strange re-zeroing level pitch and roll signal

[herishi]

when flying today it was a little windy so I had to hold a little pitch forward to stay in the same place, however I noticed if I hel it in the same place for too long it was as if it zero'd out the position so I had to increase pitch forward, this went on untill I had to land because full pitch forward was only holding the quad level.

I guess there is a setting that I need to adjust but I dont want to go randomly changing things to try it.
I thought someone must have come across this before and know what setting to change

Any ideas?

Steve

[refractions]

Check the level indicators in the gui - pitch your quad by hand and hold it there - if the pitch indicator moves back to level, you need to check your sensor orientation. I had the same thing after I switched from V1.7 to V2.0 software.

Iain

[herishi]

Check the level indicators in the gui - pitch your quad by hand and hold it there - if the pitch indicator moves back to level, you need to check your sensor orientation. I had the same thing after I switched from V1.7 to V2.0 software.

Iain

I'll have to check that, though would that not mean I would have to keep adjusting the stick backwards and forwards, it would "re level" with the quad nose down so zero stick i/p would be as if it was moving forward, the problem I have is that even with pitch forward it seems to be level then zero pitch input would have the quad flying backwards.

It's really strange :-S

Steve

[refractions]

Yep - thats exactly what was happening to me Apparantly they changed the default sensor orientation in V2. I am using a BMA180 - all i needed to do was rotate it 90 degrees - took me a while to work it out - thought i was going mad.

[herishi]

Yep - thats exactly what was happening to me Apparantly they changed the default sensor orientation in V2. I am using a BMA180 - all i needed to do was rotate it 90 degrees - took me a while to work it out - thought i was going mad.

Cant wait to get home to try it now I'm using a WMP and Nunchuck but I guess the theroy is the same.

Steve

[herishi]

would be interesting to know what
/************************* INFLIGHT ACC Calibration *****************************/
/* This will activate the ACC-Inflight calibration if unchecked */
//#define INFLIGHT_ACC_CALIBRATION

and

/* Pseudo-derivative conrtroller for level mode (experimental)
Additional information: viewtopic.php?f=8&t=503 */
//#define LEVEL_PDF

actually do ?

Steve

[herishi]

Yep - thats exactly what was happening to me Apparantly they changed the default sensor orientation in V2. I am using a BMA180 - all i needed to do was rotate it 90 degrees - took me a while to work it out - thought i was going mad.

Cant wait to get home to try it now I'm using a WMP and Nunchuck but I guess the theroy is the same.

Steve

Well just tried it again the Nunchuck ACC is working as it should and getting correct responses on the conf tool I can prop it at an angle, go of and when I come back its still showing that angle, however when I fly if I hold pitch forward 20% for any length of time is assumes that is the stick neutral position (well thats how it seems) and to keep the same pitch forward I have to go to 40% pitch forward and if I were to go back to neutral position it would result in the quad being 20% pitch reverse.

when I landed and kept throttle just below takeoff thrust it was behaving strange too! I am uploading a youtube video and will post once uploaded

Any ideas???

Steve

[herishi]

Video of quad

http://www.youtube.com/watch?v=vzLayLecZBI&feature=youtu.be

[refractions]

Weird - nice looking quad btw. Have you got level mode enabled ? Mine did something similar before i tuned it.

[herishi]

Weird - nice looking quad btw. Have you got level mode enabled ? Mine did something similar before i tuned it.

thx its the DJI flame wheel frame

I had level mode on

Still got the feeling something is not right

Steve

[herishi]

update:

After a little more testing it seems the quad performs perfectly when not in level mode.
as soon as I flick into level mode it goes wrong, I think this may be down to my PID settings :-S (that black art) I'll post my settings tonight to see if anyone can notice the problem I seem to remember that the P is set to 10 and the D is set to 100 but I cant remember the I which thinking about it could be the one I need to adjust to fix this, if anyone is reading this and doesnt have this issue and you post your PID settings for level(I know each quad is different but I'm hoping to see a trend)

Steve

[refractions]

These are the default PID settings I started with.



Iain

[herishi]

thanks Iain are you still running v1.9?

[refractions]

Hi mate,

No - running latest dev version at the moment, but still started with the same PID levels.

Iain

[herishi]

v1.9 doesnt show the D value but I'll give it a look when I get in.

[herishi]

Iain

What Gyro and acc are you using? your feedback graph seem to peak a lot higher than mine the devices must be more sensitive (or you moved the quad loads)

Steve

[herishi]

I tried changing the P part of the PID on my quad today, tested from 5 to 10 for roll pitch and yaw and 5 to 11 for level, the higher P for level made things better until you tried to move the quad then it was a crash looking for and opportunity to happen

However the quad ranges from difficult to impossible to control when set in level mode, getting harder to keep steady the longer level mode is on.

Below are my PID settings.

Steve

[refractions]

Hi,

I usea WM+ and a BMA180. The image is actually from multiwiicopter.com. Its on their wiki page - my "motherboard" is their paris v4 board.

Iain

[herishi]

Guys

I thought my issue may be down to vibration and I will be making changes to this but if you look at the screenshot you can see the quad is landed and level but the telemetry shows it at a dodgy angle.

Full video at http://www.youtube.com/watch?v=HhMU8pXcmsg&feature=youtu.be