Ender 5 Plus Dual Extruder (Part 3 — In failure…)

In failure

Ok… so, Chimera+ was a total bust.

I purchased the E3D Chimera+, the Bondtech BMG-X2 and I printed the mount, installed everything, got it all … just right… and frankly, I think the Chimera+ is a horrible piece of junk.

Before you think I don’t like E3D, let me say that the build quality was amazing, the heating elements were 100 times better than anything I’ve seen from any other vendor, the thermistors were nothing short of a piece of art, the heat blocks were a thing of beauty … the rest of the Chimera+ was utter rubbish and I hope E3D stops selling them.

The biggest problem with the Chimera+ was that it requires a piece of PTFE to run all the way through the heat sink, the heat break and down to the nozzle. I tried multiple different types of PTFE including Capricorn XS and somehow, the tubing kept deforming and jamming the whole mechanism.

I’ll stop here and suggest that it’s possible the Bondtech just has too much torque and causes too much internal pressure, but the simple truth is, this should never happen if the Chimera+ was designed properly. And it simply is a horrible design.

I have purchased from a model shop, a meter of 2mm inner diameter, 4mm outer diameter carbon tubing. I supposed I can give it one more try, but I’m simply no longer interested.

Bondtech BMG-X2

The Bondtech BMG-X2 is extremely well made. It has some serious issues which aren’t easily resolved.

First off, it weighs A LOT. I mean, I’m not sure how much we’re talking about, but with everything mounted, but not installed on the Creality X carriage, the full assembly probably weighs over a kilogram.

This is a real problem. It means no rapid printing… which is fine, but it also means that one a single piece of 20x20 V-Slot, there’s no hope for stability at all. It just wiggles and jiggles.

A hack?

I considered replacing the 20x20 V-Slot with something more sturdy. I have access to a machine shop where I could create a lovely mount to hold linear rails based on linear bearing and smooth rods. This would be a great solution… but if I do the X, then I also want to do the Y. If I do the X and the Y, then I would replace the entire V-Slot system with CNC’s 18mm plywood. If I do this, then I may as well just throw the printer away and design my own. And then I’d get my dream printer which for the Z-axis would incorporate 4 separate NEMA-17 steppers that, during auto-level would actually mechanically level everything flawlessly. I’d also include close loop drivers for the motors to detect slippage when making heavy prints and possibly a gear box to increase resolution.

I may yet still do this.

Plan accelerated

So, I’ve concluded that in order move forward with my plans, I need to do it myself.

All metal hot ends

Anyone who has ever used a Micro Swiss all metal hot end know that they’re basically the best thing ever to happen to Creality printers.

The design of the Micro Swiss hot end is basically an almost exact copy of the Creality stock hot end with the exception that it is all metal from end to end. It’s extremely well made and it’s a nearly perfect solution… except it really isn’t designed for what I’m planning.

Sacrifices

At this step of the plan, the biggest sacrifice I’m making for the moment is to go Bowden again. This is a miserable thing because ever Capricorn tubing leaves you at the mercy of long distances of tubing after the extruder. This means that flexible materials will always come out poorly and retracting will always need to be longer. I’m also a little concerned that by doing this, if filament gets jammed, there won’t be enough torque to extract.

Nozzle distance

Yeh… the old Chimera+ was 18mm between nozzles. The newer model was 20mm between nozzles. Because the Micro Swiss hot ends aren’t themselves threaded, I couldn’t make the design less than 22mm between nozzles.

The whole shebang!

I decided that if I’m going to design my own dual nozzle system, I would have to “do it right”. After all, if I’m going to be stuck with Bowden for now, let’s at least have the decency to have raised nozzles.

Raising the nozzles.

Lucky for me, my daughter started working on a school project where she would need some SG-90 servos. Of course, the only local store to sell them only carried bags of 4 (at ridiculous prices), but all said and done, I gave her the one she needed and I used two of the other ones to raise and lower the nozzles.

I had considered a single static nozzle and one in motion, but like most servos that aren’t modified for continuous motion, these servos only rotate 180 degrees. I could easily modify the servos for continuous operation, but then you lack control of position. In the design I’m proposing, by using two of these super light and super cheap servos, I should be able to raise and lower each nozzle approximately 0.175mm. As such, when one is fully raised and the other is fully lowered, there should be about 0.35mm difference in their heights.

In hind sight, this is not very good, I’d be far more comfortable with 1mm, but I think this will be fine… so long as the M4 nuts and threaded screws I use are precise enough.

Aliexpress maybe?

Ok, so if my hacked up servo and M4 threaded rod solution fails, I can order some steppers and proper lead screws from AliExpress. They range in price from about $2 to $6 each and that’s certainly reasonable. Sadly, thanks to recent changes to the Norwegian customs system, I believe I’ll have to pay an additional tax of $20 to receive the $2 package. :(

Heat sink cooling

So, two nozzles will certainly get quite warm. Something I learned from having two heat blocks too close together on the Chimera+ was that even when one is entirely turned off and there is even a heat sock installed on both nozzles, if the hot … hot end is 200C the cool hot end will warm up to 75C. So, the 22mm distance might not be the worse thing.

I’ve decided to use 30mm fans, one configured to push air, the other configured to pull it. I am considering changing fan direction when changing nozzles. I’ve pretty much convinced myself I’ll need to make my own driver board for the servos and if I’m making my own board, I might as well add thermistors and fan controls. This way, the fans will only run on high when it’s actually meaningful to do so.

Part cooling

I’m planning on printing mostly in Polymaker CoPa which doesn’t require cooling. However PolyDisolve suggests the fans are one. So in version 1, I’ve decided to make the assembly much lighter by getting rid of the two enormous 5015 blowers and simply using a Creality 4010 blower.

I know the 4010 catches a lot of flack and in a dual extruder design, people love using dual fans. But I’ve had many great results with the 4010 and thanks to the multiple Creality printers I’ve upgraded over the years, I happen to have a few of them laying around.

Statically mounted BL-Touch

I’ll never understand why everyone always makes sliding BL-Touch mounts… even worst, every BL-Touch mount I’ve ever encountered other than the statically mounted Creality metal bracket breaks all the time. As such, I’ve decided the best option is to just mount it 3mm above where I hope the nozzle will be.

Preparing to print

It’s a long print, at 0.05mm per layer and with supports I apparently will need 13 hours to print. It’s also going to be a somewhat nightmarish cleaning job… but SLA prints generally always are.

Let’s hope!!!