Those are strange spikes. If you graph the MOT_ROLL, PITCH, and YAW values, you can see them clearly there also, A lot of them on the Yaw axis especially. I can't see any inputs in the log which would explain these spikes. The MAGs are very noisy (and you can see an LED or something blinking on there), but they're the same in the old logs so that doesn't explain the spikes.
I did notice in your previous logs, and these, that there is a good amount of imbalance in the frame, meaning the PITCH/ROLL/YAW outputs tend to average higher or lower than zero. I'm not suggesting this is causing the spikes, but it would make the whole setup more efficient if you could get rid of those imbalances.
The closest equivalent to what you were flying with Menno's build is probably 7.1.b1852 (
ftp://ftp.autoquad.org/3/334693_22529/a ... d-ext_esc/ ). The last of the 7.0 series (r442 b1848) is also going to be very close if not the same (
ftp://ftp.autoquad.org/3/334693_22529/a ... d-ext_esc/ ).
From my "next" branch you may want to start with b1871 (
ftp://ftp.autoquad.org/3/334693_22529/a ... .6-quatos/ ).
If we can find a difference in the spikes somewhere between those two versions, then I can build you some other interim versions to see if we can narrow down when the problem starts. Or if neither/both of those show the problem, we'll know the issue lies elsewhere.
You also switched AQ hardware (as I understand it) so that may complicate the diagnostic process. Or did you try the original fw (Menno's version) on this new hardware as well (with no spikes)?
Cheers,
-Max