digital compass modules

Christian "Eddie" Dost ecd at brainaid.de
Wed Jan 23 23:14:26 CET 2008


You cannot calculate North (or any Heading) from Accelerometer data.

You need a 3D Gyroscope (or 3 Gyros on 3 orthogonal axis). With this
you can detect the orientation of the gyro relative to earth's 
rotational axis, and calculate gyroscope north from that. This is
the same as true north if the device does not move. If the device
moves, you need to compensate for the motion, this can be done using
latitude and speed over ground from GPS.

You can, however, calculate north from the data of two GPS antennas
about 1 meter apart, if you throw the *raw* data at a smart enough DSP.
But this has to be raw data, not filtered NMEA output.

There are devices using either concept available if you search the web,
but this is mainly high-end equipment worth 50 Neos or more. Mostly used
in Mil applications.

I once found a small device with GPS, Gyro and Accelerometer inside a
box the size of the Neo, for use in robotics or remote controlled
vehicels. But even this would sell for the price of 3-4 Neos.

The Gyroscopes need to be quite exact to get a reasonable north heading,
with "toy" Gyros you can probably better estimate north from azimuth of
the sun.

I hope this explains a little, just some thoughts I wanted to share, 
because I looked into this stuff some time ago.

Fair winds,
Eddie


Denis wrote:
> I think the accelerometers don't provide enough accuracy.
> 
> 2008/1/23, joerg <joerg.twinklephone at gmx.de>:
>> Am Mi  23. Januar 2008 schrieb Tilman Baumann:
>>> François TOURDE wrote:
>>>> Maybe the 3D accels can do that. And the GPS can be used as a bearing
>>>> indicator, when you move. No magnetic device needed in this case.
>> As stated in prev posting, GPS _can_not_ deliver bearing of device at all.
>> Imagine having the GTA in your pocket while moving - there's not the
>> faintest
>> relation between movement vector as seen by GPS and bearing of GTA.
>>
>>> Gyroscopes is what you look for. ;)
>>> Accelerameters don't see rotational movings.
>> Though gyro won't help here at all, i think TWO 3D-accelerometers placed
>> some
>> distance from each other in a system make a nice gyro with the aid of some
>> mathematics in driver. IIRC there was mentioned more than one acc-meter for
>> the GTA. I thought that's exactly for this purpose.
>>
>> jOERG
>>
>> _______________________________________________
>> OpenMoko community mailing list
>> community at lists.openmoko.org
>> http://lists.openmoko.org/mailman/listinfo/community
>>
> 
> _______________________________________________
> OpenMoko community mailing list
> community at lists.openmoko.org
> http://lists.openmoko.org/mailman/listinfo/community

-- 
___________________________________________________brainaid_____________
Eddie C. Dost           Rue de la Chapelle 51      phone +32 87 788817
                         B-4850 Moresnet            fax   +32 87 788818
ecd at brainaid.de         Belgium                    cell  +49 172 9312808





More information about the community mailing list