What AngBst should normally look like?

Log Analysis -- deprecated Copter Logs -- deprecated
1

Hi all,

I was running 3.2 firmware, while flying in AltHold, my quad suddenly output full throttle, and it clime extremely fast. I’m new to this hobby, so I panic. Tried to switch to LAND, but the problem still there. Stablize works…but it was too late … any way, I got my quad back a few hundred meters away…

After looking at the log, I found that the ThrOut was outputing at its MAX 1000 for quite a long time. While I don’t have much ThrIn. The altitude indicators looks normal too, it is the AngBst which “went to the roof”.

See the part in the red rectangle…
[attachment=1]Fly_Away_Log_rz.png[/attachment]

And this is the AngBst…
[attachment=0]Fly_Away_Log_CTUN_rz.png[/attachment]

So…

  1. How the AngBst is calculated… for I could recall, when this happens, the copter is almost not leaning… but it was fighting against some gust.
  2. Is it normal for the AngBst to be so large… even to -7000?
  3. Any suggestion on avoiding this.

Thanks
Jia

2

Can’t really understand the problem or help without having the log, sorry.

3

Hi jschall,

Thanks for reply. I’ve attached the log.

For what it worth, here are what happened.

  1. Quad hovering in AltHold, suddenly, it climbed at full throttle with no CH3 change.
  2. Switch to LAND doesn’t help.
  3. Switch to Stabilize regain control of altitude.
  4. Switch back to AltHold continue to climb at full throttle for at least 15s.

Questions:
a) Why the AngBst was so large? I understand it is ( ThrottleRequired - ThrottleIn ), so why ThrottleRequired is so large?
b) The Altitude data doesn’t looks bad, so is it cause by gyro?

[attachment=0]2014-11-08 16-10-54.zip[/attachment]

-Jia

4

AngBst is behaving normally, albeit maybe a little inscrutably - the throttle output is being constrained so AngBst is going big negative.

But what you should be focusing on is CTUN.CRt, CTUN.Alt and CTUN.BarAlt. Alt tracks BarAlt poorly, CRt is hugely negative when the slope of BarAlt clearly is not so. You’ve just got vibration issues.

5

Are you referring to the second half of log? I thought that’s already lost control, and the slop of climb rate is expected.

What I still don’t understand is that, the Alt follows BarAlt close to me before the throttle went roof. Are you suggesting that the ThrOut is ahead of Alt and BarAlt???

Jia

6

I’m suggesting that the vertical inertial navigation screwed up, because the climb rate from the inertial navigation does not match the climb rate from the barometer.

There are really only two possible causes:

  • Too much vibration
  • Attitude error

Typically we see the first one.

ThrOut hitting limits is the result of the copter thinking that it is falling when it isn’t.