This is a modular Z-Carriage upgrade for the Anet A6-printer. It replaces the stock carriages (those pierced with lots of screws).

The stock carriages don't have a belt tensioner which I consider to be crucial for good quality. I was a bit unsatisfied with the existing upgrades, as I had troubles to get the perfect measures due to warping issues.

The model I developed is 100 % parametric, so everyone can edit measures and tolerances at will.

My priorities were roughly:

• perfect distance of the x-rods
• perfect fastening of the x-rods
• modular design to be able to replace single parts and reduce print time on failure (implicitly supports multi color prints)
• reducing the need for supports
• ability to upgrade to plain bearings (like Igus)
• optics and customizability

Some detail decisions:

• The holes for the rods in the braces seem to be oddly shaped: They aren't circles but beveled 8-sided polygons. The rod fits in easier and straight-lined holes can be printed more exactly than curved walls. The polygon is oriented with vertices in each main direction so the rod can be exactly centered without touching the clamp's gap. The polygon is beveled for making the hole not bigger than necessary - some walls are already very thin. Another beveling makes a slight funnel for easier insertion and reducing an elephant's foot which could make the hole too small for the rod.
• Some screw holes where you'd expect to have the shapes of hex nut & screw head simply have two hex nut holes. For most of those holes this allows you to decide which way the screw should go in. For others it simply wasn't worth the effort to specialize the screw head hole ;)
• Most angles are 45 °. This angle typically can be printed as overhang without any support.
• The tensioner knob is rather long. I wanted to provide the longest possible range for tensioning. To get the wheel holder close to the knob I had to bury the knob very deep into the carriage.
• I decided to provide my own belt fasteners as the most common ones are rather thick-walled and have sharp edges. Mine can be be customized for your needs, have beveled edges and their measures influence the size of the belt tunnels within the carriages - try to keep them small.
• The carriage covers have a grid as bottom layer. This allows to peel them off fairly easy from the bed, but still prevented bending during the print. If you experience any problems with bed adhesion, try to remove the grid or use a raft.

1 Shopping list

Lots of screws! That's the main downside of my design. I tried to mitigate this by allowing to use the screws from the original z-carriages.

All screws are M3 (3 mm thread diameter)

• x_rod_brace_clamps: 8 x 14 mm (12 mm minimum, 18 mm possible, if you use the xz_stop, one screw has to be 20 mm as absolute minimum)
• x_rod_adjustment: 4 x 12 mm
• x_rod_brace to z_carriage: 8 x 10 mm (8 mm minimum, 18 mm possible)
• motor screws: 4 x 10 mm (just use the original screws)
• z_rod_bearings: 4 x 16 mm (14 mm minimum, just use the original screws)
• z_spindle_nut: 8 x 16 mm (just use the original screws)
• x_belt_tensioner_knob: the longer, the better (the original 18 mm screws are ok)
• x_belt_tensioner (wheel): 16 mm (the original 18 mm should fit, but might touch the back x rod)
• z_carriage_cover (optional): 8 x 6 mm (8 mm possible)
• hex nuts: 35 (add another one if you want to secure the tensioning screw against turning within the wheel holder; remember to use one lock nut to assure the belt wheel turns freely and the screw cannot get loose)

Also needed: a belt wheel

2 Calibration

First of all you should print right_inner_x_rod_brace and check for the following issues:

• The hex nuts should fit in with little to no force
• The screws should fit in with little to no force
• The clamp should have a small visible gap
• The x-rod shall be a tight (but possible) fit. If the clamp's gap is visibly bent open after the rod has been inserted, something is wrong. On the other hand: if the rod isn't held by the clamp at all, that's no better.

If anything is wrong, take a look into the tolerances.scad. The variable _ can be altered to get a better fit. Have a look at https://www.thingiverse.com/thing:2805660 to figure out the perfect values.

3 What to print

Components from left to right (roughly):

• left_outer_x_rod_brace: holds the left ends of the x-rods and provides the motor support
• left_z_carriage: the core part of the left carriage, connecting the z-rod, z-spindle, x-rod-braces
• left_inner_x_rod_brace: holds x-rods and optionally the xz_stop
• right_inner_x_rod_brace: holds x-rods, can be used for calibrating the tolerances (see above)
• right_z_carriage: the core part of the right carriage, connecting the z-rod, z-spindle, x-rod-braces and the belt tensioner
• right_outer_x_rod_brace: holds the right ends of the x-rods, has a hole for the tensioner knob and has text on its surface
• x_belt_tensioner: holds the right wheel of the x-belt
• x_belt_tenioner_knob: contains a hex nut to pull the x_belt_tensioner away from the motor

Optional:

• xz_stop: This is necessary when not using a proximity sensor for auto-leveling and when using the original position of the x-stop switch. If you're not sure about this: just print it.
• left_z_carriage_cover, right_z_carriage_cover: that's mostly for optics and maybe dust protection
• belt_holder (2 x): An alternative to zip ties to fasten the x_belt to the x carriage. They're smaller than the popular ones and can be configured for you needs

4 How to assemble

Please have a look at the overview and the photos I made. If anything is unclear, don't hesitate to ask in the comment section and I'll do my best to explain. This section will be updated over time.

5 What's still open/planned

• Holder for plain bearings
• allow the x-stop-switch to be moved from the x-carriage to the left z-carriage (no good ideas so far)
• As with other designs, the right carriage is too light-weighted. The motor on the other side leverages the right z-carriage, when the x-carriage is on the left side. I'm still looking for something useful to weight it down. A spool would be too heavy, maybe a second motor ... ;)