The “Prusaic” extruder – X5S goes direct

As mentioned in the Upgrade List and the Extruder Shootout post, I had seen those “Cloned Btech Dual Gears DIY” gears on Aliexpress and ordered. I always thought that Bondtech’s dual, counter-rotating drive is superior to anything else, but spending 60,- € for a set of genuine gears was just beyond all question. The chinese store pointed to a Thingiverse design – and there are more of them online. But I had the old E3Dv5 Hot End lying around and that combination (Bondtech gears and E3D J-Head) was just what the Prusa i3 Mk3 uses. Time to get rid of the long Bowden tube…

Prusa Research stick to their Open Hardware credo (there is a tattoo on Jo’s arm after all) and doing so properly means (at least IMHO) not using Makerbot/Stratasys Thingiverse. So if you’re one of the lazy guys only searching there, you’re missing their STLs. πŸ˜›
Their extruder design is direct/direct: gearless and non-bowden. And you can bet they got themselves FAE support. (Their forum suggests, that at least their stepper drivers were tuned properly by Trinamic, so it is highly likely that Bondtech at least reviewed their design.)

Bondtech reacted to the clones by lowering the price, BTW. 30,- € is still a lot IMHO, but now we could actually discuss the advantages of the “properly made” gears. And you also get the versions for stepper motors with metal gearboxes (bigger shaft diameters) there.
In the meantime some chinese manufacturers have startet to clone the complete BMG extruder with it’s additional 3:1 gearing. So maybe the way I went is not ideal to anyone anymore. Anyway…

To be honest: I tried to design something myself first (and didn’t get that right at the first attempt), before I read about “some improvements” the Prusa guys had implemented. My workmate had his MMU incoming (it arrived this week, BTW) and the blog post about it’s release also pointed the R3 / B7 extruder parts. The superb documentation on how to upgrade and the forum thread on what it’s all about had me say “Screw that!” to my own design effort – and take a look at their GitHub. (Another source would be their own site, but you don’t get the OpenSCAD code there.)

Yes, the v5 is longer than the v6 – and it had other issues. Otherwise there would have been no E3Dv6. There is a post in the E3D Blog descibing everything in detail.
But I had no intention to order a v6 or a Lite6 (which would have been great to start with on a first 3D printer – but I’m beyond that point), so I set out to make it fit.
If you want to get a v5 heatsink nevertheless, Aliexpress has got you covered. The v5 heatsink has an overall height of 50.1 mm and a diameter of 25 mm. Search for “J-Head”, as they confuse the mounting option with the type. FYI: The corresponding v6 heatsink is 42.7 mm high and has a diameter of 22.3 mm.

The modified parts are available on TinkerCAD. SRY again – still no account for uploading them somewhere else. Edit: As stated here, I finally found something – the extruder parts can now be found here, the adaptor for the vertical cable chain here on Libre3D!

There is a small bowden tube guide for the v5-specific geometry, the modified extruder body, the corresponding cover, and an altered idler – without the opening for cleaning the optical filament sensor. (It’s absence is also the reason, why I could print in blue and didn’t have to use black.) Instead of the sonsor there is now an M4 nut to hold a reverse bowden tube.

The body is not only adapted to the springs I had (these gears work best with sparse preload – use Loctite on your screws!), but also a TronXY X5S specific mounting plate.
My bag from Conrad says Part No. 1197628 and that it’s a stainless steel compression spring with 6.3mm diameter and 26mm length. If you search for that part now, you see the measurements have changed to 5.6×25.3 – so I have no idea if you would get the same parts…

Having printed everything and tried fitting it together, I fetched the X5S from the cellar and started removing the old carriage and the notoriously loose “vertical cable chain”. Mounting everything was quite complicated, so I don’t see this writeup as an invitation to recreate this. But maybe someone can seize on it and do it properly. πŸ˜‰

My belt tension had cracked ljbrumfield‘s carriage. And I couldn’t print a reinforcement without the X5S, so J-B Kwik had to come to the rescue. And I tore off a solder pad from the blower I had used on the last “fang” cooler. Re-soldered and glued that also.

The 45Β° prusa duct looks like this. (Said in advance: I was sceptical about the duct not cooling the counterside properly and my later prints proved me right about it, but that’s another post.)

The TronXY drag chain came apart and the single-bearing “vertical mount” also didn’t stay in place over the full movement range, so I made an adaptor for theΒ Tronxy X5S Top Mount Cable Chain Remix by Vetterick and installed that. But the cable management startet with some proper connectors for the steppers (could borrow the proper crimping tool for that series). The nice thing about crimp contacts is that you can test the electrical connection before finally snapping them into their housing.

Then came a fresh drag chain from the bits box. (First picture below shows it’s predecessor.)

Zip ties did not only manage my cables, but also supported my motor. πŸ˜‰

While I was at it, I also tried to correct the belt levels. Those were rubbing againt the idlers a bit. Afterwards I returned the whole machine to the cellar…
…and realized, that I hadn’t thought about filament guides beforehand.

Anyway: Time to adress that broken carriage. And add some metal, to get it unbreakable. (Can’t publish the STL of that plate, as the author has set the license to “no derivates”.)

Then came the filament guidance issue.

…and then I could finally start calibration and tests. The “print head” on the X-carriage got a lot heavier after all.

With spring tension, PID regulation, acceleration, jerk, extrusion factor (took the multiplier from the i3 Mk3 firmware on GitHub) and so on dialed in, I could finally tryout the TPU sample I had lying around for quite some time. Neither the Prusa extruder nor the E3D Titan Aero is completely foolproof with this material, but I got some result. πŸ™‚

And (of course) the first retraction test turned out pretty good – finally no “coasting”, “pressure advance” (or anything trying to fix what’s wrong with bowden extruders in general) anymore.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.