Tuning
The order of operations matters more than the numbers. Filters first, then I, then P, then D, and the throttle curve last — change two at once and you will not know which one helped.
The order that works
Fix the mechanical problems first: balance the props, check the motor bells and soft-mount anything that carries a sensor or a camera. Then set the filters (a gyro lowpass roughly at half the frame resonance, the dynamic notch on, RPM filter on with bidirectional DShot). Then set the rates, because they change how the aircraft feels far more than the PID does. Only then tune I, then P, then D. Throttle curve is the last thing you touch on a long-range build and the first thing you touch on a ducted one.
Symptom to term
The table below is how to read what the aircraft is telling you. Almost every symptom has one term that fixes it and three that make it worse.
| Symptom | Change this | Not this | Why |
|---|---|---|---|
| Nose drops when you cut throttle | Raise I | Raise D | D damps oscillation, it does not hold attitude against gravity |
| High-frequency shake at full throttle | Lower D | Lower P | That is D amplifying motor noise through the frame |
| Slow wobble after a hard stop | Raise P | Lower I | The aircraft is not returning to attitude fast enough |
| Motors hot after three packs | Lower D, check filters | Lower P | D is where most of the electrical noise ends up |
| Camera shake only in fast descents | Fix the mount | Change PID | The mount has a resonance the tune cannot damp |
| Yaw drifts in a hover | Check motor bells and frame | Change yaw P | A bent bell is a mechanical fault, not a tuning one |
| Aircraft is twitchy around centre stick | Raise expo | Lower P | Expo is the correct tool for centre-stick feel |
| Oscillation that grows instead of fading | Lower P and D together | Raise any term | A growing oscillation is always too much gain |
Rates decide more than PID
Two aircraft with identical PID values and different rates feel like different machines. Actual rates at 780 deg/s with 0.58 expo puts 110 deg/s on the first 20 percent of stick travel; the same 780 deg/s with 0.3 expo puts 220 deg/s there. Cinematic builds want the high expo, racers want the low one, and freestyle sits around 0.5 to 0.6. Change the rates before you change a single PID value.
Blackbox before opinions
A blackbox log turns a debate into a measurement. Look at the gyro trace first: if the oscillation is visible in the gyro at a frequency above 150 Hz it is noise, not tuning, and filters or mechanical work will fix it. If it is between 20 and 80 Hz it is almost always a PID term or a frame resonance. Log a 30 second flight with a full-throttle punch and a hard stop in it; that is enough to diagnose most problems.
Filter settings on file
| Airframe | Filters | PID profile | Tune sheet |
|---|---|---|---|
| 5in-freestyle-baseline | PT1 120 Hz / PT1 110 Hz / Biquad 90-300 Hz | 48 / 82 / 42 | Actual 780 deg/s · expo 0.58 |
| 7in-longrange-cruiser | PT1 90 Hz / PT1 85 Hz / Biquad 70-220 Hz | 42 / 96 / 32 | Actual 480 deg/s · expo 0.72 |
| 3in-cinewhoop-ducted | PT1 150 Hz / PT1 140 Hz / Biquad 120-400 Hz | 62 / 118 / 28 | Actual 420 deg/s · expo 0.66 |
| 65mm-whoop-indoor | PT1 200 Hz / PT1 190 Hz / Off | 78 / 140 / 34 | Actual 600 deg/s · expo 0.4 |
| 5in-racing-lowslung | PT1 160 Hz / PT1 150 Hz / Biquad 140-500 Hz | 58 / 90 / 32 | Actual 950 deg/s · expo 0.28 |
| 4in-sub250-longrange | PT1 130 Hz / PT1 120 Hz / Biquad 100-320 Hz | 52 / 104 / 30 | Actual 520 deg/s · expo 0.65 |