3D Printing Stack Exchange is a question and answer site for 3D printing enthusiasts. It only takes a minute to sign up.
Sign up to join this community
I'm trying to build a DIY 3D printer for myself. I've been exploring many different styles of printers and found this type of printer that has a fixed bed that stays fixed in one place and the whole gantry moves which includes all axes.
Check out this video for reference and skip to 10:50:
Why is this so rare?
Such kind of printers usually harder to assembles, calibrate, and maintain because 3 axes machine is a bit more complex than 2 axes. For instance, it's can be tricky to move an entire extruder among all 3 axis and some of such printer's designs may require even dedicated exruder's design like Bowden Extruders.
Are there flaws in this design?
The key disadvantage of such kind designs is complexity with moving of an extruder among all 3 axes. Moving platform by at least one axis simplifies that.
Will print quality be affected by using this approach?
It depends on the exact printer's design, so, potentially you can have issues with ease of assembling and maintenance due to more complicated construction and as a consequence higher risk of low printing quality due design, assembly or configuration mistakes.
On the other hand, if you already have some device with precise enough 3 axis machine, like CNC milling machine, you can upgrade it to 3D printer by installing an extruder, however, it would also require update of software and, probably, electronics.
As a frame challenge, they kind of are. It's just that the optimal, and in some sense only reasonable, design for a fixed build platform that doesn't move on any axis is the delta robot geometry. This design is not the most popular, but it's far from obscure - there are lots of cheap entry-level delta printers available as well as higher-end ones.
What makes delta optimal? Keeping a single gantry (like the Ender 3 has) square with a fixed bed height is hard enough; it requires a very rigid frame and perfect rails/rods. If you want to have a multi-axis motion system over a completely fixed bed, you have a whole extra dimension in which it can be non-square. Mechanically (at least for a plain cartesian configuration) it's like having a gantry that moves between two other gantries, each of which already has concerns about remaining square, and the resulting system might not even end up being planar.
The delta configuration avoids this by not having a 2-axis motion system that's constrained to a particular movable Z height relative to the bed, but instead calibrating the transformation between a constrained motion system with free degrees of freedom and normal cartesian coordinates. Any error can just get calibrated out.
On top of that, delta has an extremely low toolhead mass, allowing extreme speeds and accelerations.
Within the realm of Cartesian machines (eg having orthogonal X, Y, Z axis) consider the sort of design you do see: the Ultimaker-like X-Y stage over a build platform which can drop.
The question is then, why drop the build rather than raise the mechanism?
Likely the reason is mechanical simplicity. The X-Y gantry is the part that moves around quickly and may vibrate. It's also the part that is rather complex and has lots of electrical cables. Making that fixed in the housing is generally a simplicity. In contrast, the built object in a typical consumer printer has lower mass, and the build platform only moves slowly so variation in dynamics from its increasing mass isn't really an issue (increasing build mass typically isn't even really taken into account in designs where the build rides on a rapidly moving Y axis either).
You might ask, what would a machine like this look like if instead of dropping the build you raised the mechanism?
It turns out such a thing does exist, in the form of a clever hack where you run a linear rail up the wall of your hackerspace, mount your entire Ultimaker-style printer on it (less build platform and bottom panel), and tap out the Z control signals such that the printer climbs itself up the wall leaving some towering sort of build on a fixed platform below. Conceivably if you wanted to start there you could also saw off all but the top third or so of the cabinet to leave the X-Y gantry with some rigid frame. But it's a bigger, more expensive product that doesn't work simply by lifting it out out of the shipping box. Apart from very unusual or "proof of concept" builds, it doesn't make a whole lot of sense.
The thing about the 3d printer realm is that you're free to try anything you like. Some ideas work. Some ideas don't. Some that go a little bit against orthodox mechanical design as taught in Mech E. departments turn out to work a bit better than they should fairly be expected to, and make it into products. But generally what is on the market is what has proven to present a good balance between cost and utility.
Why is this so rare?
This is rare because machine designers basically copy the designs that work, without trying to innovate too much. As a result of this, Adrian Bowyer's original RepRap design (a travelling type) is currently the most prominent desktop 3D printer design around.
Are there flaws in this design?
There aren't any flaws in the concept for a static bed, after all, it is just a design concept; there are advantages and disadvantages for each type (static, floating and travelling).
Will print quality be affected by using this approach?
The individual implementations will have varying levels of success. Certain designs provide more quality than most due to the rigidity of the frame that is required (e.g. CoreXY) which in turn provides better print quality. Any printer design can produce a good print quality when the machine is well designed and well maintained.
Now that I learned the truth about supporting the XY gantry, I can answer this. No, it’s not about complexity, it’s about stability. There’s nothing complex about 3-axis movement. However, moving the gantry has more requirement than moving the bed, and this additional requirement is the reason why many implementations stick to moving the bed instead.
The XY gantry moves around quickly, horizontally. Which means that it must be supported by at least 4 points of supports forming a square, otherwise, it’ll wobble. Yes, 3 points define a plane, yet the XY gantry cannot be stably supported with just 3 points, which is why this is challenging. Synchronizing the movements of those four supports is not as easy as linking them all together with a single belt. This introduces another challenge. To do this correctly, typically, it requires multiple motors for the Z axis (e.g. Voron 2.4 has four Z-axis motors.). And even with 4 points of supports, they can still fail to stably support the rapidly-horizontally-moving gantry.
On the other hand, the stuff on the bed does not move around. The weight does not move around. Meaning, the bed is absolutely fine with less than 4 supporting points (e.g. 1, 2, or 3 leadscrews).
Now you should see why moving the bed is considered as easier and more stable.
