Read your dyno like a pro. Make changes that matter.
Everything you need to prep, run, and interpret dyno sessions the DMS way — so your graphs tell the truth and your tune gets faster, safer, and smoother.
1) What a dyno actually measures
Power (hp/kW) and Torque (Nm) are derived from the dyno’s roller/brake load and engine speed. Your graph is a story of airflow, fuel, and timing under controlled conditions.
- Power curve: How fast the engine does work (top-end breathing).
- Torque curve: How strongly it twists (mid-range response).
- AFR/Lambda: Combustion health under load.
- Boost/MAP: Air mass and control stability.
- IAT/ECT: Heat management and repeatability.
Key concept — repeatability
We care less about peak numbers and more about repeatable trends. If the setup repeats within a small margin across runs, we can trust changes in the tune or hardware.
2) Pre-dyno checklist
- Fresh fuel (correct octane), no leaks, correct oil level.
- Tire pressure set, wheel nuts torqued, no exposed threads/rubbing.
- Intake/IC piping secured, clamps tight, vacuum lines intact.
- Cooling system bled; radiator fans operational.
- ECU has no critical DTCs; CEL concerns addressed.
- Laptop + cable ready (logging works, power supply OK).
- Battery on charger/stabilizer if flashing on site.
- Baseline data ready (previous logs, mods list, tires/gear ratio).
- Agree on run ranges (e.g., 2,500 → rev limit) and gear.
3) DMS dyno methodology
Setup & strapping
- Strap at hub height, car centered, zero toe steer from straps.
- Front/rear straps tensioned evenly; suspension settled.
- Fans directed to radiator + intercooler; replicate road airflow.
Gear choice
Use the gear closest to 1:1 (MX-5 NC: typically 4th) unless speed limits dictate otherwise. Consistency beats ego-numbers.
Run procedure
- Stabilize temp at cruise load; start logs (AFR, MAP, IAT, trims, knock).
- Full sweep pull (e.g., 2,500 → 7,000 rpm) with steady throttle.
- Cooldown to equalize IAT/ECT; repeat for validation.
We keep smoothing moderate and correction consistent across runs for fair comparisons.
4) Reading the graphs
Power/Torque behavior
- Flat torque + rising power: Healthy turbo spool and top-end.
- Torque dip mid-range: Timing pulled, boost control soft, or knock/IAT.
- Power falls early: Backpressure, cam timing, or ignition ceiling.
AFR/Lambda
- NA: typically ~12.8–13.0:1 at WOT; Turbo: ~11.5–12.0:1 (fuel & hardware dependent).
- Wavy AFR: fuel system limits, MAF scaling, or O2 feedback boundaries.
Boost/MAP control
- Clean rise → slight taper is normal; oscillation = duty/gain mismatch or leaks.
- Early boost cut: check target vs limiters and solenoid polarity/duty cap.
Temps
- Rising IAT shifts knock threshold; expect timing trims.
- ECT creep = airflow/fan strategy or coolant issues; pause and cool.
5) Corrections, smoothing & comparability
Environmental correction
- Use the same SAE/DIN correction across sessions.
- Log ambient temp, pressure, humidity for transparency.
Smoothing
- Low smoothing exposes noise; high smoothing hides issues.
- Pick a middle value and keep it fixed when comparing.
6) Common issues & quick fixes
Oscillating boost
- Lower gain or adjust duty ramp; verify plumbing and check valve orientation.
- Wastegate preload too high/low → set baseline properly.
Fueling limits
- High IDC/leaning out → injectors/pump limits or voltage drop.
- MAF pegging → resize housing or switch to SD (where appropriate).
Heat soak
- Extend cool-downs; improve fan angle/CFM; shield turbo hot side.
- Compare first vs third pull to quantify soak effect.
Knock activity
- Revisit timing, AFR, IAT; inspect plugs, fuel quality, and mechanical noise sources.
7) MX-5 NC specifics (quick notes)
- 4th gear is typically closest to 1:1; confirm per diff/box.
- Stock airbox vs open filter can change MAF behavior on the rollers.
- Turbo NC: prioritize IC airflow (fan angle), not just radiator face.
- Cam swaps: expect VE shifts — re-map MAF/SD and revisit spark maps.
8) A clean run plan to copy
Session: 5 pulls total 1) Baseline warm pull (2,500 → rev limit) — log AFR/MAP/IAT/ECT/knock 2) Repeat for validation (same settings) 3) Make change #1 (e.g., WG duty +5%) → single pull 4) Cooldown (IAT within +2–3°C of first pull) → validation pull 5) Make change #2 (e.g., +1° mid timing if knock-free) → validation pull Keep correction/smoothing constant • Save graphs + raw logs • Note ambient
Save graph images and raw CSV logs for each stage. We prefer before/after overlays with identical axes limits.
9) Mini glossary
AFR / Lambda
Air-fuel ratio; Lambda = AFR ÷ 14.7 (for petrol). Use both for sanity checks.
IAT / ECT
Intake air / engine coolant temps; both shift knock limits and repeatability.
IDC
Injector duty cycle; sustained 95–100% indicates fueling headroom is gone.
10) FAQ
Do dyno numbers equal road performance?
Not always. Use the dyno for relative changes and shape, confirm on the road/track with real IAT/RAM effects.
Which smoothing value should I use?
Pick a middle value that doesn’t hide problems; keep it identical across comparisons.
How many pulls are enough?
Two matching baselines + one after each change, with cooldowns that keep IAT/ECT comparable.