Been flying FPV drones for about 2 years now. Love the flying part, but PID tuning? That's always been the part I dread.
You know the drill - spend a weekend doing test flights, recording blackbox logs, staring at PIDtoolbox graphs trying to figure out what to change. Tweak P a bit, test again. Still got prop wash? Maybe try D term filtering. Rinse and repeat until you give up or get lucky.
The Problem
PIDtoolbox is great for analyzing logs, but it doesn't tell you what to actually change. You still need to understand all the theory - step response, noise floors, filter delays. Most of us just want to fly, not become control systems engineers.
What I Built
So I built FPVtune - a tool that reads your blackbox logs and spits out PID suggestions automatically. It uses neural networks trained on thousands of flight logs to figure out what changes would improve your tune.
The workflow is dead simple:
- Upload your blackbox log
- Wait ~30 seconds for analysis
- Get recommended PID values
That's it. No graphs to interpret, no theory to understand.
How It Works (for the curious)
The core is a neural network that learned patterns from analyzing blackbox data. It looks at things like:
- Gyro noise characteristics
- Motor response times
- Current PID behavior
- Oscillation patterns
Then it predicts what PID changes would reduce issues like prop wash, oscillations, and bounce-back.
The whole thing is open source on GitHub: github.com/chugzb/betaflight-pid-autotuning
Results
Tested it on my 5" freestyle build that had annoying prop wash on throttle chops. The suggested PIDs reduced it noticeably on the first try. Not perfect, but way better than my hours of manual tuning got me.
Try It
The tool costs $9.90 (one-time), but I have a beta code if anyone wants to try it free:
Code: FPVTUNE-BETA-2026
Just go to fpvtune.com, upload a log, and enter the code on the payment page.
Curious if anyone else has tried auto-tuning approaches? Would love to hear what's worked (or not worked) for others.
Top comments (0)