Use Actual, Quick, and KISS Rates in FPVSIM
Compare and configure the Actual, Quick, and KISS rate systems available in FPVSIM.
Read a step-response graph and use it to find a practical balance between P and D gains for your FPV quad.
What this tutorial covers
This guide uses FPVSIM Blackbox Analyzer's step-response view to find a practical P-to-D balance. The goal is to make roll, pitch, and yaw follow commanded movement promptly without excessive overshoot or oscillation.
Tune conservatively and verify current firmware behavior. The video demonstrates an older Betaflight workflow. Remove propellers while configuring on the bench, make small changes, watch motor and component temperatures, and use the names and ranges documented for your installed firmware.
Prepare a useful log
- Use a mechanically sound quad with secure hardware, healthy motors, and undamaged propellers.
- Back up the current configuration so every change can be reversed.
- Configure Blackbox logging at 1 kHz as demonstrated, if the flight controller and storage support it reliably.
- Choose a safe open area and avoid testing near people, property, or obstacles.
Establish the P-to-D balance
Isolate the feedback response
In the video's Betaflight setup, Feedforward and Absolute Control are set to zero, D-MIN is disabled, I is kept around 30 to 40, and TPA is set to 1.0. Current firmware may expose different controls, so verify each setting before changing it.
Choose a conservative starting point
The tutorial starts P and D around 35 to 45 on a typical five-inch quad, or closer to 30 for a rougher build. Treat these as historical examples, not universal values.
Record a short test flight
Use the demonstrated basement-style input sequence: hover steadily and make several quick roll or stick movements that give the analyzer distinct command steps. A flight of roughly one minute can provide enough clean examples.
Open Step Response Overview
Put the flight controller into storage access mode, open FPVSIM Blackbox Analyzer, load the log, and select Step Response Overview. Inspect roll, pitch, and yaw separately.
Read the response curve
The horizontal axis is response time in milliseconds. The vertical axis compares the commanded step with the quad's measured response. A large peak past the target indicates overshoot; repeated ringing suggests oscillation; a slow rise indicates a sluggish response.
Adjust P in small increments
If the response is too slow, raise P by a small amount such as the five-point change demonstrated. If it overshoots or rings, reduce P or revisit the relationship with D. Change one axis or one parameter group at a time.
Find the upper boundary carefully
After the P-to-D ratio looks balanced, the tutorial raises P and D together to explore overall strength. Stop and back down at the first sign of sustained oscillation, hot motors, noise amplification, or unstable flight.
Re-enable the full tune deliberately
Restore any feedforward, D-MIN, TPA, or other features you disabled for the test, then log and verify the complete configuration instead of assuming the isolated result remains correct.
What to look for
- Compare multiple clean step events; a single bump or prop-wash event can distort the average.
- Tune roll and pitch independently when their mechanics or response shapes differ.
- Do not chase a perfectly shaped graph at the expense of cool motors and predictable real-world flight.
- Keep the original log and configuration beside each revision so you can identify whether a change actually helped.
Edited transcript
This transcript has been cleaned for readability. Parameter names and example values reflect the firmware and FPVSIM interface shown in the video.
00:00–00:55
This tutorial demonstrates a simple way to find a good P-and-D balance with FPVSIM Blackbox Analyzer. Configure Blackbox for a 1 kHz logging rate so the step-response calculation has enough samples.
00:56–02:02
Start with conservative P and D values. The example uses approximately 35 to 45 for a regular five-inch quad, or around 30 when the build is not performing well. The exact number matters less than beginning from a stable baseline.
02:03–03:18
To focus on the basic feedback response, the demonstration sets Feedforward and Absolute Control to zero, turns D-MIN off, keeps I around 30 to 40, and sets TPA to 1.0. These are temporary test conditions for the workflow shown.
03:19–04:23
Fly for about a minute using a series of quick, distinct stick inputs while maintaining a safe hover. The method is based on Brian White's basement tuning approach and is meant to produce recognizable response steps in the log.
04:24–05:17
Connect the flight controller in storage mode, open FPVSIM Blackbox Analyzer, load the log, and choose Step Response Overview. The page provides separate plots for roll, pitch, and yaw.
05:18–06:31
Time in milliseconds is shown along the horizontal axis, and the measured response is shown against the requested step vertically. Too much P relative to D tends to overshoot and may ring. Too little produces a slower, softer response.
06:32–07:42
Adjust P in small steps, such as five points, then repeat the flight and analysis. Compare the curve rather than making several unrelated changes at once.
07:43–08:40
Once the shape shows a reasonable P-to-D balance, increase P and D together only as far as the build remains clean and stable. The onset of oscillation marks a boundary; back away from it and confirm the result in another log.
Frequently asked questions
Can I copy the P and D values from the video?
No. They are starting examples for the demonstrated quad and older firmware. Frame stiffness, motors, props, filtering, firmware, and payload all affect the safe tune.
What does overshoot look like?
The measured response rises beyond the commanded level before settling. Repeated movement above and below the target is ringing or oscillation.
Why disable Feedforward and D-MIN for this test?
The demonstrated workflow temporarily removes features that can obscure the basic P-and-D feedback relationship. Restore and verify the complete configuration after that relationship is established.
Should every axis use the same values?
Not necessarily. Roll, pitch, and yaw can have different inertia, mechanics, and response curves, so inspect each axis on its own.
When should I stop increasing gain?
Stop at any sustained oscillation, unexpected noise, hot motors, or unstable behavior. Back down to a conservative setting and confirm it with another safe flight and log.
