After resolving an initial issue with Z-axis movement, my Creality Ender-3 V2 enters my 3D printing pool. I mainly got it because it was becoming a hassle to switch my MatterHackers Pulse XE between PLA and PETG printing. My intent is to leave the Ender-3 V2 loaded with PLA and the Pulse XE loaded with PETG. I’m sure that’ll only last for as long as it takes for one of these printers to develop a problem, but I’ll figure out how to cross that bridge when I come to it.

The best news is that the extra cost for V2 was worthwhile: this printer operates without the whiny buzz of older stepper motor drivers. It’s not completely silent, though. Several cooling fans on the machine have a constant whir, and there are still some noises as a function of movement. Part of that are the belts against motor pulley, and part of that are roller wheels against aluminum.

These rollers are the biggest mechanical design difference between Creality’s Ender line and all of my prior 3D printers. Every previous printer constrained movement on each axis via cylindrical bearings traversing over precision-ground metal rods. One set for X-axis, one set for Y-axis, and one set for Z-axis. To do the same job, the Ender design replaces them with hard plastic rollers traversing over aluminum extrusion beams.

The first and most obvious advantage to this design is cost. Precision ground metal rods are more expensive to produce than aluminum extrusions, and we need them in pairs (or more) to constrain motion along an axis. In contrast, Ender’s design manages to constrain motion by using rollers on multiple surfaces of a single beam. In addition to lower parts cost, I guess the assembly cost is also lower. Getting multiple linear bearings properly lined up seems more finicky than bolting on several hard plastic rollers.

Rollers should also be easier to maintain, as they roll on ball bearings that keep their lubrication sealed within. Unlike the metal guide rods that require occasional cleaning and oiling. The cleaning is required because those rods are exposed and thus collect dust, which then stick because of the oil, and then the resulting goop is shoved to either end of range of travel. Fresh oil then needs to be applied to keep up with this migration.

But using rollers also means accepting some downsides. Such a design is theoretically less accurate, as hard plastic rollers on aluminum allow more flex than linear bearings on precision rods. Would lower theoretical accuracy actually translate to less accurate prints? Or would that flex be completely negligible for the purpose of 3D printing? That is yet to be determined.

And finally, I worry about wear and tear on these roller wheels. Well-lubricated metal on metal have very minimal wear, but hard plastic on aluminum immediately started grinding out visible particles within days of use. I expect the reduced theoretical accuracy is minimal when the printer is new but will become more impactful as the wheels wear down. Would it affect proper fit of my 3D printed parts? That is also yet to be determined. But to end on a good note: even if worn wheels cause problems, they should be pretty easy to replace.

The first test print on my assembled Creality Ender-3 V2 showed some artifacts. General symptoms are that some layers look squished and overall height is lower than it should be.

These two parts were the same 3D file on two separate print jobs. The piece on the right was printed before my modification, showing many problematic layers and the overall height is lower than it should be. Hypothesis: my Z-axis is binding, occasionally failing to move the specified layer height. Verification: With motor powered off, it is very hard to turn the Z-axis leadscrew by hand.

This put the Z-axis motor mount under scrutiny.

Looking closer, I saw it was not sitting square. There is an uneven gap forced by the motor that is slightly fatter around the black midsection than its silvery ends. This means when the motor mounting block is tightened against the vertical Z-axis extrusion beam, that motor rotates and output shaft tilts off vertical.

A tiny gap would not have caused a problem, because the shaft coupler could translate motion through a small twist. However, this gap is larger than the shaft coupler could compensate for, causing it to bind up. I saw two ways to reduce this gap. (1) Grind the side of Z-axis stepper motor to eliminate the bulge, or (2) insert a spacing shim into the motor mount. I don’t have a precision machining grinder to do #1, but I do have other 3D printers to do #2. I printed some shims of different thicknesses on my MatterHackers Pulse 3D XE.

I tried each of those to find the one that allowed smoothest Z-axis leadscrew turning by hand.

After this modification, the print visual quality and Z-axis dimensional accuracy improved tremendously. The piece on the left was printed after this modification.

So that’s my first problem solved, and I think this printer will work well for the immediate future. However, looking at how it was built, I do have some concerns about long-term accuracy.

From what I can tell, Creality’s Ender-3 is now the go-to beginner 3D printer. It works sufficiently well in stock form out of the box. And if anyone wants to go beyond stock, the Ender 3 has a large ecosystem of accessories and enhancements. These are exactly the same reasons I bought the Monoprice Select V2 several years ago. So when Creality held one of their frequent sales, I picked up an Ender-3 V2. The key feature I wanted in V2 is Creality’s new controller board, which uses silent stepper motor drivers. A nice luxury that previously required printer mainboard replacement.

When the box arrived, I opened it to find all components came snugly packaged in blocks of foam.

The manual I shall classify as “sufficient”. It has most of the information I wanted from a manual, and the information that exists is accurate. However, it is missing some information that would be nice, such as a recommended unpack order.

And this is why: the Ender-3 came in many pre-assembled components, and when they are all tightly encased in foam it’s not clear which ones were already attached to each other. Such as the fact the printhead was already connected to the base. I’m glad I didn’t yank too hard on this!

That minor issue aside, it didn’t take long before all pieces were removed from the box laid out.

The Z-axis leadscrew is something to be careful with, vulnerable to damage in the sense that the slightest problem would cause Z-layer issues in the print. It was cleverly packed inside a bundle of aluminum extrusion beams, protected by a tube of what looks and feels like shrink wrap tubing.

All fasteners are bagged separately by type and labeled. This is very nice.

As far as I can tell, all of the tools required for assembly are bundled in the box. The stamped-steel crescent wrenches weren’t great, so I used real wrenches out of my toolbox. In contrast the hex keys were a pleasant surprise, as they had ball-ends for ease of use. I considered that a premium feature absent from most hex keys.

I was initially annoyed at the instructions for the filament holder spool, because it looked like the channel was already blocked by some bolts.

But then I realized the nuts are not perfectly rectangular. Their shape gives them the ability to be inserted into the slot directly, without having to slide in from the ends of the beams. As the fastener is tightened, they will rotate into place within the channel. These are “post-assembly nuts” because they allow pieces to be added to an extrusion beam after the ends have been assembled. These are more expensive than generic extrusion beam nuts and a nice touch for convenience.

Here is the first test print. It’s pretty good for a first print! But not perfect. Uneven vertical wall indicates issues with Z-axis movement.