paparazzi-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Paparazzi-devel] Magnetometer behavior


From: Sergey Krukowski
Subject: Re: [Paparazzi-devel] Magnetometer behavior
Date: Sat, 29 Nov 2014 10:58:19 +0100
User-agent: Opera Mail/11.61 (Win32)

Nice tip!
HMC5xx3 is widely used, and I saw only once or twice, that it was used with real tantalum capacitors. But anyway, the problem could also be because of the bus stuck. And imo the alternative ahrs/imu behavior would be a better solution in failsafe case.


Maybe this could be helpful for the hardware.

When I find a magnetometer not responding, one possible reason might be
related with the "pump capacitor" (in the case where it is required).
Normally this capacitor is a tantalum one. Depending on the quality of such
capacitor, its series resistor might change dramatically with low
temperatures, leading to a malfunction on the magnetometer.
On Nov 29, 2014 2:15 AM, "Sergey Krukowski" <address@hidden> wrote:

Hi guys!

As I understand, currently if the magnetometer freezes (for any reason),
the on_mag_event function will not be called anymore, which leads to slow
(or even not slow) drift of yaw angle and resulting spinning of the
rotorcraft. Which could be worse for possible failsafe recovery (because
even in attitude mode the rotorcraft will spin around), rather than
updating the ahrs with the last received magnetometer values (in that case
the yaw angle will stay constant and the rotorcraft at least won't spin
itself).
If I understand it right, what do you think about possible changing of the
ahrs architecture in that case or should it better be done on particular
boards IMU level?

Best regards,
Sergey

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel



--
Erstellt mit Operas revolutionärem E-Mail-Modul: http://www.opera.com/mail/



reply via email to

[Prev in Thread] Current Thread [Next in Thread]