A first layer that sticks fine on one side of the plate and not the other is almost always a leveling problem, even on a machine with an automatic sensor. This page covers what leveling actually corrects, how to run a manual paper-method calibration, and how often to revisit it.
Mechanical leveling means adjusting the physical position of the bed (or the head, depending on the machine) so its surface runs parallel to the X and Y axes. Mesh bed leveling changes nothing mechanically: the firmware probes a grid of points across the plate, builds a map of height differences, then adjusts Z during printing to follow those variations.
According to the Prusa Knowledge Base, the Bed Level Correction feature compensates for residual imperfections by virtually raising or lowering the bed on its left, right, front, and back sides, within a range of plus or minus 100 microns. That range is intentionally limited: it catches small leftover deviations, not a genuinely warped or poorly mounted plate, which needs mechanical adjustment first.
That is why a badly warped plate keeps causing problems even on a machine with automatic leveling: the firmware can follow a gentle slope, but it cannot compensate for a significant warp without sacrificing first-layer height in some spots.
Printers with automatic bed leveling do not all use the same sensor technology. On the Bambu Lab A1 and P1S, Bambu Lab's documentation describes strain gauges (force sensors) mounted under the plate: the nozzle taps the surface at a grid of points, and the sensor detects contact at each one to build the height mesh. Other brands such as Creality or Anycubic fit part of their lineup with inductive probes mounted near the nozzle that measure distance to the plate without touching it; the output in both cases is a comparable compensation map.
Either way, how reliable the measurement is depends on how clean the nozzle and plate are at the time of probing. Bambu Lab's troubleshooting page for A1-series leveling failures notes that filament residue on the nozzle or debris on the plate throws off contact detection and can fail the leveling cycle before the print even starts.
Auto-leveling corrects a height map, not the reference height itself; that is the job of Z-offset (or Live-Z, depending on the brand), which sets the gap between the nozzle tip and the plate at the very start of the first layer. A perfectly calculated mesh with a Z-offset set too high still gives a first layer that will not stick; one set too low crushes the filament or scratches the plate, regardless of how good the auto-leveling is.
A greasy or dusty plate also throws off a force sensor's contact reading or an optical probe's reflection, which can shift the calculated mesh slightly. The cleaning method suited to each surface is covered on the bed adhesion page, before leveling even comes back into the picture.
Z-offset calibration is best done at printing temperature, with a hot nozzle and bed: metal expands slightly as it heats up, so a value set cold can end up slightly off once the machine reaches working temperature.
On a machine without an automatic sensor, or to check the mechanics before trusting a mesh, the paper method is still the standard reference. Start by heating the bed and nozzle to the intended printing temperature, since thermal expansion slightly shifts heights measured cold. Then slide a standard sheet of paper between the nozzle and the plate, directly under each leveling screw, usually located at each of the four corners.
For each corner, lower the nozzle until it nearly touches the plate, then turn the leveling screw until you feel light friction as you slide the paper back and forth: not enough resistance to jam the paper, but not a completely free slide either. Repeat at all four corners, then recheck the first corner, since adjusting one corner slightly changes the plate's overall tilt.
Once the mechanics look right, rerun the Z-offset routine or the slicer's first-layer test, because the paper method sets the plate's parallelism, not the exact first-layer distance. The first layer guide covers how to dial in that distance once the plate is level.
A poorly leveled bed shows up as an uneven first layer rather than a uniform defect: one corner where the filament forms a thin, translucent line that does not fuse into the next pass, opposite a corner where the nozzle scrapes the plate, leaves marks, or even jams extrusion. A part that consistently peels off the same corner of the plate while the rest adheres fine almost always points to a leveling gap rather than a pure adhesion problem.
On an auto-mesh machine, repeated leveling-cycle failures at startup, or an error message mentioning contact detection, usually point to a hardware issue (a dirty nozzle, a dirty plate, sensor wiring) rather than a setting that just needs redoing.
On an auto-leveling machine, the mesh is generally recalculated before every print or on a firmware-adjustable schedule, which limits how often manual intervention is needed day to day. A mechanical check is still worth doing after moving the printer, swapping the plate, a knock to the frame, or if auto-leveling failures become frequent despite a clean plate.
On a manual machine, rechecking leveling every few weeks of regular use, or as soon as an uneven-adhesion symptom shows up, keeps you from piling adjustments onto other settings when the real cause is the plate.
256 × 256 × 256 mm
More space on every axis than the A1 mini, while keeping the A1 family workflow.
An open frame does not provide the same thermal conditions as an enclosed printer. Choose materials accordingly.
Specifications and seller
256 × 256 × 256 mm
The same nominal volume as the A1, with an enclosure and a different motion system.
Check the precise AMS version, adapters and accessories included. Enclosure alone does not establish material suitability.
Specifications and sellerWe organize manufacturer and community documentation; we have not measured these products ourselves. The manual for your exact model takes precedence over general guidance.
Use the free buying checklist and budget worksheet before comparing offers.