I’m Daniel and I’ve started building a printer project for our company early this year, inspired by David’s large format machine. If everything works as planned, an even larger one might follow.
It features:
-1 by 1 by 1 meter build volume (what everyone strives for, right)
-heated moving bed (125°C and four zones)
-belt driven gantry
-Dyze’s new Atom Extruder
-Duet 6XD and Duet 23CL w/48v
-watercooling
-enclosure/heated chamber (80°C hopefully)
-BL Touch inspired bed probe (industrial style)
-it’s purpose is to print rough stock for cnc milling, so precision is not first nor second priority
I’m running the first tests right now after getting everything working AND keeping the magic smoke contained! There was so much stuff to figure out, I was used to Enders and Marlin before, but Duet’s been a first. Plus none of this stuff I’ve learned, just hobby tinkerer for life.
We’re a CNC shop so all the custom parts were made in house.
Z height will never be maxed out, probably 250 mm at most, but If we ever lack a proper chair, boy are we ready
The duet CAN system is balls to the walls man.
I’ve been brave and used a shielded 5 strand wire for power and comms to the motors, so basically one wire does all!
What still needs to be done;
-getting the enclosure done
-figuring out a door system
-figuring out how to best provide pellets and how to close up the upper portion
-add larger screen for DWC
-wheels get swapped for machine feet, they dislike the weight a lot (360 kg’s so far) the pid drive tuning cycle hated them as well (wobbly)
-getting a real water chiller for cooling
-figuring out materials and bed adhesion. The Atom demands SMALL pellets. 3.5 mm the most
According to MATHS, (chatgpt) the bed drive is able to lift 330 kg’s per side. Even more If planetary reduction is used ( but 10mm XL belt hast left the chat undoubtedly until then)
Each Zone is supported by 3 standoffs; one fixed, one movable in one axis and one only supporting Z to give freedom for expansion. Theres supposed to be a gap between the plates one the picture, needs to be adjusted again.
Hi, wish you good luck with commissioning of the printer. Please let me know how the setting procedure of the Dyze pellet printer looks and what the road to the first products will look like. Practical experience is what I am looking for and forward to. From what I hear the Dyze swallows pellets up to 5 mm, but may have non-ideal printconsistency when the pellet variability is high. This might be the reason Dyze specifies 3 mm pellets - so the prints looks nice.
I am looking at building a LFPP myself at 1,5x1x1 m size or longer. Will start CAM design on a friendly 1x1x1 m printer here in Prague probably next week. Since the Dyze is one of the candidates for the pellet extruder, would you please share some documentation info if they supply any?
I wonder why normal square or rectangular closed profiles (jäckel in EU) are seldomly used instead of open construction profiles. Can be drilled even manually (as Ivan Miranda on youtube proves), have about 50% lower weight 1/3 price and double the stiffness (lower bending). Or for the same price about 5x more stiff - less bending excursion on load. Since the stiffness is all we want in a printer I find using open profiles suboptimal.
I myself am an acoustical engineer, so dynamical systems are my bread and butter. Calculating excursion of a beam on a specific load staticaly and dynamically is an easy task . Therefore comparison between profiles is straigtforward … so I wonder why to use suboptimal sollution.
I am aware that this project is a 10 000 variables to solve together and due to time constrains it is easier to just overengineer it and call it a day … but this topic came up easterday in a pub after a few beers…
Of course we can discuss the extruder in detail. The production version will be more forgiving in pellet size, this one swallows 4mm ish in diagonal size pellets only. I’ve got some fine PLA with uneven size distribution, some are elongated like long grain rice, one of them is enough to block everything up. So it really has to be even. If you can and weight is no issue, go with the typhoon i reckon. Our next build will be equipped with larger one as well.
To be honest, I chose the profiles I’m familiar with in an automation environment, and can bolt stuff onto-and together without having to drill or tap anything. They even come cut to length. Otherwise I’d have to block machine hours and do cam work, which costs way more money. remember this is an industrial build paid and used by/inside our company. If built at home inside the garage, I’d probably approach it differently.
Plus, we want to print rough stock with excess material only, so it can be mill finished afterwards. Beauty isn’t even a requirement, just pure volume.
When using a pellet head, speed anyways isn’t your first priority. Of course you don’t want a flimsy machine by any means, but if you choose 8080 profiles or even 8040 you can rest assured that you’ll reach more than plenty of stiffness, no matter what tube design you go for.
In the end, if you choose to pick a 15 kg`s or more extruder, a 2 kg vs. a 4 kg gantry tube won’t matter anyway
If you plan to build a high speed high dynamic system with filament like a voron, that’s a completely different story. Low weight, outmost stiffness is key there.
I really appreciate you taking the time to document your efforts in building this awesome printer. This is the first pellet extruder printer that I have seen with a moving bed. I am also curious to see how well the Dyze’s new Atom Extruder performs. Personally, I feel like “small” and “pellet extrusion” are incompatible. Even the 17mm ID feed throat on my MDPH2 would experience occasional bridging, which could lead to a print failure.
Are you able to provide any more details on the application?
Try and purchase spherical pellets that are pelletized underwater. I talk more about it here: DIY Large Format 3D Printer
Sourcing the right pellet material AND size is going to be the biggest hurdle I believe (at least with the Atom) . I work in the plastics industry by trade (mould making) and pellet size is usually the least thing on your mind. Just make sure you don’t throw metal scrap in there. Usually unfilled PA’s come in the shape of ‘micro lentils’ most of the time. Those would be great, but we want to avoid unfilled.
Currently i’ve got Nanovia’s PC CF that extrudes great but sticks zero to my build plate. So i’ve ordered their petg CF for testing. They’re shaped cylindrical though. But pretty tiny.
But those are just for testing since they’re priced far from reasonable.
Our use case is thermoforming plugs for plug assist, maybe you’re familiar with this. Right now they’re being cut from blocks (PU or epoxy tooling board) which is very messy and wasteful. So we plan on printing rough stock with ~2mm surplus to be finished. But they need to withstand some temperature and the less I have to fight warping the better! Your shattered glass bed actually made me sigh lol
Wow, that is an excellent use! If you are able to share a couple images (after printing, after machining, and during the molding), I would love to share this as 3D printing application for my class. I understand if you are not able to do due to time or proprietary parts.
By the end of the month I should see some progress on a large pellet extruder mounted on a robot arm. local university offspring startup: http://www.3deposition.com/
Also been looking for: https://greenboy3d.de/
But as the guy is young he is in for a surprise. From my understanding its a toy not worth my while.
1/ big filament Voron by the end of the year
2/ 1,5-2,5x1x1 m pellet printer next year.
or …
3/ 1,5x1x1,5 with dual gantry - fast filament head + slow pellet head. Learn how to use it simultaneously. 5 layers of outer filament skin then one big chunky layer of pellets inside it.
Of course, I can provide regular updates on the projects we try to tackle on this machine.
Today I did some testing to get smoother movements and I’m pretty happy with the results:
Basically I raised the motor current from 1600 to 2500, got rid of some whining; furthermore switched from x16 to x32 microstepping which solved most of it; then tweaked the A parameter (lower) to get rid of the last bit of raspiness.
Yet I still struggle with some jittery slower moves.
Would you mind sharing your XY motor closed loop settings so i get a rough idea of where I’m at overall ?
There was so much stuff to figure out, I was used to Enders and Marlin before, but Duet’s been a first. Plus none of this stuff I’ve learned, just hobby tinkerer for life.
