Horizontal ridges that repeat every 8 mm (or your lead screw’s pitch) point straight to Z banding. The pattern is periodic — not random — and that predictability is your best diagnostic clue. Tracing the mechanical chain from motor to gantry saves hours of guessing.
1. What Z Banding Looks Like (and Why It’s Not Always Wobble)
Z banding appears as a consistent series of horizontal ridges with a spacing equal to the lead screw’s lead. On a typical T8 screw (8 mm lead, 2 mm pitch, 4 starts), you’ll see a ridge every 8 mm of vertical travel. If the spacing is wider or chaotic, look elsewhere — random extrusion issues, filament diameter swings, or temperature fluctuations create non-periodic scars. Resonant ghosting leaves a decaying echo, not a steady drumbeat.
The screw rotates once per 8 mm of Z travel. Any lateral runout or binding in that rotation applies a cyclic side load to the gantry or bed, shifting the layer position. Even a square frame can’t absorb it if the screw shoves sideways every revolution. Ridges may come in pairs or vary subtly in intensity if runout interacts with the coupler or nut preload, but the fundamental period stays tied to the screw lead. Measure the centre-to-centre distance between two ridges with callipers. If it matches your screw lead, you’ve found the mechanical culprit.
2. Root Causes: Tracing the Mechanical Chain
Z banding starts with a rotating lead screw that pushes the moving assembly sideways once per revolution. The chain: motor → coupler → screw → nut → linear guides. Any defect that converts rotation to lateral displacement will repeat in the print.
2.1 Lead Screw Straightness
A bowed screw applies a cyclic side load. A 0.05 mm bow at the midpoint of a 300 mm screw can shift the nut enough to produce a 0.04–0.06 mm ridge. The force direction rotates with the screw, so ridges may be stronger on one side. This is forced runout, not wobble.
2.2 Coupler Misalignment
Rigid couplers transmit motor shaft runout directly to the screw. A 0.03 mm eccentricity at the motor can produce 0.15 mm of shake at the top of a 300 mm screw. Flexible jaw couplers absorb angular misalignment, but leave a 0.5–1.0 mm gap to avoid preloading the motor bearing. If the screw bottoms out against the motor shaft inside a rigid coupler, bearing play becomes layer error.
2.3 Nut Constraint Versus Floating
A rigidly mounted nut amplifies screw runout because it can’t shift to average lateral force. A spring-loaded anti-backlash nut allows slight radial float while maintaining zero axial play, decoupling the carriage from the screw’s side force. But excessive spring tension can cause stiction, turning a small wobble into a stick-slip band.
2.4 Guide Rail Binding
Binding linear guides cause the carriage to stick and release, producing ridges that mimic over-extrusion. Dusty rods, misaligned bearings, or tight bushings all contribute. Disconnect the screw: the gantry should fall under its own weight; if not, binding is present.
3. Quick Checks to Narrow It Down
-
Measure banding pitch. Use callipers across multiple ridges and divide by the number of gaps to get the average spacing. Compare it with your lead screw lead. For most T8 screws, that’s 8 mm; for a TR8x2 (single start) it’s 2 mm. A match means the banding is almost certainly screw-linked.
-
Rotate by hand. Disable steppers and spin the screw manually. Feel for a tight spot that repeats each turn — that’s a bent section. Watch the nut: if it oscillates side to side with the same frequency, the screw runout is transferring directly.
-
Listen to the motor. Run the Z axis at 1–2 mm/s. A smooth whoosh that rises and falls in pitch every revolution points to a bent screw loading and unloading the motor.
-
Check runout with a dial indicator if you have one. Measure at the midpoint, top, and near the coupler. Runout under 0.05 mm total indicator reading across the threaded length is good; over 0.1 mm will likely need attention. If runout is worst at the midpoint and reduces near the bearings, the screw is bowed, not just eccentric at the coupler.
| Pitch of ridges (mm) | Pattern | Likely cause |
|---|---|---|
| 8 (exactly one per revolution) | Regular, often asymmetric | Bent lead screw or eccentric coupler |
| 2 (4 ridges per 8 mm) | Faint, repeating closely | Thread-form defect or debris in nut (less common) |
| Variable, not 8 mm | Chaotic, irregular | Extrusion, temperature, or filament issue |
| 8 mm but intermittent | Random intensity | Guide binding interacting with screw runout |
If your measured pitch doesn’t match the screw lead, look at the extruder or hot end first — dial in your extrusion multiplier before chasing the Z axis.
4. Fixes That Actually Work
I’ve ordered these from the least invasive to the most involved. Start at the top and work down only if the symptom persists.
4.1 Realigning and Isolating
Most banding comes from coupler misalignment, not a bad screw. Loosen motor mount bolts and coupler grub screws. Run the Z axis by hand to let the screw centre itself. With a flexible jaw coupler, leave a 0.5–1.0 mm gap between motor shaft and screw end to avoid preloading the motor bearing. Retighten in the relaxed position. If the motor is fixed, slacken the coupler, run the nut to the top, let the screw hang, then tighten. Re-test; banding often disappears.
4.2 Straightening a Bent Screw
If runout exceeds 0.07 mm at the midpoint, straighten it. Roll the screw on glass or granite, mark high spots, and apply gentle pressure. Use a V-block and soft-jaw vise, or an improvised jig. Bend in 0.5–1° increments, aiming for under 0.05 mm total runout. If you can’t get below 0.1 mm, replace the screw.
4.3 Nut and Guide Adjustments
Clean the screw with IPA and a toothbrush, then lube with light PTFE grease. If the anti-backlash nut is adjustable, back off the spring until axial play just disappears and the screw turns freely by hand. Disconnect the screw and move the carriage; it should glide smoothly and fall under its own weight if vertical. If not, loosen bearing mounts, slide the axis full travel, and retighten at centre.
4.4 When to Replace
If runout remains above 0.1 mm after straightening, or the threads show visible galling and pitting, the screw is done. A higher-grade lead screw (C7 rolled or ground) keeps runout typically under 0.05 mm over 300 mm. A two-screw Z-axis synchronization (dual motors with a belt) can average out residual errors, but it adds complexity.
5. Preventing Banding Through Design Choices
Preventing banding is easier than fixing it. A few design choices keep the screw from causing trouble.
Minimise unsupported screw length. A 300 mm screw with only a motor-end bearing will have more runout than a 200 mm one. For long screws, use a floating nut that allows 0.2–0.3 mm radial shift to decouple the carriage. Rigid nut mounts only work with perfectly aligned dual-end guidance; otherwise, they transmit every wobble.
Adding a top bearing can stabilise the free end only if alignment is perfect. Any misalignment between motor, screw, and top bearing will bind and increase banding. A motor-integrated lead screw eliminates the coupler and its runout, at the cost of mounting flexibility — but for most Cartesian printers, the improvement in Z consistency is worth it.
| Configuration | Typical unmounted runout tolerance | Banding risk | Notes |
|---|---|---|---|
| Rigid coupler, rigid nut, no top bearing | 0.05–0.08 mm | Medium–High | Alignment-critical; any coupler runout goes to print |
| Flexible jaw coupler, floating nut, no top bearing | 0.05–0.10 mm | Low | Nut absorbs small lateral movement |
| Dual rigid bearings (motor + top), rigid coupler | 0.03 mm or better | High if misaligned | Any misalignment causes binding that amplifies bands |
| Motor-integrated screw, floating nut, top guide bearing only for radial support | 0.02 mm typical | Very Low | Removes coupler error; top bearing |