Hemera Direct-Drive Extruder Carriage WIP

Hey all. I’ve been away from this community for a while. I was an early Eustathios builder back in the Google+ days, but since that all disappeared, I ended up kinda drifting away.

In any case, I’ve been getting a bit sick with the inconsistent bowden extrusion on my printers (When I print 4 perimeters, the inner perimeters under extrude while the outer ones over extrude) so I figured it would be time to start thinking about giving a direct-drive extruder a shot on one of my printers. A brief look through these forums seemed to indicate there wasn’t an available direct drive mod ready-to-go, so I figured I’d have a go at designing my own around the E3D Hemera. @Maxime_Favre seemed to have the beginnings of something similar going, but I don’t think he released anything.

I’ve managed to design a carriage that uses 5 of the 8 mounting points on the Hemera (data sheet says only two are required, though that seems low to me) and fits a 40x40x20 blower. Much of the cooling duct geometry is shamelessly stolen borrowed from the v2.5 carriage.

Some dimensions:

  • Maximum extents in the X direction is 78.5mm
  • Maximum extents in the Y direction is 88mm (excluding the press-fit bearings)
  • The nozzle tip is 55.2 mm below the centerline of the lower cross-rod.

Some pics of the CAD are shown below. This represents about 2 nights of work.

I’d love to get some feedback on possible improvements. I’m sure there are ways to optimize the design, but I think I’ve knocked out all the obvious (to me) stuff.

6 Likes

Welcome to Maker Forums! Glad you found your way here. :slight_smile:

4 Likes

It looks great! I can’t tell from the images, so please pardon the dumb question: can this be printed without support material?

1 Like

The main carriage may need a little support to help bridge the gap that I had to make around the motor connector. Otherwise, it should be pretty good. Printability was definitely a requirement of mine when designing.

The cooling duct will definitely require some support. As I mentioned, it’s based on the existing v2.5 design found on Github, so it should be just as easy or difficult to print as that.

5 Likes

That design looks awesome. Sorry I didn’t reply earlier, just playing catch up while I do a bunch of work travel.

2 Likes

Thanks! I’m going to have to make a couple other mods. The new V2.5 “belt tensioners” (side-rod carriages) will work for one of the cross-rods, but not the other, as it will shrink the Y-axis more than necessary. I’ll have to use the old design, or make one that biases the cross-rod towards the rear of the printer.

Would you be able to give some insight into the design changes for the belt-tensioners/side carriages? The more compact bent tensioner mechanism is nice, but I’m trying to see the benefit of the other changes.

Based on a symmetric central carriage (v2.5), there was room to space the bearings apart on the side carriages without lowering the total travel. In fact the more compact newer carriage (v2.5) actually increased the printable area. But if you have an asymmetric rod on the center carriage because of the body of the hermera then yeah, the new side carriages would limit your travel. I think biasing the hole one way on the side carriage would be a solution for your new design. The further you get those bushings apart from each other the easier it is for alignment and the more stable things are.

The side care design was actually stolen/modified from an update that Zane did on the HercuLien and I liked the way that they looked/worked. Hence the integration into the latest version Eustathios.

Yeah, I noticed the V2.5 changes result in a larger build area. The new carriage seems to lower the hotend, but I guess the new bed design still gives larger vertical build height? It’s interesting. I’m not prepared to re-do the bed. I’ve never come close to using the full build height of the printer. I’ll probably leave one printer a bowden version, and the other the direct drive extruder (Maybe call it “Eustathios Direct”?) so that I can still have one with the slightly lager build envelope if necessary.

Thinking through it, I can see how more widely spaced bearings helps a bit with alignment. I may indeed have to switch to this, albeit with a modified design for this carriage. I’m also still not totally sold on using set screws to hold the rods.

1 Like

The features of the carriage have been dressed up a bit and fasteners added:


I found some time to figure out the gantry. I wasn’t totally in-love with the existing v2.5 side carriages, so I remodeled them to look a bit more “clean” and, more importantly, to be more easily printed. I did end up sticking with the set-screw design since I couldn’t seem to make a clamping design I liked.


Once I was happy with that, I made offset versions that would attempt the maximize the print area of the new carriage.


The new side carriages do require a much longer screw for belt tightening than the original v2.5 design since I didn’t want to tighten against a heat-set insert. Getting that heat-set insert in that deep pocket would also be quite difficult.

Overall, the gantry looks like this:


I get a theoretical maximum printable area of 289.5mm in X by 279 mm in Y. This may have to be reduced somewhat in X to find room for the end stop switch, and somewhat in Y so I don’t pinch wires going to the extruder on the back of the carriage.

Overall, I’m fairly happy with it. I’m not in a position to build this myself quite yet, but I’m more than happy to release my files in case someone else wants to give it a go. Please let me know if that’s the case.

4 Likes

The models look great. Wonderful work. Once you get things tested out would you mind submitting the changes via a pull request to the GitHub? I want to add a community mods folder to v2.5 like I had for V2.

Thanks for sharing. Very clean work.

2 Likes

I’m not going to have a chance to build and test this thing until late May or June at the earliest. I’d be happy to put in a pull request with the caveat that the parts are untested and represent an alpha or beta release. Would that be okay with you?

I also need to confirm how it works with the print bed. My old Onshape model has decayed, so I’m re-doing everything in Fusion360 and I haven’t yet CADed everything up. It seems the newer bed model relies on the nozzle being a bit lower, so I have to work out the heights. My new carriage has a similar nozzle length to the V2 carriage.

2 Likes

Since I will still need something on the fixed side to guide the filament, I figured it might be good to make a filament run-out sensor. It’s pretty simple, though maybe a bit over engineered. It is to work with Bowden couplings and of course PTFE tubing, which isn’t pictured. I think it’s pretty obvious where they fit, however.



There are provisions for strain relief for the switch wires, as well as for the wire harness going to the carriage and hot end.

There’s no reason this couldn’t be used in combination with the existing Bowden design, though I think a mirrored version would work better in that case.

3 Likes