Direct Granules Extruder V7.0, first prints

Added: 2026-05-06

[00:00:05]I haven't posted anything about my direct granulate extruder project for a while.

[00:00:10]The reason was that I only wanted to record another video once I had found a really, really reliable design, which is now the case.

[00:00:21]The last version was 6.2, and it seemed quite close to the goal - but during long-term testing, reliability turned out to be not so good: The extrusion rate occasionally dropped during printing, leading to poor or even unusable results.

[00:00:41]Once again, I built, tested, and rebuilt the extruder - long story short: here is version 7.0, and in this video, I'll briefly explain the most important considerations that led to this design.

[00:00:57]The auger is still a standard 6mm wood screw from the hardware store.

[00:01:02]The extruder tube is also still based on a 40mm long, M6 threaded sleeve made of stainless steel.

[00:01:10]One difference to previous versions is noticeable: The sleeve has a notch approximately halfway along its length.

[00:01:18]The reason for this lies in the fact that the 7.1mm bore is significantly smaller than the 8.5mm bore of the previous version.

[00:01:29]My experiments have shown that molten plastic at the cooler end of the tube can solidify again and then stick to the wall.

[00:01:39]During retraction and printing at varying speeds, some of the molten material inevitably rises too high.

[00:01:47]If the gap between the screw and the extruder wall is too large, this material cannot be immediately removed.

[00:01:55]Not until the "clump" has grown large enough, it breaks away, leading to fluctuating extrusion rates.

[00:02:03]Therefore, the gap needed to be reduced, which in the simplest case can be achieved with a screw of larger diameter.

[00:02:11]A 7mm screw leaves a gap of 0.75mm to the wall, which is small enough to prevent the resolidified plastic from sticking to the extruder.

[00:02:23]At the same time, the gap remains large enough to allow the flow of liquid plastic between the threads and the escape of air.

[00:02:32]This type of pressure equalization is necessary because I'm working with a screw that has a constant pitch along the entire length.

[00:02:41]As explained in several videos of this series: On the way from loose powder to a compact, molten mass, the material is compressed, and the trapped air must get out of the tube.

[00:02:54]However, the 7mm screw has two disadvantages: Firstly, this screw size isn't very common, at least not at my local hardware store, which is why I ordered a couple of them online.

[00:03:07]The second disadvantage is that the larger screw diameter requires higher torque.

[00:03:15]By reducing the size of the pinion on the stepper motor, I was able to increase the gear ratio to 32:9, just enough to apply sufficient torque to the screw.

[00:03:27]Printing worked, but not with a very high material throughput.

[00:03:32]So, plan B: reducing the bore to 7.1 mm.

[00:03:38]Since a higher wall thickness results in undesirably higher heat transfer, I removed material at the transition zone from the hot to the cold end with my lathe.

[00:03:48]A good side effect: The transition zone is more sharply defined, the transition from solid to liquid plastic is faster, which reduces the overall friction in the tube.

[00:04:01]The tube wall is smoother, simply because I'm using better, more expensive drill bits and also injected some coolant and lubricant into the sleeve before drilling.

[00:04:12]There are also changes on the cold end side: I'm using an insert made from a brass tube with 8mm outer and 7mm inner diameter.

[00:04:24]Brass has a lower static and kinetic friction than stainless steel, as I had already observed in earlier extruder versions.

[00:04:32]The high thermal conductivity of brass also ensures better cooling.

[00:04:38]Five filed grooves along the axis ensure that friction is high in radial direction, thus transporting the plastic grains downwards happens with minimal slippage.

[00:04:49]I notched the brass insert at the top so that the screw can grip the grains and press them downwards without them rotating along the upper edge.

[00:04:58]A drop of superglue keeps the insert in place - the temperature at the cold end doesn't exceed 50°C.

[00:05:06]The extruder tube is drilled out to 8mm, 20mm deep to almost the point of the outer groove.

[00:05:14]The brass tube is only 15mm long, creating a step along the extruder tube.

[00:05:21]This pre-compresses the powder and forces out some of the trapped air.

[00:05:28]This step must not be too large, as the powder cannot be compressed indefinitely - at least not with the finite force available to me.

[00:05:38]Too large and the extruder will get clogged; too small and the desired effect won't be achieved.

[00:05:44]The 0.5mm wall thickness of the brass tube works well.

[00:05:50]I also notched the top edge of the extruder tube.

[00:05:54]The inner diameter of the brass insert is slightly smaller at 7mm than the bore in the threaded sleeve, which widens the extruder's inner diameter slightly towards the hot end.

[00:06:07]This is better than a narrowing, as that would make the extruder more prone to clogging.

[00:06:14]I experimented a lot with enamelled extruder tubes because this glass coating significantly reduces friction with the tube wall. Since version 7.0 also works without the coating, I'm leaving it as is.

[00:06:28]Coating at least the lower half of the tube is an option for further experimentation.

[00:06:34]The water cooling system has only been slightly modified and now comes without front covers: The aluminum block is drilled so that the water is still routed around the extruder tube.

[00:06:46]The rear cover is milled from aluminum and is attached with two screws - glue serves as the sealant.

[00:06:53]The connections for the water hoses are screwed on to the left and right.

[00:06:58]The extruder tube is inserted into the aluminum block up to the notch and locked in place with two screws.

[00:07:05]It sticks out 5mm on the top of the aluminum block.

[00:07:10]The screw is no longer floating - brass washers ensure it is firmly connected to the coupling on the gearbox.

[00:07:20]A little bit of movement remains due to the ball bearing and the long lever arm of the screw.

[00:07:26]The screw axis again runs through the center axis of the tube - in the previous version, I had used a slight offset.

[00:07:34]With a total length of 60mm, the screw extends 31mm into the tube.

[00:07:42]During experimentation, spacers of 3, 2, and 1mm thickness proved helpful for varying the insertion depth.

[00:07:50]Too deep results in non uniform extrusion, too high prevents retraction, leading to stringing.

[00:08:00]The aluminum block on the hotend, with 15mm height, doesn't quite reach the notch of the tube.

[00:08:06]As I said, this version prints well and reliably.

[00:08:11]I'm performing the first print with the newly manufactured extruder using a 0.6mm nozzle - the particles leftover from drilling and filing will pass through the larger opening more easily.

[00:08:25]The print shows a bit over-extrusion.

[00:08:28]Manufacturing tolerances apparently mean that this extruder, that I made for the video, feeds slightly more material per screw revolution than the first extruder from version 7.0, despite identical specifications.

[00:08:46]The fact that the screw feeds the material with optimal grip is evident from the fact that even small variations of 5% or less in the slicer settings have a noticeable effect on the print.

[00:09:00]In previous versions, at least 10% or more was required to produce visible signs of over- or under-extrusion.

[00:09:11]The material is extruded with almost no slippage.

[00:09:15]Despite slight over-extrusion, the first print is successful - the walls of my standard track link are nice and smooth.

[00:09:26]The next print is done with a 0.4mm nozzle and a reduced filament factor.

[00:09:33]In terms of reliability, the extruder in version 7.0 sets new standards.

[00:09:40]I've already printed a small fleet with the first extruder from version 7.0, and there are no gaps visible on any of the Benchys.

[00:09:50]The second extruder of the series is working perfectly from the beginning.

[00:09:55]Infill and perimeters are printed at significantly different speeds without any gaps appearing on the outer walls - as seen in the previous video, the predecessor version had weaknesses in this regard.

[00:10:11]Infill is set to 80 mm/s, even on the top layers at 100% infill. The outer perimeters are printed at 50 mm/s, and small structures even at just 30 mm/s.

[00:10:30]Retract is set to a very low value with respect to my previous extruders, and there's no stringing.

[00:10:38]For the very first Benchy printed with a brand-new extruder, the result is quite impressive.

[00:10:47]In the next video, I won't be tinkering with the basic extruder design anymore. Instead, I'll be examining the specifics of printing with plastic powder and explaining which settings need adjusting to achieve the best possible results.

[00:11:02]Among other things, I'll be taking a closer look at stringing, bridging, and material throughput.

[00:11:09]That means printing, printing, printing.

[00:11:15]As always, the build instructions for Extruder 7.0 are available for download on my website.

[00:11:22]There's still a lot to do, so a huge thank you to everyone who has already supported me on my long journey to a working direct granulate extruder, whether through Patreon or the donation button on my website.

[00:11:38]A special thanks again to my anonymous major sponsor.

[00:11:53]Thanks for watching and: "I'll be back!".