You plug in, arm, and the quad wobbles like it is drunk. Or it flies fine for two minutes then spins out of the sky. Or the video feed goes to snow 100 meters out. Every FPV pilot hits these problems. The difference between a grounded quad and a flying one is knowing how to diagnose symptoms quickly and fix the right thing on the first try. This guide maps every common failure mode to its most likely causes — and exactly what to do about each one.
Symptom-to-Cause Diagnostic Table
| Symptom | Most Likely Cause | First Thing to Check |
|---|---|---|
| Fast oscillations / jello in FPV feed | P-term too high; mechanical vibration | Reduce P by 10%; check for bent motor shaft |
| Slow wobble after flips / dives | Propwash — I-term too low or D-term too low | Increase D by 3–5 points; increase I by 5 |
| Motor screeches and stops mid-air | Motor desync | Increase motor timing; check demag compensation |
| Quad drops from sky, RX Loss OSD | Failsafe — link lost | Check antenna connection; verify LQ/RSSI |
| Horizontal lines in video under throttle | Electrical noise on VTX power rail | Add capacitor; check VTX ground point |
| Motors too hot to touch after 30 seconds | D-term noise; PWM frequency too high | Lower D; check filtering; reduce PWM to 48kHz |
| Quad flips on arm / death roll | Idle speed too low; ESC desync; wrong motor direction | Increase idle; check motor ordering and direction |
| Random twitching in level flight | Gyro noise; frame resonance; loose stack screws | Tighten all screws; check blackbox spectrogram |
Oscillations: Fast Shakes and Jello
If your quad has visible shaking in the FPV feed — especially at high throttle or during punchouts — you have oscillations. The first diagnostic step is determining whether it is electrical (tune) or mechanical (hardware).
P-Term Too High
The P-term corrects proportionally to error. Too much P gain, and the correction overshoots, causing the quad to oscillate around the setpoint. Signature: fast, rhythmic shakes that are present even in gentle hover. Fix: reduce P gain on the affected axis (usually roll and pitch) by 5–10 points. If oscillations stop, you have confirmed a tune issue.
Mechanical Noise
A bent motor shaft, chipped prop, or delaminated frame arm produces mechanical vibration that the gyro reads as attitude error. The PID loop fights it, producing oscillations that worsen with RPM. Signature: shaking that is worse at specific RPM ranges and changes when you swap props. Fix: spin each motor individually in the Betaflight Motors tab and watch the gyro trace. A motor with a bent shaft produces a clean sine wave at its RPM frequency. Replace the motor or shaft.
Loose Stack or Motor Screws
A flight stack that is not firmly mounted transmits frame vibration directly to the gyro. Signature: oscillations that appear suddenly after a crash and worsen with throttle. Fix: tighten all stack standoffs. Use soft-mount gummies if the frame has hard-mounted standoffs. Check motor screws — a single loose motor screw creates an imbalance that looks like a bent shaft.
Propwash Handling
Propwash is the wobble you get when the quad flies through its own turbulent air — after a dive, sharp turn, or hard stop. It is a low-frequency oscillation (5–15 Hz) that the PID loop struggles to correct because the disturbance is larger and slower than motor noise.
The fix is a combination of D-term and I-term tuning:
- D-term: D dampens rapid changes. Increase D on roll and pitch by 2–3 points at a time until propwash oscillations settle within 1–2 cycles. If motors come down hot, you have gone too far — back off D and accept slightly slower propwash recovery.
- I-term: I accumulates error over time. Higher I helps the quad “hold attitude” through propwash instead of getting pushed around. Increase I by 3–5 points on pitch first (propwash hits pitch hardest), then roll.
- TPA (Throttle PID Attenuation): TPA reduces P and D at high throttle where oscillations are most likely. Start TPA at 1500 (throttle point where attenuation begins) with 20–30% attenuation. This lowers PID authority at full throttle, giving the quad a looser feel that is less prone to oscillation when motor forces are highest.
Motor Desyncs
A motor desync is terrifying: one motor stops mid-air, the quad spins violently, and you disarm and watch it fall. The ESC loses synchronization with the motor’s rotor position and stops driving it. Causes and fixes:
- Demag compensation too low: Demag comp helps the ESC recover from back-EMF spikes that confuse rotor position sensing. Set to Low or Medium in the ESC configurator. High setting reduces power — avoid unless necessary.
- Motor timing too low: Timing determines how early the next phase fires. Auto timing (default on both BLHeli_32 and AM32) works for most setups. If desyncs persist, set fixed timing to 20–25°.
- Min/max throttle mismatch: If Betaflight’s idle throttle is lower than the ESC’s minimum startup power, the motor may stall on rapid throttle changes. Set dynamic idle in Betaflight to 25–35 (RPM × 100). This is higher than the old “motor idle percent” approach and far more reliable.
- ESC protocol errors: DShot is CRC-protected — if the ESC receives a corrupted packet, it holds the last valid command for a few milliseconds, then disarms. Poor ESC signal ground can cause packet corruption. Check that the ESC signal ground wire is connected. Never run DShot without a ground reference.
Failsafes: RX Loss and Link Quality
A failsafe happens when the flight controller stops receiving valid data from the receiver. In the OSD, you will see “RX LOSS” or “FAILSAFE.” The two sub-cases:
- True RX Loss: The radio link is gone. RSSI dBm drops below the receiver sensitivity threshold (typically -112 dBm for Crossfire, -117 dBm for ELRS). The receiver literally cannot hear the transmitter. Fix: check antenna placement (both antennas should be at 90° to each other, away from carbon fiber and battery), check that the antenna is the correct frequency (868/915 MHz for Crossfire, 2.4 GHz for ELRS), and verify TX power settings.
- Link Quality Drop: LQ (link quality) is separate from RSSI. You can have strong RSSI and terrible LQ if there is interference. LQ below 80% means the receiver is dropping packets. Fix: change frequency/channel, move away from noise sources (WiFi routers, other pilots), or increase TX power.
Configure your OSD to display both RSSI dBm and LQ. RSSI tells you signal strength; LQ tells you data integrity. A failsafe at -95 dBm with 90% LQ means a hardware issue (loose antenna connector), not a range problem.
Video Noise
Clean video is non-negotiable. Noise in the feed is either electrical (from the power system) or RF (from the VTX environment).
Electrical Noise: Horizontal Lines
Horizontal bands that change with throttle are electrical noise from the ESCs coupling into the VTX power rail. Fixes in order of effectiveness:
- Add a low-ESR capacitor: Solder a 35V 470–1000µF capacitor across the battery pads. This is the single most effective fix for 80% of video noise problems. Panasonic FM or Rubycon ZLH series are recommended.
- LC filter: If the capacitor alone does not fix it, add an LC filter between the battery pads and the VTX power input. Many 4-in-1 ESCs have a filtered 9V/12V output — use it for the VTX instead of direct battery voltage.
- Ground loop diagnosis: If video noise changes when you touch the VTX antenna or frame, you have a ground loop. Ensure the camera, VTX, and FC share a common ground point. Do not ground the VTX to the frame — ground it to the FC’s VTX ground pad.
RF Noise: Snow and Breakup
If video degrades with distance but not with throttle, it is an RF problem: wrong antenna polarization, damaged antenna, or VTX overheating. Verify the VTX antenna matches the receiver antenna (both RHCP or both LHCP). Check that the VTX antenna’s SMA connector center pin is not pushed in. Replace any antenna that has been in a crash — internal damage is invisible.
Hot Motors
Motors that are too hot to hold after a flight are wasting energy as heat instead of thrust. The primary causes:
- D-term too high or too noisy: D amplifies gyro noise. If your D gains are high and your filtering is light, the motors receive high-frequency commands that sound like buzzing. The motor tries to follow these commands, generating heat without producing useful thrust. Solution: lower D by 5 points and enable RPM filtering if not already active.
- PWM frequency: Higher PWM frequency (48–96 kHz) runs the MOSFETs harder, generating more heat in the ESC and motor. Drop to 48 kHz or even 24 kHz if motors are consistently hot.
- Filtering too aggressive: Paradoxically, too much filtering can increase motor heat because the filter delay causes the PID loop to fight old data. Raise the gyro lowpass cutoff slightly (from 200 to 250 Hz) if RPM filtering is handling motor noise.
Death Rolls / Flip of Death
The quad arms, you raise throttle slightly, and it instantly flips over and eats dirt. This is almost always one of three things:
- Motor direction or ordering wrong: In Betaflight Motors tab, spin each motor individually. Verify that motor 1 is front-right, motor 2 is rear-right, motor 3 is rear-left, motor 4 is front-left (standard configuration). Verify each motor spins in the direction shown on the diagram. Use the “Motor Direction” wizard in Betaflight 4.4+ to set this up correctly.
- Idle speed too low: If idle throttle is set too low (old-style “Motor Idle” at 4–5%), the motors may stall when Airmode activates at zero throttle. Use dynamic idle (RPM-based) at 25–35 instead of a percentage. This keeps the motors spinning fast enough to maintain authority.
- ESC protocol mismatch: All four ESCs must run the same protocol. If one ESC is on DShot300 and the others on DShot600, the odd one out may fail to arm. Check in the ESC configurator.
Twitching and Random Glitches
Small, random twitches in level flight that do not look like oscillations are usually a noise problem reaching the gyro:
- Frame resonance: Every frame has a resonant frequency where it vibrates like a tuning fork. If this frequency falls in the gyro passband, the FC sees phantom movement. Use the blackbox spectrogram to identify the resonant peak, then place a static notch filter at that exact frequency.
- Loose screws: A loose stack screw, motor screw, or arm screw introduces high-frequency vibration that the filters cannot fully reject. Tighten everything — use threadlocker (Loctite 242) on motor screws.
- Damaged gyro: If twitching persists after all mechanical and filtering fixes, the gyro IC itself may be damaged. This happens after hard crashes. The only fix is a new FC.
Keep this guide bookmarked. The next time your quad does something wrong mid-session, check the symptom table, try the most likely fix, and get back in the air instead of packing up and going home.