[Space] draft blog post

[Jonathan Moore]

I am working of a python script to run the analysis that we can use
and distribute with the android code. I  have attached a sample of my
work so far. Still has a long way to go but you can see the direction
it is heading.

-Jonathan

On Tue, Feb 9, 2010 at 2:36 AM, Christoph Maier
<cm.hardware.software.elsewhere at gmail.com> wrote:
> On Tue, 2010-02-09 at 00:48 -0800, Michael O wrote:
>> The orientation sensor is supposedly showing pitch,roll,yaw angles as
>> derived from the magnetometer.
>>
>> The accelerometer is supposedly showing meters/second^2 acceleration
>> per X,Y,Z axis. So it should show 9.8m/s^2 on the ground.
>> I seriously mis-trust the accelerometer data. It's showing
>> acceleration values that are way too high or too low at different
>> parts of the flight.
>>
>> The GPS data is also mildly strange. It looks like the GPS is low-pass
>> filtering altitude, and it's unable to cope with the high ascent and
>> decent rates. That's my best guess for why it steps altitude by up to
>> 500 meters on occasion. The lat/long data looks fine; it's just the
>> altitude.
>>
>> I'm yet to go and take a close look at the acceleration directions (I
>> just did the magnitude) so it may be that the payload really was
>> swinging that wildly...
>
> I'll put my attempts to make sense of the data at
> http://pony.noisebridge.net/~cmaier/EpicBalloonWin/
> until shown a better place.
>
> I agree that the GPS seems to limit ascent and descent rates to
> something like 5 meters/second. Latitude and longitude _does_ seem to be
> affected somewhat, but this is less apparent than in the altitue
> dimension.
>
> The data from the accelerometers and magnetometers drifts substantially
> while on the ground, which makes me doubt the accuracy of the readings.
>
> In the air, _all_ the readings oscillate wildly ... so far, I haven't
> zoomed in on a region to verify that the oscillations are consistent
> between sensor outputs, but it looks like the swinging was real. The
> pictures and movies support this assumption, too.
>
> _______________________________________________
> Space mailing list
> Space at lists.noisebridge.net
> https://www.noisebridge.net/mailman/listinfo/space
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sample-out-1.png
Type: image/png
Size: 76665 bytes
Desc: not available
URL: <http://www.noisebridge.net/pipermail/space/attachments/20100209/429fdb62/attachment.png>