Mastering the Temp Tower: A Complete Guide

There is a familiar ritual for every 3D printing enthusiast: you buy a fresh roll of filament, load it up, and immediately wonder, "What temperature does this actually want?"
The manufacturer’s label usually gives you a wide range—something vague like "190°C – 220°C." That thirty-degree swing is the difference between a pristine, glossy print and a stringy, drooping mess.
The standard advice is simple: "Print a temp tower." But here is the catch—most people print the tower, look at it, pick the one that looks "prettiest," and move on.
If you are strictly printing Baby Yoda figurines, that method is fine. But if you want to understand how your printer actually behaves, or if you need parts that won’t snap under load, you need to go a step further. We are going to look at how to evaluate a temp tower for both aesthetics and layer adhesion.
This guide is designed for general-purpose calibration—finding the balance between aesthetics and strength. However, if you are printing functional parts where mechanical performance is the primary goal, you should read our deep-dive on Printing Functional Parts instead.
Step 1: The Setup (Don't Just Hit Print)
Before we melt any plastic, we need to ensure the test is valid. A common mistake is downloading a pre-sliced G-code file from the internet because it’s "easier."
While convenient, pre-sliced files use someone else's retraction settings, speeds, and cooling profiles. If their retraction is set to 6mm and you have a Direct Drive extruder, the test will fail—not because of the temperature, but because the retraction settings were wrong.
Ensure you have the following ready:
- STL File: A standard Temp Tower model.
- Slicer: Cura, PrusaSlicer, or OrcaSlicer.
- Filament: The new spool you want to calibrate.
- Pliers: For the destructive testing phase.
Setting Up the Slicer
Import the temp tower STL into your slicer. Most modern slicers (like Cura or OrcaSlicer) now have built-in calibration generators.
- Set the Range: Look at your filament spool. If it says 180-220°C, set your tower to cover that range.
- Increments: We prefer 5-degree increments. It offers enough resolution to see changes without making the tower 30 inches tall.
- Your Settings: Ensure the print speed, fan speed, and retraction settings match your standard profile. You want to test the temperature, not the other variables.
If you are manually setting up the script to change temperatures, always preview the G-code or watch the first layer transition. A common error is setting the script wrong, resulting in the entire tower printing at a single temperature. Watch your printer's screen during the print to verify the target temp actually changes as the nozzle moves up.
Step 2: The Beauty Contest (Visual Inspection)
Once the print is done, take it off the plate and put it under good lighting. This is the step everyone knows, but let's break down exactly what you are looking for.
Scan the tower from bottom to top. You are judging the "Cosmetic Quality."
- Bridging: Look at the underside of the bridges. As the temp gets hotter, the plastic stays molten longer, causing it to droop. The "cold" zones usually bridge best.
- Stringing: Look between the two posts. Higher temps usually equal more stringing because the material oozes out of the nozzle more easily during travel moves.
- Detail: Look at the small numbers or cones. Are they sharp? Or do they look like melted candle wax?
Write down the temperature that looks the best. For our example, let’s assume 190°C produced the cleanest bridges and zero stringing.

| Signs of Too Cold ❄️ | Signs of Too Hot 🔥 |
|---|---|
|
Weak Layer Adhesion (Delamination / Brittle) |
Sagging Overhangs (Drooping Bridges) |
|
Under-extrusion (Gaps in walls) |
Stringing & Oozing (Wisps between parts) |
|
Matte / Dull Finish (Can Be Favorable) |
Loss of Detail (Melted / Blobs) |
Step 3: The Strength Test (Destructive Testing)
This is the step most people skip, but it is critical. A print can look perfect at 190°C, but if the layers snap apart like a stack of crackers, it’s useless for anything other than a shelf ornament.
The Snap Test
We are going to verify Layer Adhesion.
- Start at the coldest zone (e.g., 180°C).
- Use your hands (or pliers) to try and snap the block off the tower.
- Look closely at how the plastic separated.
Analyze the break:
- Clean Snap (Delamination): If it snaps cleanly along the layer line, the plastic was too cold to bond. This is a fail.
- Jagged Break: If the plastic tears, stretches, or breaks across multiple layers, the bond is strong. This is a pass.
Work your way up the tower until the layers become difficult to separate. Write down the temperature where the strength became acceptable. For our example, let's assume at 210°C the parts became nearly impossible to break by hand.
A "dry" snap sound usually indicates the material is brittle or cooled too fast. A dull bending or stretching indicates the material is fully melted and bonded. For PLA, you want it stiff, but it shouldn't separate at the layer lines effortlessly.
Step 4: Finding Your "Goldilocks" Temperature
Now you have two data points:
- Best Visuals: 190°C
- Best Strength: 210°C
So, which one do you choose? This is where engineering judgment comes in. You need to decide based on the intent of your project.
Scenario A: The Display Piece (Visuals Win)
If you are printing a bust, a statue, or a non-load-bearing architectural model, go with 190°C. The improved overhangs and lack of stringing will save you hours of post-processing (sanding/painting). Strength doesn't matter if it just sits on a shelf.
Scenario B: The Functional Part (Strength Wins)
If you are printing a bracket, a tool holder, or a replacement part for your dishwasher, go with 210°C. It might look a little glossy, and you might have to trim a few strings with a hobby knife, but it won’t shatter when you apply force.
Scenario C: The Compromise (The Daily Driver)
For most general printing, you want a balance. You want it to look decent but hold together if dropped. In our example, 200°C would likely be the winner. It offers decent bridging from the lower-temp characteristics but enough heat to ensure the layers bond reliably.
Step 5: Document Results (The Pro Move)
This sounds boring, but it is the difference between a novice and a pro. Don't throw the tower away immediately.
Take a simple permanent marker and write the "Winner" temp on the spool itself, or log it in a spreadsheet. We prefer a simple database that tracks:
- Filament Brand/Type/Color
- Best "Pretty" Temp
- Best "Strong" Temp
- Retraction settings used
By taking 30 seconds to document this, the next time you load that "Galaxy Blue PLA" six months from now, you won't have to guess or run the test again. You can just look at your notes, slice, and print with confidence.
Don't rely on loose scraps of paper. We created a free Temp Tower Calibration Sheet to help you log your results for both cosmetic (visual) and functional (strength) tests.
- Don't trust the spool: Manufacturer ranges are suggestions, not rules. Every printer sensor reads slightly differently.
- Consistency is key: Use your own slicer settings (speed/retraction) for the test, not a generic G-code file.
- Looks aren't everything: A print can look perfect but have zero strength. Always perform destructive testing.
- Context matters: Choose lower temps for statues (cosmetic) and higher temps for brackets (strength).
- Write it down: Label your spools with your findings to save time later.
Do I need to run a temp tower for every new spool?
Ideally, yes. Even different colors from the same brand can behave differently due to the additives used in the pigment. However, if you stick to the same brand and type, you can usually use previous results as a very close starting point.
Why does my temp tower look bad at every temperature?
If the entire tower looks terrible (stringing everywhere or blobs), your issue might not be temperature. Check your filament for moisture (wet filament strings badly) or calibrate your retraction settings first.
Can I just use the temperature with the strongest layer adhesion?
You can, but be aware that printing too hot can cause "heat creep" (clogs) or sagging on detailed parts. It's usually best to print at the lowest temperature that still provides full strength.
Is 5-degree increments enough?
Yes. 3D printer thermistors generally have a variance of a degree or two anyway. trying to tune within 1 or 2 degrees is usually splitting hairs and won't yield noticeable differences for general printing.