Got bitten by AQ

[leopold1]

Last week I was investigating small oscillations with DIMU installed I wanted to check that it wasn't hw related so I switched back to the official fw without DIMU from the download section. After removing the DIMU, programming the FC went allright and I restored my parameter file and motormix. When I wanted to arm the quad, nothing happened. So I set the tx aside and got closer to the quad to check for possible connector problems. Green led was blinking. Then suddenly the led didn't blink anymore and about 5 seconds later the motors started full power and tilted in my direction. I jumped right up and the props hit my lower legs and knees. The quad was upside down and I put my foot on it in order to disconnect the battery. Luckily for me I only got cut a few times and not to deep. Nevertheless it was a frighting experience.
After cleaning my self up I wanted to know what went wrong.
I uninstalled all the props and hooked it up to the pc. Without the TX on everything was fine. But when I switched the tx on, about 15 sec later the quas armed itself and 5 seconds later the motors started turning. Then I noticed that the radio type was set to SUMD whilst I'm usung Sbus. This was the problem.

Sometimes I have the impression that when I want to change fw, parameters, motor mix, .... not all changes are uploaded. Both with wireless and usb connection.
Something I noticed to strengthen this impression is when I upload the latest DIMU version with GCS 1.4 all goes wel but I have to restore the old pids (devided by 4.82). This allready happened twice and I discussed it with Menno but couldn't quite lay his finger on it.

Conclusion for me is to double checked changed parameters! And don't take for granted fw changes.

Now for the positive side of things. I started out tinckering with multirotor bout 3 years ago starting out with APM and the pirate code. I learned al lot from that time and had a lot of broken props, sometimes bent armes and then and now a fly away.
Last year I got my hands on an AQ board and have to say that since then I never have damaged a props nor frame due to fw problems. (only due to piloting error )
So keep up the great work and stay safe.

[kinderkram]

Dammit!
That´s why I put a global announcement here to warn everyone:
viewtopic.php?f=24&t=3522

Experimental and beta fw always needs extra caution, especially when changing hardware or the like.

Fortunately you weren´t hit in the face. And good to know you´re still learning from the project, though it shouldn´t have been the hard way.

Any proposals how the team can manage to avoid such incidents in the future? A newsletter maybe?

[leopold1]

Norbert,

In the end it was my fault not checking the important parameters like radio type. Even though I would never have thought that the use of a different radio type would have such a dangerous impact.
And even with a different approach of arming the ESC's it still would of happened.
But my lessons learned and will read important announcements

[Max]

I'm very sorry this happened. Changing the meaning of the radio type parameter should have been avoided in the first place, IMHO. I'm glad you're OK and it wasn't worse, but that doesn't make it alright. Your point of checking vital parameters before flight is of course valid and important, but this kind of dangerous mixup could/should have been avoided entirely in the firmware to begin with. Hopefully we can remedy this in the near future.

-Max

[leopold1]

Thats correct. I'm a little bit worried about the parameters not validating even when changing fw.
I'm now flying with r406 with DIMU but with pid setting from the official fw release.

This is the second time in a row. I'm allmost certain that when I would reload the fw again I would be able to fly with the new pids. So something really isen't right.

[kinderkram]

Sorry, Vincent - my last post didn´t sound like an appology - but we owe you one!

Max has removed the affected fw versions and we hope no one else will stumble upon this...

We always try to play safe and except for one incident where we had to revoke a firmware, we did. Until now.
We thought an announcement would be enough to warn those using experimental fw. It obviously wasn´t.

We simply didn´t know what happens when the wrong radio is selected. Now we know.

Sorry again,

Norbert