most people (myself included) run 4k-10k accel for every day use.
2. If you are aiming for biggest nozzle and layer heights, then the max speed and acceleration is not going to matter much *because you will be limited by the max flow rate of the hotend, not the speed of the printhead*. As a point reference, I just put a huge 1.0mm nozzle on mine, and used a line width of 0.6mm and layer height of 1.3mm. Even going as "slow" as 100mm/s results in ~80 mm^3/s flow rate which is higher than most nozzles can handle. I try to keep mine below 35-40 mm^3/s even with my Phaetus "ultra-high-flow" (UHF) hotend which is one of the highest-flow-rate hotends out there. They claim 75 mm^3/s though most think that number is a bit optimistic. I've seen a couple hotends recently that claim to go up to 115 mm^3/s (like the super-volcano), which is great but you'll probably need a custom EVA mount for it.
3. Most people go with 3mm or 6mm panels though the rear electronics panel must be 4mm or less. Honestly I don't think the thickness will matter that much unless you're trying to push the chamber higher than 60-70C but that would require a lot of other upgrades (high temp belts, bearings, water cooling the printhead). There is a "#high-temp" channel in the RatRig discord where people will give you advice if that's what you're aiming for. If you're just planning to print ABS/ASA, then I can assure you that 55C-60C chamber is good enough and you'll have no problem with 3-4mm panels.
Finally, I encourage you to join the RatRig discord, there's a lot of dedicated channels there for all these kinds of topics. For instance, there's a guy in #high-temp claiming to be working on a printer that can get up to 220C chamber! There's also tons of mods discussion where you will find people that have redesigned parts to accommodate different belts, rails, etc.
I would have gotten rid of those printed parts that didn't work, its only going to frustrate you every time you work on that printer.