Recovering from a stall/oscillation/rotor wash issues

while messing with gains and props etc, it seems to be not uncommon to get into an oscillation the results in a fall of the craft.

Whats the best procedure regain control if the your multiroter isn't self correcting tumbling out of the sky.

I recovered from one last night with a full vertical thrust, but thought I'd open the discussion.

First I'd like to say that if you are experimenting with the gains you shouldn't be flying higher than 6-12 feet, probably no higher than eye level. If you can't see the affects of your changes at a low altitude you certainly won't be able to see them way up in the air.

I have three in-flight propeller failures under my belt and in my opinion giving full throttle is the best thing you can do...assuming you are not near any people you may injure or property you can damage. Giving the flight controller as much power as it needs to make any corrections is the best thing you can do.

This brings me to a thought and might be something to think about in your gain settings that nobody has yet mentioned. When I originally setup my Naza-M I had read somewhere that the vertical gain should be 20% less than maximum. The reasoning was that if you were at full stick the Naza would still have reserve power to do whatever it thinks needs being done. I believe that this was a transmitter setting, not a Naza Assistant setting.

A good source for settings is here:
http://wiki.dji.com/en/index.php/Naza-M_Autopilot_Setting

Thanks for the insight.

Please let us know what you find out.

Ugh, total calamity!

Was downtown last night, the the parking lot adjacent to the News Paper. I had pretty must just reset my gains back to default, with a touch backed off the attitude. She was flying great, very stable and I just had a great run down by Baker park.

Then at about 200 meters, I was minding my own business (feeling really good about the stability and all) when.....it just freaked out again and tumbled all the way down to the pavement. Phantom is okay, the gimbal took the brunt of it, and broke in half, tearing a control ribbon in the process. :/

I'm thinking I may have been in decent at the time.

So here's my new theory. I'm probably running into instability with rotor wash on a decent and it's being compounded by the flight dynamics of the carbon props.

I since learned that I need to make sure I have some other movement besides vertical decent to reduce the instability.

I'm seriously considering being done with the CF props. Now I have to see if I can salvage my gimbal which is pretty depressing.

Found this thread at phantom Pilots that looks useful
http://www.phantompilots.com/viewtopic.php?f=19&t=7902

here was some advice

Sorry to hear about your crash. Sounds like you are not giving up though which is good.

How fast are you trying to descend? I have yet to see a multi rotor that does not get wonky when descending fast.

Well I couldn't really tell, and it could have even been a gust of wind that started it off, I dunno. I wasn't trying to decent fast, but from what I've read there is sounds like the 9443 props of the Phantom 2 are more prone to creating the problem, and I suspect the carbon fiber props may have been a factor as well for the actual continued oscillation.

None the less, the results we're the same, just the crazy flipping around until it hit the ground. I was trying everything I could think of to try to regain control, mostly trying to push for more vertical, but it sounds like what I should have tried was to try hold down on a forward motion pretty much anything but up or down

Jonas, are you using a Phantom transmitter?

Yep. Stock Phantom transmitter.

It's hard to understand your issue without seeing it in person. I have yet to see a multirotor that does not get wonky when descending quickly. It also seems the small ones get wonkier than the bigger ones.

I've found that flying down seems to be the fastest way. Check out the last minute of this video.

Yep I see what you mean by flying it down.

Descending into your own rotorwash is called "Settling with power", you need to fly forward out of it, your just falling with your own dropping column of air. Real helicopters will crash when more power is attempted, they reduce collective pitch and fly forward to recover. RC has a lot of power and lift and this won't happen until your loaded down heavy with FPV stuff.

Another name for this condition is called Vortex Ring State.

Here is a Phantom in VRS, good stuff at the end obviously, seems as though this model has an issue with it for some reason. Adding more throttle will just cause you to crash faster because you push more dirty air below right what your falling into. Hard to say what the right thing is to do, a collective pitch can keep rotor speed up and glide out while producing less lift for a few seconds. A multi needs to be powered to really do anything at all.

Yep that's exactly what happened. So I've been making a concerted effort to never just drop straight down and try flying it down, so far so good.

Last Sunday I also practiced recovering out of a stall, which like you say my error was trying to give vertical thrust, not left right forward or back. Or at least when I tried during the fall, it was not giving it enough time to take effect. Watching the video it looks like I might have been starting to recover, but didn't notice from the ground and switch to another position hence not really getting out of the VRS.

I think probably the CF 9443 props (more thrust ), the gimble load (more downward force) and it being a bit choppy (turbulet air) all probably exasperated the situation.

I'm hoping my VRS days all the way to the ground are behind me and when I do find myself in that situation I'll be able to recover, but hopefully just avoid it now that been schooled

Thanks guys.