Ender 5 Plus Dual Extruder (Part 4 — IDEX…)
So, I managed to buy a Raise3D E2 which is an insanely great printer. It’s also very very very expensive at least on a home user scale, but it’s really a great printer and I’m working through the “how do I make it print perfect” steps, which I’ll share with everyone when the time comes.
This is great news because it means I can break my Ender 5 Plus again… I mean upgrade it. The E5P has been my primary printer for a few months now. I’ve made some very interesting upgrades to it which I’ll share when I get to it. But let’s just say, Creality themselves would probably not be able to identify the E5P anymore. I think the key difference that would confuse Creality is that mine works.
Raised nozzle isn’t going to work
I’ve drawn maybe 10–15 different raised nozzle designs and each one gets better and better, but the raised nozzle simply isn’t going to work on the Ender 5 Plus. I think could make it work if I really reinvented the printer, but in that case, I’d rather just toss it out and buy one of the new configurable printers and do a tool changer design instead.
Single nozzle is a terrible idea
I’ve tried single nozzle multi-material extrusion on the Prusa i3 MMU2S and… frankly, Prusa should just stop selling it. Everything they make is amazing… except that… which is pretty junky. The mechanism is really cool, but the material waste is enormous.
Mosaic Palette 3 Pro
Yes please… I mean no… maybe? Well, I’m going to be testing a Palette 3 Pro as soon as it finally arrives. But I’ve been watching Joel Telling’s videos on it and other than support for a LOT of filaments, the results I’m seeing just aren’t good enough. I think when you’re using 8 types of PLA, it’s beautiful… except for the psychotic purge tower. If I were to consider it, I’d have to add a purge bin to the side of the printer rather than losing all my real estate. And then, I’m not really happy.
As for changing materials… I think that the Palette 3 Pro is not the perfect solution for this. I won’t go into details, but I’ve had too many problems over the years with switching different filaments back and forth and ended up hating life while cold pulling and removing nozzles and such. It’s just not a great solution.
Tool Changer?
I think by the time I added all the parts and the circuity to the Ender 5 Plus to make it a tool changer, I probably could have just bought the E3D tool changer kit. At the moment, I really don’t like E3D as you’ve seen from my previous articles. But if I ever need 3 filaments, I’m almost 100% sure it’ll be time for a tool changer.
Independent dual extrusion (IDEX)!!!
So, thanks to the Raise3D E2 which is the first independent dual extrusion printer I’ve ever seen that genuinely works, I’ve decided that IDEX is not only possible, but a brilliant idea. In fact, I’m even mentally experimenting with making a 3rd nozzle possible.
IDEX must be skinny to work
In order to fit multiple carriages on the x-axis, especially on the Ender 5 Plus which really just isn’t made for two x carriages.
This really limits my off the shelf options.
This is a well known, high end assembly and if it’s printed in a good material instead of the crappy filament they use for it, it’s a great option.
This is currently my #2 choice as I think it will cost me a lot of money in parts, shipping, taxes, etc… to make happen
The BondTech solution with a mosquito hot end is a great option and is actually what I’m designing for at the moment.
The only drawback to this design is that it doesn’t come in a mirrored version. Also, to keep the mounts on the V-Slot narrow enough, it’ll be very difficult to us cut aluminum for the mounts.
In fact, it’s possible I could end up resorting to the DDX v2 since it the fan bracket mount on the DDX v3 is a bulky kind of monster.
The stock V-Slot is not good enough
So, there’s just no way to mount a two transmission systems on V-Slot 2020. Lucky for me, a local shop happened to be selling a V-Slot 2040 1 meter piece I could use for this.
So, to fit two belts, I’d need to get it and I rushed to the store and bought it!
This is a peek at what it would look like when assembled. I’m making use of standard V-Slot wheels and a new back plate. I’ll have to run a few iterations to make sure that the slots for the belt are good enough. But what I like best about this is that the wider V-Slot will make the carriage far more stable.
Don’t reinvent the whole printer!!!
It was pretty important to me that I could avoid rebuilding all the parts of the printer. It may seem as though I’m doing that anyway… but if I do too much metal work, I’ll need to either pay for it or beg some machinist friends of mine to help and so far, this whole project doesn’t require any new metal parts except the V-Slot
Did you say no new metal parts?
Yep, thanks to FormLabs convincing me to buy some Rigid 10K Resin which I’ve just completed my very first test print on, I believe whole heartedly that this entire design could be 3D printed on a Form 2 printer.
FormLabs Rigid 10K resin costs about $600 in Norway to get started with (assuming you already have the printer, curing and washing sorted). And it’s not for the faint of heart. But Rigid 10K Resin seems to be perfectly suited for a project like this. All the carriages and plates will be printed with it.
Where you get these parts?
When I publish the design, I’d recommend using a cheap SLS printing service (there are actually cheap SLS printing services out there now) and you can print the parts like that… or you can take a stab at using rigid resins for cheap resin printers… I seriously think ordering is smarter though.
Alternatively, I may… to recover some of my investment setup an Etsy store or something like that. We’ll see.
Above is my first print with FormLabs Rigid 10K. I’ll go into a lot more detail in a video series dedicated to the material. I was most interested in testing :
- how it would print high detail
- how rigid it would be
- could I cure it without an actual Form Cure (I did)
- what kind of marks would be left by removing supports
- what tolerances could I expect
In short, the model above is hollowed to have only 1.5mm walls and the material was amazing. I will say that the supports were rigid to the point of fragile, but I believe when printing 4mm thick plates, it will be impeccable and I also think I can print almost entirely without supports for this project.
A clean replacement
I hate doing stupid things like breaking something that already works and risking parts while I experiment. So, I’ve specifically designed this so far to be a direct replacement for the entire X carriage on the printer. It really should require anything more than unscrewing the existing axis and then screwing in a new one. 4 screws… that’s it. If something goes wrong, I can do those 4 screws again and go back to what I already have which is working REALLY REALLY well.
The key drawbacks to this design
Bondtech vs. Prusa
First of all, the main reason I’m interested in using Bondtech is availability of parts. I live in Norway and shipping a lot of little packages can be very expensive here. I found almost all the parts I need for the Prusa extruder either on AliExpress or direct from Prusa. Either way, I found the prices to be quite a lot, so I decided that I’d rather use Bondtech if only because they’re a Swedish company (in fact, I think all the mails they sent me on past orders were in Swedish because of my location). It means that if I need something, I can order using the cheap postal service and receive packages in a day or two from them.
If all goes well with my Bondtech experience, I’ll pop over to visit them if they’ll have me at some point.
But, the drawback is of course cost.
A full DDX v2 with Mosquito is something like $150-$200. And to make this, I’ll need two of them.
Controller
I’m currently using a BigTreeTech SKR v1.3 with TMC2209 drivers. There is one more driver available to me to complete this design, but to build this, I’d need another. This is going to be a problem.
Controller Option 1
I can split a single driver. Since I don’t need to operate both X steppers simultaneously, I can actually just wire up a relay board using 4PDT relay (or two DPDT relays) to switch between steppers. This is probably in a way the nicest solution since there’s no real reason I need to have a full extra driver.
Controller Option 2
I can replace the BigTreeTech SKR v1.3 with a BigTreeTech Octopus v1.1 which is one of the most impressive looking controllers ever made. It’s really a beast with everything on it. I’m thinking this is likely how I’ll end up going but only after trying option 3.
Controller Option 3
Use a second controller for the extruders.
This is a real option where I can compile and run the latest Klipper firmware and control two controllers from a single Raspberry Pi. Since the extruders are generally quite benign in their control pattern, a stock Creality Ender 3 motherboard would do just fine here. I’m not worried about noise because the steppers on the extruder don’t really profit from microstepping or from noise suppression.
Why isn’t it done yet?
Well, simple… I just got the new Raise3D yesterday morning and I’m playing with that. Also, while trying to find tolerances to print for, it takes a lot of time to experiment with the FormLabs Rigid 10K resin. For example, I need to know how big the screw holes should be for mounting the motors and the extruders.
Also… I have probably the best dual extruder printer on the market already. I’m honestly only doing the project on the Ender 5 Plus because I feel like it. I don’t actually need it anymore. That said, I’ve already started spending money on this next step because a LOT of people have asked me either publicly or privately to finish this project and share how I did it.
I just spent like $4500 on printers, filament and resin. It’ll take me a little while to recover enough to write this. I’ve sent a mail to Bondtech asking if they’d sponsor the project in hopes that they would. After all, as I said, I don’t really need it and if only 1% of the people who read the full contents of this article decided to spend some money on it, that’s big bucks for them. I’m pretty sure that I’ll end up spending money and find myself with a printer that doesn’t work quite right for a month or three. So, other than the approximate $5 (total) I’ve made writing on Medium, I have nothing to gain here except satisfaction from a job well done.
Where will I go from here?
I will work a bit more on the CAD model for the DDX v2. I’ll also consider making a model using the parts I gut from the Bondtech BMGx2 I have no use for from the previous posts.
I don’t know what I’ll do for heat sinks and heating blocks if I make my own design. I’m very likely just to borrow a small 5-axis mill from the office and machine my own since I have some spare stock laying around. Alternatively, I might see if I can buy the cheapest E3D v6 clones ever made. The hot end really isn’t the issue here… so long as I have two identical units.
I won’t use the Microswiss blocks because I have concluded that they are pretty useless for direct drive. Sure, I could make my own adapter for them, but overall, I think I’d be better off with a V6 or similar in this case.
Well, wish me luck. This is my last attempt to make this happen. If I don’t manage to pull this off, I’m just reverting the printer back to how it is now and forgetting it.
P.S. — I’m also very much considering yoinking out the Meanwell 500w power supply which is REALLY noisy and replacing it with a 150w fanless power supply for the motors and extruders and then either a 350w silent power supply (if I can find one) or finding an AC bed replacement. I’ve already made it so when the printer isn’t printing, it shuts off the power supply by using the Meanwell remote control connection. I installed an extra Meanwell 5v 50w power supply for the Raspberry Pi and the Raspberry Pi touch screen. I’ll share this all when I’m done with this project.
