Your slicer tells the extruder to push 100mm of filament. The extruder pushes 93mm instead. Every single dimension on your print is now 7% wrong — walls are thin, top layers have gaps, and you’re chasing “flow rate” in the slicer to compensate for a mechanical error. E-step calibration fixes this at the source: it makes the extruder actually deliver what the firmware tells it to deliver.
I’ve calibrated over 40 printers in the last five years. The process takes 10 minutes, requires a ruler and a marker, and fixes extrusion problems that slicer settings can’t compensate for.
Step-by-Step E-Step Calibration
Prerequisites
Before calibrating E-steps, verify two things:
-
The extruder is assembled correctly. The tension arm should grip filament firmly without deforming it. A loose tension arm causes variable slip that makes any calibration meaningless because the effective steps/mm changes from one test to the next. Tighten the tension screw until you can’t pull filament out by hand, but the extruder gear doesn’t leave deep bite marks.
-
The nozzle is clear and at printing temperature. A partially clogged nozzle creates back-pressure that the extruder has to fight. E-steps measured against a clogged nozzle will be artificially high — the extruder motor works harder and the calibration accounts for resistance that shouldn’t be there. Heat the nozzle to your primary filament’s printing temperature (200°C for PLA, 240°C for PETG) and purge 50mm to clear any residue.
The Measurement Method
Step 1: Mark the filament. Measure 120mm from the entry point of the extruder (where filament enters the housing). Make a mark with a fine-tip permanent marker. Then measure another mark at 150mm from the entry point. You now have two marks 30mm apart, with the second mark at 150mm.
Step 2: Extrude 100mm. In your printer’s control interface (LCD menu, OctoPrint, or terminal), send:
M83 ; relative extrusion mode
G1 E100 F100 ; extrude 100mm at 100mm/min
The extrusion speed of 100mm/min (F100) matters. At high speeds, the extruder motor can skip steps — you’d be measuring motor stall, not calibration error. At 100mm/min, the motor is well within its torque curve.
Step 3: Measure the remaining filament. After extrusion completes, measure from the extruder entry point to the 150mm mark. If the mark is now at 53mm from the entry, you extruded 97mm (150 – 53 = 97mm) instead of the commanded 100mm.
Step 4: Calculate the new E-steps value.
New E-steps = Current E-steps × (100 / Actual Extruded)
If your current E-steps is 93.0 and you extruded 97mm:
New E-steps = 93.0 × (100 / 97) = 93.0 × 1.0309 = 95.87
Read your current E-steps from the printer:
M503 ; report settings
Look for the line: M92 Exxx.xx — that’s your current E-steps value.
Step 5: Set and save the new value.
M92 E95.87 ; set new E-steps
M500 ; save to EEPROM
Step 6: Verify. Repeat steps 1-4 with the new E-steps value. The second measurement should be within ±1mm of 100mm extruded. If it’s off by more than 1mm, your extruder has a mechanical issue — inconsistent tension, a worn drive gear, or filament diameter variation that’s changing the effective grip radius.
Temperature Matters
E-steps calibration is temperature-dependent because the extruder motor’s torque varies with the filament’s melt viscosity. PLA at 200°C flows differently than PETG at 240°C, and the extruder back-pressure changes. For multi-material printers, calibrate E-steps at each material’s printing temperature and note the values:
| Material | Print Temperature | E-steps (example) | Notes |
|---|---|---|---|
| PLA | 200°C | 95.87 | Base calibration |
| PETG | 240°C | 95.87 | Usually same as PLA (same extruder mechanics) |
| TPU | 220°C | 96.20 | Slightly higher — TPU compresses in the extruder |
| ABS | 250°C | 95.87 | Same as PLA for most extruders |
| Nylon | 260°C | 95.40 | Slightly lower — nylon is stiffer prior to melting |
Most printers keep the same E-steps across materials because the extruder mechanics (gear diameter, motor steps) don’t change with temperature. The 0.5-1% variation for flexible materials comes from filament compression in the extruder path — set E-steps for your primary material and use slicer flow rate adjustments to compensate for material-specific behavior.
Fine-Tuning After E-Step Calibration
After E-steps are accurate, fine-tune extrusion with a single-wall cube test:
- Print a 20x20x20mm calibration cube with single perimeter walls (0% infill, 0 top layers, 1 wall line count)
- Measure the wall thickness with calipers at 3 points per wall (top, middle, bottom)
- If the wall measures 0.42mm and your slicer is set to 0.40mm extrusion width, reduce flow rate to 95% (0.40/0.42 ≈ 0.95)
- Reprint and verify the wall measures 0.40mm ±0.01mm
This compensates for filament diameter variation, die swell (filament expands after exiting the nozzle), and extrusion path geometry — factors that E-step calibration alone can’t address.
Parameter Comparison Table
| Extruder Type | Typical Steps/mm | Adjustment Sensitivity | Drift Over Time | Notes |
|---|---|---|---|---|
| Stock Creality (single gear) | 93.0 | Coarse (full steps) | Moderate — brass gear wears | Upgrade to dual-gear for consistency |
| BMG Clone (dual gear) | 415.0 | Fine (geared reduction) | Low — hardened gears | The standard for reliability |
| Bondtech LGX | 400.0 | Very fine (50:1 ratio) | Very low | Premium option, overkill for most users |
| Orbiter V2.0 | 690.0 | Very fine (7.5:1 ratio) | Very low | Lightweight direct drive champion |
| E3D Hemera | 398.0 | Fine (dual drive) | Low | Good torque, noisy at speed |
| Titan Aero | 420.0 | Fine (3:1 reduction) | Low | Older design, still widely used |
Higher steps/mm values mean more micro-steps per millimeter of filament, which gives finer control but can push some stepper drivers past their micro-step resolution limit. A BMG clone at 415 steps/mm with 16x micro-stepping moves 25,937 steps per millimeter — the TMC2209 driver handles this cleanly, but older A4988 drivers may have resolution issues above ~10,000 steps/mm.
Common Mistakes & What Most Users Get Wrong
1. Calibrating E-steps through the nozzle instead of into free air. E-steps should be measured with the filament exiting the extruder into free air (no nozzle, no hotend). The extruder’s job is to push filament, not to fight nozzle back-pressure. If you calibrate through a hot nozzle, you’re measuring extruder + hotend resistance — remove the Bowden tube or unscrew the nozzle, and calibrate the extruder alone. Then use flow rate in the slicer to account for the system’s extrusion characteristics.
Correction: most modern guides recommend calibrating through the hot nozzle because that’s the real operating condition. The “free air” method measures the extruder in isolation but the assembled system behaves differently. I calibrate through the hot nozzle at printing temperature — it measures the real system, and the slicer flow rate (not E-steps) handles any remaining discrepancy. Pick one method and be consistent.
2. Extruding too fast during calibration. At 300mm/min or faster, most Bowden extruders skip steps intermittently, and the measured extrusion length becomes inconsistent. Use F100 (100mm/min) for reliable results. The calibration takes 60 seconds instead of 20 — accept the extra time.
3. Not saving to EEPROM after M92. The M92 command changes the value in RAM only. If you power-cycle the printer without issuing M500, the calibration is lost and your prints revert to the old underextrusion. I’ve done this three times — the third time, I taped a “SAVE EEPROM” note to my printer’s control box.
4. Using a worn extruder gear for calibration. A brass extruder gear with 500+ hours of use has a groove worn into the teeth profile. The effective grip diameter is smaller than a new gear, and the steps/mm value that works with the worn groove will over-extrude when you eventually replace the gear. If your extruder gear shows visible wear (groove where filament sits, dull teeth edges), replace it before calibrating.
5. Calibrating E-steps and never rechecking. A Bowden tube that’s worked slightly loose from its fitting, a spring that’s lost tension, or a gear that’s accumulated filament dust — these shift the effective extrusion by 2-5% over weeks. Recheck E-steps every 3 months or after any extruder maintenance.
⚠️ Safety Notice: When performing extrusion calibration, ensure the hotend is at operating temperature. Never force filament through a cold hotend — it can damage the extruder gears or shear the filament inside the heat break. Keep hands clear of moving extruder gears during calibration. The stepper motor has enough torque to pinch fingers. Ensure your printer is on a stable, fire-resistant surface during all heated operations. In 2026, thermal runaway protection should be enabled in firmware (M143 S275 for hotend) — verify this with M503 before starting any calibration procedure.
For the full tuning workflow that follows E-step calibration, our 3D Printer First Layer Calibration Guide covers the bed-side adjustment that pairs with accurate extrusion. The extrusion mechanics also relate to our Direct Drive Extruder Conversion Guide — Bowden vs direct drive changes your E-steps significantly. And if you’re upgrading to a dual-gear extruder as recommended in the table above, our Dual Gear Extruder Upgrade Guide has the installation steps.
Accurate extrusion starts with an extruder that grips filament consistently. The Trianglelab BMG clone dual-gear extruder uses hardened steel drive gears with a 3:1 reduction ratio — that means 415 steps/mm, which gives you roughly 4x the control resolution of a stock single-gear extruder at 93 steps/mm. The dual-gear design grips from both sides so the filament doesn’t slip even with TPU, and the hardened gears won’t wear into a groove after 500 hours like brass alternatives.