I'm expanding the capability of the ship analysis tool to measure not only vibration (analysis of the ACC values) but also:
o determine mid-throttle for use in setup
o analyze the stability of the ship in hover (one of many indications of PID tuning, but a good place to start)
Right now the worksheet goes after Roll/Pitch/Yaw from the ATT records in the log. But should I be looking at the Gyr data in the IMU record instead?
Thought?
Once I have a some reference points and input, I'll release the worksheet for general use.
Replies
Not sure what stability index/score you will produce (neither how), but here my 2 cents:
In an ideal world, perfectly stable ship (perfect machine and no external influences like wind, magnetic fields, etc) , Roll/Pitch should remain equal to zero during hover, throttle should remain at 50% (for an ideally powered craft). Yaw would remain free because not involved in the stability process. Yaw should remain a fixed value when you let the yaw stick go.
So i guess measuring relatively (between ships /configurations) four parameters : Roll/Tilt/Yaw and throttle would be a good start to relatively compare the inherent stability between ships (assuming you'd be able to reproduce the exact same external conditions : not only wind but also magnetic fields ->hard to do except if you carry your experiments in a Faraday Cage big enough to hover a ship)
About throttle, ascent and descent of the ship can occur because of PIDs not rightly optimized ->difficult to decouple from pure throttle stability. Also, radios are not perfect-> if you look closely at PMW values, they vary a little even without touching the sticks. This is also a parameter difficult to isolate, except if we have a perfect ideal radio which does not exist.
I'm also wondering how you would decouple the effect of prop sets from one ship to another in your relative comparison because they will obviously also impact stability (or you have to keep perfectly identical sets of props and throw away any prop with little dents that would influence the comparison).