Why Prints Fail at the Same Height Every Time

A failure that happens at the same Z height on every print is one of the most useful clues a printer gives you. Random failures are hard; repeatable ones are almost diagnostic.
There are only two families of cause, and the first test splits them in one print.
The test that splits it
Rotate the model 90 degrees around Z and print it again.
| Result | Meaning |
|---|---|
| Fails at the same height | The machine. Something at that Z position. |
| Fails at the same feature | The model. An overhang, a thin wall, a bridging section. |
This costs one print and removes half the possibilities. Everything below assumes it failed at the same height.
Cause 1: the Z movement itself
On a bed-slinger, the bed moves in Z and the gantry does not. On a CoreXY like a Bambu, the bed drops. In both cases a specific band of Z can be affected by:
- A binding lead screw or Z rod. Clean, then lubricate with a light grease, not oil. A dry rod binds at the same spot every time.
- A bent lead screw. Watch the coupler while the bed moves through the problem range. If the rod wobbles, no amount of lubrication fixes it.
- Z-hop collisions. If the failure starts exactly when a hop begins, the nozzle may be catching the part.
Cause 2: the hotend changing behaviour
A partial clog does not appear instantly. It builds, and it often becomes visible at a consistent point because the extrusion demand changes there — a solid infill section after a sparse one, or a large flat layer after a small one.
Signs that point to the hotend rather than the Z axis:
- The top surface gets progressively worse before the failure
- Under-extrusion lines appear on the layer before the break
- The failure moves when you change infill density
Cause 3: cooling and warping at a specific height
If the part crosses from a thin feature to a wide one, the cooling demand jumps. The layer curls, the nozzle hits it, and the print fails — always at that transition. This is the one case that looks like a machine problem but is a model problem.
Test: print the same model with cooling reduced by 30% and see if the failure point moves.
Cause 4: filament diameter at that point
A spool with a diameter variation will under-extrude in one stretch. It shows up as a band, not a point. If the failed print has a visible thin band a few layers below the break, suspect the spool rather than the machine.
What to do first
- Rotate and reprint. Determines machine or model.
- Clean and lubricate the Z rod. Cheapest fix, most common cause on bed-slingers.
- Watch the coupler. A wobble means a bent screw.
- Cold pull the hotend. Rules out a partial clog.
- Change cooling. Only if the failure sits at a geometry transition.
FAQ
Why does it always fail at the same layer even after I clean the nozzle? Because the nozzle was not the cause. A repeatable height points at Z motion or at a geometry change, not at a random clog.
Can a warped bed cause a failure at one height? Not directly. A warped bed affects the first layer. If the failure is high up, the bed is not it.
My print fails at 80% every time — is that the spool? Possibly. If most of the spool has printed fine and the last section fails, measure the filament diameter in that section.
Does Z-hop cause failures at one height? It can, if a hop starts at that layer and the nozzle catches a curled edge. Turn Z-hop off for one test print.
Is it worth replacing the lead screw? Only after you have seen it wobble. A wobbling coupler with a straight screw is a loose grub screw, which is a two-minute fix.