Traditional heli violent oscillations in alt hold mode

Hi all,

I had just gotten my traditional heli running missions when I dumb-thumbed it and disarmed mid flight instead of swiching modes (ch8 switch was right next to my mode switches) causing the heli to fall about 15m - I had enough control to get it on it’s skids but I destroyed the boom and chewed out all the helical gears that drive the tail. I fixed the major damage done to the heli yesterday and took it for a test flight about an hour after the crash. I ran it successfully in alt hold, loiter and did a quick mission with auto take-off and land without issue.
After that I rebuilt the head and re-calibrated the mag and compass (plates on the bottom of the bird where the pixhawk is mounted were cracked and everything got moved because I replaced them.). Now the heli flys the best it ever has in stabalise (still fine-tuning loops) but when I try alt-hold I get violent up-down oscillations. I’ve been playing with the throttle PIs without any success except when they are far enough out that the heli just shoots meters up or down (I can throttle the heli hard without any oscillations in stabalise mode too)/

I haven’t been able to find any similar issues discussed so I just assumed that it must be bad PI set-points but I guess it could be hardware? I loaded my log but it doesn’t seem to be the one that I wanted (cleared all after I grabbed it and it’s raining now -will post more data tomorrow - not sure what would be most useful?). Is altitude hold done solely off the GPS - it looks rock solid in the log that I have? -GPS is obviously not quite accurate enough to capture the oscilations I am taking about (probably only 5-10cm up and down - very rapid) so I’m guessing there is some other smarts that handle altitude control?

Setup:
Gauii X4ii
Pixhawk
3dr 915mHz mavlink radio
3dr mag and gps (this one: http://diydrones.com/profiles/blogs/3dr-gps-mag-back-in-stock)

Thanks for any help

Fisher92,
Alt Hold uses the Barometer I think. Have you tried playing with the THROTTLE ACCEL parameter. I still fly 3.2 so I’m not sure what the parameters are in 3.3. Post some logs. Its strange this problem started after replacing some parts on the Heli.

Regards,
David R. Boulanger

Thanks David,

I made some progress troubleshooting this today. It seems the issue is directly related to the throttle I set- prior to my crash I had it set at 85% but I re-calibrated the radio and forgot to reset this- when the issue started the throttle was at ~105%. I’ve been playing with this today and the maximum throttle I can set and get some alt stability is ~87% but it runs best at about 80%.

I’m guessing that there is some negative pressure right where I have mounted the pixhawk?? Not sure what the best way to beat this issue is - maybe I should mount the pixhawk on it’s side.

-Weird thing is that the baro data doesn’t look noise during oscillations (see attached). My pixhawk is mounted pretty much directly under the main shaft about 20cm below the rotors.

Any pointers appreciated.

pics

http://postimg.org/image/50f6fnxgn/
http://i66.tinypic.com/e9dvyb.jpg
http://postimg.org/image/qzhwlpdfh/