Nozzle temperature is the setting most likely to wreck a print if it is wrong, and the one most people set once from the spool label and never touch again. A temperature tower gives you a faster way to check that number than printing the same object five times at five temperatures. You print one object, once, and every level tells you something different, which pairs well with the calibration order laid out in this site’s calibration order guide and with the material-specific starting points in the PLA settings guide.
What a temperature tower actually tests
The model is a narrow tower split into blocks, usually 5-10 mm each, stacked on top of one another. As the print climbs, the slicer inserts a temperature change at the start of each block, so the bottom of the tower might print at 220 degrees and the top at 195. Everything else about the print stays fixed: same speed, same fan curve, same retraction. That is the whole point. If only temperature changes between blocks, then whatever changes in the print can be attributed to temperature and nothing else.
Most towers also carry small overhangs, a bridge across a gap, and sometimes lettering on each block, because a plain cube would only show you stringing and not much else. Look at the bridge to judge sag, the overhang to judge how well the plastic holds its shape, and the vertical walls to judge stringing and surface texture.
Setting up the test in your slicer
OrcaSlicer, PrusaSlicer, Bambu Studio and Cura all ship a built-in temperature tower object or a downloadable one, so you rarely need to hand-edit g-code anymore. Pick a range that brackets what the filament maker printed on the spool. If the label says 190-220, do not test 150-250; you are wasting plastic on temperatures nobody would ever run. A 5 degree step between blocks is the common default across these tools and gives enough blocks to see a trend without making the tower absurdly tall.
Before you print, make sure the rest of your profile is one you would actually use, not a generic default. Match your usual speed, layer height, cooling and retraction settings, because a temperature tower run at a strange print speed is a test of that speed and temperature together, and the result will not transfer to your real prints. Dry filament matters here too. PETG and nylon in particular will string and bubble from moisture in ways that look like a temperature problem when it is really a drying problem.
Reading the results block by block
Work from the top of the tower to the bottom, since the top is usually printed at the lowest temperature in most slicer presets (check yours, some run the other direction). At the cold end you are watching for under-extrusion: gaps between wall lines, a rough or grainy finish, and poor adhesion between layers that you can flex apart with your fingers. At the hot end you are watching for the opposite: stringing between features, sagging overhangs, a glossy or slightly blobby surface, and dimensional overhang problems where the corners droop.
Somewhere in the middle there is usually a run of two or three blocks that all look acceptable. Pick a spot inside that range rather than the single best-looking block, since block-to-block differences of only a degree or two are within the noise of home printer thermistors anyway. If you plan to print faster than your test or push higher flow rates later, lean toward the warmer end of the good range, since faster extrusion needs more heat to keep up with the flow. If stringing matters more to you than raw strength, for a display piece rather than a functional part, lean cooler instead.
Why the ideal number depends on more than the material label
Two spools of the same brand and color can print differently, and colorants, especially some blacks and blues, absorb heat and behave slightly differently at the same set temperature. Your hotend’s thermistor placement and heater cartridge wattage also affect how close the reported temperature is to the plastic’s actual temperature at the nozzle tip. This is part of why manufacturer temperature ranges are given as a spread and not a single number, and why a tower run on your own printer beats copying someone else’s number from a forum post. If you are calibrating a material this site already covers in detail, like ABS or TPU, use that guide’s range as your starting bracket and the tower to narrow it down for your specific spool.
When to run the test again
A temperature tower is not a one-time chore for the printer, it is a one-time chore per filament. New spool, new color, new brand, or a nozzle swap to a different material (brass to hardened steel, for instance) are all good reasons to rerun it. If you start seeing stringing or weak layers on a filament that used to print cleanly, check that the spool has not absorbed moisture before assuming the temperature has drifted, since a damp spool will mimic a too-hot printer even at a setting that worked fine before.
Checklist
- Bracket the range printed on the spool rather than guessing a wider window.
- Keep speed, cooling and retraction identical to your real print profile so temperature is the only variable.
- Check the bridge and overhang features on each block, not just the vertical walls.
- Pick the middle of a run of good-looking blocks, not a single best block.
- Favor the warmer end of the good range for speed or strength, the cooler end for cosmetics.
- Rerun the test for every new spool, color or nozzle material change.