Tail keeps dropping in acro mode since dev20120504

[wilco1967]

Hi,

Did not read anyone else reporting, but it seems since dev20120504 the tail on my Tri keeps dropping (nose up) if I increase throttle, or if I fly forward.
Without constant forward stick (nose down command), the tail seems to drop, as if wanting to brake automatically.
If I keep the pitch stick neutral, and come from fast forward flight, the tails drops initially, but then seems to level off again.
It's kind of the behaviour you get from a trainer plane, which also tends to pick up the nose in a dive.
In steady hover, all is fine.

Just re-loaded dev20120414 and that is perfect again. It will maintain pitch attitude perfect were I left it... PID settings default on both.

Anyone else noticed this.

I guess this might have been added on purpose, but I find this behaviour very annoying
Is there some way of de-activating so it maintains the attitude we put it in ?

I'm definitely not critical. I really like this project, and very much appreciate the work that has gone into it.... just curious if this was intentional or not, and if it can be 'deactivated' It still flies fine like this, but I liked it a lot better as it was before.

Trying to get the EOSBandi's GPS version dialed in, which is great. But as that is based on dev20120504, is suffers from the same pitch behaviour.
I would stay at 0414 if it wasn't for the GPS...

BTW: flying a Flyduino mega, with WM+, BMA-020, HS5883, BMP-085. Perfect level mode also, no vibration related issues. Tried with gyro smoothing on/off.... no improvement. in 20120414 also without gyro smoothing everyting perfect.

[Crashpilot1000]

@wilco
My tri did the same! I flashed (after eepromclear) the new dev and set my PID in the GUI and did the calibration. My Hardware is a 800gr TRI (incl LIPO) with promini, orig WMP, orig nunchuk and bmp085 (ext Pullups,LLC, NO I2C ERRORS!). It was nearly unflyable in the wind today in just the gyromode! The tri dopped the tail all the time and i had to pitch forward like crazy. The tailmotor was hot like never before! The Tri flew without problems with the 1.9 version. Whats up?
BTW Wilco, can you post the dev20120504? I cant find it anymore.

[wilco1967]

I think dev20120504 already has this issue, (if I'm not mistaken).... dev20120414 was still OK

For the file, please sent me a PM with your email, I'll mail it to you.... (cannot attach here it seems).

[Crashpilot1000]

@Wilco1967 from Crashpilot1972
Thx for your offer! I tried my tri today with heavy winds with dev20120504. It seems to be better than the 522 Version. In the middle of nowhere i had no chance of flashing my old 1.9 for comparison with my 10minutes-i-am-dead-6Years old-Sony Laptop. My Email is: r.balko et yahoo.de.

So long
Rob

EDIT: I found a copy of the dev_20120414 !!

[polo_fly2]

I had the same with my tri and the last dev. Adjusted only the pitch " I " from
0.038 to 0.058 and now the tail holds the pitch stick input nearly forever.

Georg

[wilco1967]

I had the same with my tri and the last dev. Adjusted only the pitch " I " from
0.038 to 0.058 and now the tail holds the pitch stick input nearly forever.

Georg

OK, first I didn't think this would be the cause, but tried anyway.... You were right 100% spot on.....

increased the pitch I from 0.030 to 0.10, and it holds so much better. I think it could go even up further before becoming too much....
I always had the impression 'I' setting was already quite agressive, but that was as I found out now definitely not the case. In fact it was very soft....

I have no LCD, so my PID tuning always was very minimal, as it flew pretty well with stock PID's. I never bothered to bring the laptop out to the field, but perhaps I should do (or better, get an LCD). I do a reasonable amount of PID tuning for my work, so it's not I don't understand PID's, or are afraid to touch the settings.... it just it 'seemed' already quite good, and it 'seemed' version related. Wrong on both accounts
With PID settings, you need to 'feel' the difference, as you change them, and that's not possible without the laptop... PID tuning is an more of an art, and not an exact science.

So conclusion: there is absolutely nothing wrong with either software version (at least not with the pitch). Perhaps something was changed which affected overall gain which made me think is was version related, but nothing that cannot be corrected by proper PID tuning.

Thanks Georg and all others who helped on this project !....

Just came back with the Tri, flying with R33 GPS code and without any wind.... just beautiful.... now I (almost) don't want to mess with it anymore, as everything is perfect !

Now see if my quad also responds so well to this treatment ....