3D Printer FYIThe first week is not about speed or exotic filament. Success is a printer that can reprint the same small test with a consistent first layer. Vendor wizards differ, but the order of work does not: inspect, assemble per the manual, check motion, set Z, load an easy filament, print a known-good model, then change one variable if something is wrong.
What success looks like before you start
Before you power on, decide what “done” means:
- The frame sits on a rigid, level table (not a folding card table that rocks).
- Shipping ties, foam, and bed clips are removed. Prusa’s current unpacking manuals still warn people to keep the original foam for warranty shipping — do that for any brand (Prusa CORE One+ Gen 2 unpacking).
- You have the exact handbook for this printer, not a YouTube video for a cousin model.
- You have isopropyl alcohol or the cleaner the bed manufacturer specifies, a scraper rated for that sheet, and one roll of basic PLA (or the vendor’s starter filament).
If the printer arrived with a self-test / first-run wizard, run that after the physical checks below. Wizards assume the machine is already free of packing damage.
Physical inspection and safe assembly
Work through the vendor unpacking guide in order. Typical first-run jobs, using Prusa as an example of a current assembled CoreXY, include mounting a panel that was removed for shipping, seating the print sheet on the locking pins, and cleaning the sheet before heat (Prusa Knowledge Base). Kit machines add frame squaring and belt routing — follow the kit manual, not a generic “i3” blog.
Safety, not folklore:
- Confirm mains voltage and grounded outlet match the PSU label.
- Keep the area clear of filament spaghetti and paper near hotends and beds.
- Do not skip the vendor’s self-test. Load-cell, PINDA/SuperPINDA, or Bambu lidar/calibration flows are not interchangeable.
Mechanical sanity checks: belts, wheels/rails, frame, fasteners
Before the first print, move each axis by hand (power off, if the manual allows). You want smooth travel, no grinding, no loose idler that walks.
| Check | What “good” feels like | If it fails |
|---|---|---|
| Belts | Even tension side-to-side; teeth seated on pulleys | Re-seat, then tension per the manual — do not guess a note on a frequency app unless the vendor says to |
| Wheels / eccentric nuts | No slop, no binding at the ends of travel | Nudge the eccentric; over-tight wheels flat-spot |
| Rails | Carriage slides without grit | Do not dunk in solvent; wipe and use the grease the rail vendor specifies |
| Fasteners | Frame corners do not rack when you press diagonally | Tighten in the order the kit uses |
Voron and other rail-heavy kits document this as a build step, not a later “upgrade” (Voron documentation).
Firmware, bed leveling/mesh, Z offset, and why names vary by printer
Three different systems, one job: the nozzle’s first-layer height is correct across the bed.
- Vendor UI (Bambu, Prusa, Creality “auto level”): run the stock mesh / load-cell / lidar routine. Then watch the first layer; auto-level is not a substitute for a visual check.
- Marlin: bed leveling + babystep Z (
M290) while the first layer goes down; store withM500when happy. - Klipper: rough probe or endstop calibration (
PROBE_CALIBRATEorZ_ENDSTOP_CALIBRATE), then liveSET_GCODE_OFFSET, thenZ_OFFSET_APPLY_PROBEorZ_OFFSET_APPLY_ENDSTOPandSAVE_CONFIG(Ellis: First Layer Squish).
Ellis’s practical tell: on a smooth sheet you should still see extrusion lines; gaps between lines mean not enough squish; a glassy smear with vanished lines is too much.
Names you will hear that are not the same thing: Z offset, probe z_offset, position_endstop, babystep, “live Z”, mesh fade. Changing the wrong one is why people “level for an hour” and still fail.
Loading a forgiving baseline filament
For week one, use a dry, brand-name PLA (or the roll in the box). Skip TPU, carbon-filled, and bargain mystery silk until you can reprint a cube cleanly.
Load until you get a consistent purge. If the printer has a filament sensor wizard, finish it. If the filament is wet (pops, steam, rough extrusion), dry it per the spool vendor — do not “just turn the nozzle up.”
When you are ready to pick the next material, use the PLA vs PETG vs ABS comparison instead of mixing materials and settings in the same evening.
Choosing a known-good first test model
Do not start with a 12-hour dragon. Use a small, well-sliced object with a decent first-layer footprint:
- A first-layer patch / square set (Ellis publishes these in the Print Tuning Guide test-print folder).
- 3D Benchy once the first layer is trustworthy — it exposes overhangs, bridging, and cooling, not just adhesion.
- A 20 mm cube if you only want to see layer consistency.
Slice with the vendor profile for that printer + PLA. Do not import a random “fast” profile from a Facebook group.
Reading the first-layer result
Stand there for the first layer. Cancel early if it is a bird’s nest.
| What you see | Likely meaning | First change (only one) |
|---|---|---|
| Lines with tiny gaps, easy peel | Not enough squish | Lower Z a little (more negative offset on most machines) |
| Ridges, no visible lines, elephant foot | Too much squish | Raise Z a little |
| One corner fails, rest OK | Mesh / twist / dirty sheet | Clean the sheet; re-run mesh; check the high corner |
| Nowhere sticks | Contamination or wrong sheet temp | Wash the sheet; confirm the filament profile’s bed temp |
| Spaghetti immediately | No extrusion | Heat soak, check path, reload |
Ellis’s photo set for good / too much / not enough squish is the best free visual reference in the community (First Layer Squish). Teaching Tech’s calibration site is the other standard walkthrough (Teaching Tech calibration).
If the first layer is good and later layers fail, stop “leveling” and go to troubleshooting.
The 10 beginner mistakes that create most self-inflicted problems
- Skipping the printer’s own unpacking / self-test.
- Printing on a dirty or finger-oiled sheet.
- Changing temperature, retraction, speed, and Z in the same print.
- Using a random slicer profile for a different nozzle or firmware.
- Starting with PETG/ABS/TPU because a video said it was “stronger.”
- Tightening belts or eccentrics until the motors skip.
- Ignoring a grind in the extruder (that is a hardware problem, not a “flow” problem).
- Leaving shipping foam under the bed or a clip on the rail.
- Chasing speed before a baseline cube exists.
- Never saving the working profile, then “tuning” from a dirty starting point.
Establishing a baseline profile before “optimizing”
When a first-layer square and a small cube both work:
- Name the profile (
Printer-PLA-baseline) and do not keep editing it in place. - Duplicate it before experiments.
- Write down nozzle, layer height, and filament brand on a note in the profile.
That baseline is what calibration and slicer tuning starts from. Optimizing a broken machine just encodes the damage.
What to do next: calibration, slicer tuning, material selection
- First layer still random → troubleshooting and FDM maintenance.
- First layer good, walls ugly → calibration workflow.
- You need heat or toughness PLA cannot give → PLA vs PETG vs ABS.