Re: Open Source AHRS project: giving away hardware

2014-05-11 Thread Jake
On 05/11/2014 07:08 AM, joerg Reisenweber wrote:
 On Sat 10 May 2014 11:57:30 Pascal Gosselin wrote:
 calibrate the compass via GPS track by an easy calibration or
 self-calibration method
 
 GPS doesn't offer any data to calibrate magnetometer from. 
 magnetometer aka compass is about orientation of device, GPS is
 about position and movement vector. They are 100% unrelated.

In a static situation this is correct, but while moving it is possible
to get the current heading from GPS.

Jake

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Open Source AHRS project: giving away hardware

2014-05-11 Thread Pascal Gosselin

Indeed. When you taxi an airplane on the ground in a straight live or drive a 
car in a straight line, GPS TRACK (adjusted for magnetic declination) = 
Magnetic Heading.

The other complementary magnetometer calibration techniques involve driving 
around in a full circle in about 60 to 75 seconds or doing a 360 degree turn 
and stopping every 30 degrees.

-Pascal

Sent from my iPhone

 On May 11, 2014, at 4:47 AM, Jake jak...@rommel.stw.uni-erlangen.de wrote:
 
 On 05/11/2014 07:08 AM, joerg Reisenweber wrote:
 On Sat 10 May 2014 11:57:30 Pascal Gosselin wrote:
 calibrate the compass via GPS track by an easy calibration or
 self-calibration method
 
 GPS doesn't offer any data to calibrate magnetometer from. 
 magnetometer aka compass is about orientation of device, GPS is
 about position and movement vector. They are 100% unrelated.
 
 In a static situation this is correct, but while moving it is possible
 to get the current heading from GPS.
 
 Jake
 
 ___
 Openmoko community mailing list
 community@lists.openmoko.org
 http://lists.openmoko.org/mailman/listinfo/community


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Open Source AHRS project: giving away hardware

2014-05-11 Thread Pascal Gosselin



 On May 11, 2014, at 12:03 AM, Ben Wong lists.openmoko@wongs.net wrote:
 
 Sounds like a fun challenge. I'd take it up if I wasn't already so
 busy. After all, who in their life hasn't at one point wanted to write
 an Unscented Kalman Filter? ;-)
 
 Just out of curiosity, is OpenPilot.org not a viable solution?

My understanding is that the code written for micro-controller platforms will 
need significant changes to work in a Linux environment.

Of course re-using as much of the code as possible that is used for controlling 
most small drones totally makes sense.

-Pascal



___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Open Source AHRS project: giving away hardware

2014-05-11 Thread joerg Reisenweber
On Sun 11 May 2014 09:25:37 Pascal Gosselin wrote:
 Indeed. When you taxi an airplane on the ground in a straight live or drive
 a car in a straight line, GPS TRACK (adjusted for magnetic declination) =
 Magnetic Heading.
 
 The other complementary magnetometer calibration techniques involve driving
 around in a full circle in about 60 to 75 seconds or doing a 360 degree
 turn and stopping every 30 degrees.
 
 -Pascal
 
 Sent from my iPhone
 
  On May 11, 2014, at 4:47 AM, Jake jak...@rommel.stw.uni-erlangen.de 
wrote:
  On 05/11/2014 07:08 AM, joerg Reisenweber wrote:
  On Sat 10 May 2014 11:57:30 Pascal Gosselin wrote:
  calibrate the compass via GPS track by an easy calibration or
  self-calibration method
  
  GPS doesn't offer any data to calibrate magnetometer from.
  magnetometer aka compass is about orientation of device, GPS is
  about position and movement vector. They are 100% unrelated.
  
  In a static situation this is correct, but while moving it is possible
  to get the current heading from GPS.
  
  Jake
  

This all assumes a locked and defined mounting situation for the magnetometer. 
Then yes. For an embedded device however this method tells you nothing about 
the magnetometer heading. The embedded device can change relative orientation 
to the vehicle that's driving.
PS: you must be very sure about the vehicle moving exactly straight ahead as 
well, for anything but a non-sliding car that's not guaranteed, think boat, 
even airplane
-- 
()  ascii ribbon campaign - against html e-mail 
/\  www.asciiribbon.org   - against proprietary attachments
(alas the above page got scrapped due to resignation(!!), so here some 
supplementary links:)
http://www.georgedillon.com/web/html_email_is_evil.shtml  
http://www.nonhtmlmail.org/campaign.html
http://www.georgedillon.com/web/html_email_is_evil_still.shtml
http://www.gerstbach.at/2004/ascii/ (German)


signature.asc
Description: This is a digitally signed message part.
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Open Source AHRS project: giving away hardware

2014-05-11 Thread Pascal Gosselin

On 2014-05-11, 11:28 AM, joerg Reisenweber wrote:

This all assumes a locked and defined mounting situation for the magnetometer.
Then yes. For an embedded device however this method tells you nothing about
the magnetometer heading. The embedded device can change relative orientation
to the vehicle that's driving.
PS: you must be very sure about the vehicle moving exactly straight ahead as
well, for anything but a non-sliding car that's not guaranteed, think boat,
even airplane



We're talking calibration here.  Yes the unit should be rigidly mounted 
for calibration of the sensors.  There are also periods of stay still 
for X seconds at various points in such a calibration.   I have done 
and continue to do lots and lots of AHRS calibrations of various types 
on aircraft (airplanes and helicopters).  I would be more than happy to 
share the information that I have on various calibration techniques.


Once the sensors are calibrated (i.e. figuring out the drift of the Rate 
Gyros when sitting still) and the magnetic environment of the device is 
known, what's left is the alignment procedure. Every time you start the 
AHRS code on the device, it would need to be motionless for a while.  
Lying the GTA02/GTA04 flat on a table for example for perhaps 2-3 
minutes might be sufficient.


I feel that this likely the reason why even the latest mainstream phones 
don't have AHRS or IMU capability (an IMU would enable indoor navigation 
over only very short distances in a smartphone, given the rather crude 
quality of the MEMS sensors and horrific gyro drift expected if you are 
bouncing around with the smartphone in your hand and moving rather 
slowly without GPS aiding the Kalman filter).


Application for an AHRS on a smartphone would be for enhanced 
geo-referencing of photos and Google Glass type applications that 
don't make you look like a Glasshole.  The other obvious application is 
as an emergency backup attitude (Pitch and Roll) and Heading indicator 
for airplanes and helicopters or simply recording of the sensor data 
and doing post-processing on a server to process the data later (think 
extreme sports, like playing back a skydive for example).


-Pascal



___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Open Source AHRS project: giving away hardware

2014-05-10 Thread Ben Wong
Sounds like a fun challenge. I'd take it up if I wasn't already so
busy. After all, who in their life hasn't at one point wanted to write
an Unscented Kalman Filter? ;-)

Just out of curiosity, is OpenPilot.org not a viable solution?

--Ben


On Sat, May 10, 2014 at 8:57 AM, Pascal Gosselin pas...@aeroteknic.com wrote:
 Hello,

 With the idea of spurring interest in the development of a software Attitude
 Heading Reference System for the GTA02, I am going to be giving away twelve
 (12) Golden Delicious NavBoard V2s to developers that are willing to work on
 this idea.

 I have been frustrated by lack of an available AHRS for Linux-based systems
 for many years now.  The closest thing available is the FoxAHRS:

 https://github.com/FedericoLolli

 I believe it runs on the ARM-based G20 board:
 http://www.acmesystems.it/FOXG20

 I might be wrong but anyway the idea is to get a working 100% open
 source AHRS than can run on the GTA02 and GTA04.  Ideally it should be
 entirely self-calibrating (i.e. figure out what's level and calibrate the
 compass via GPS track by an easy calibration or self-calibration method).


 This is what the boards look like:

 http://wiki.openmoko.org/wiki/Freerunner_Navigation_Board_v2

 It has a 3D ITG-3200 Rate Gyro, the HMC 5843 magnetometer and BMP 085
 barometric sensor.  It's a board you must solder internally to the GTA02.

 Anyone that actually ends up contributing to this project will also be given
 a Navboard V3, which has a better magnetometer, to make sure it works that
 one as well (I paid good money for the V3 boards so I don't want to give
 them away to people who may not ever end up installing them).

 If anyone that is a serious contributor needs some help, I would be more
 than willing to help out with stuff like free brand new GTA 02s (850Mhz or
 900Mhz, I have both brand new in stock) , new batteries.   Serious hackers
 don't need to worry about damaging their GTA02, I will provide replacement
 units to anyone who does and is serious about this project.

 I also have a early GTA04 fully assembled with new case/screen/battery to
 give away to a serious contributor who would be willing to ensure that the
 AHRS code would run properly on this platform as well.

 You may reply to me privately or via this list.

 Pascal Gosselin
 pas...@wi-flight.net










 ___
 Openmoko community mailing list
 community@lists.openmoko.org
 http://lists.openmoko.org/mailman/listinfo/community

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: Open Source AHRS project: giving away hardware

2014-05-10 Thread joerg Reisenweber
On Sat 10 May 2014 11:57:30 Pascal Gosselin wrote:
 calibrate 
 the compass via GPS track by an easy calibration or self-calibration 
 method

GPS doesn't offer any data to calibrate magnetometer from. 
magnetometer aka compass is about orientation of device, GPS is about position 
and movement vector. They are 100% unrelated.

/j
-- 
()  ascii ribbon campaign - against html e-mail 
/\  www.asciiribbon.org   - against proprietary attachments
(alas the above page got scrapped due to resignation(!!), so here some 
supplementary links:)
http://www.georgedillon.com/web/html_email_is_evil.shtml  
http://www.nonhtmlmail.org/campaign.html
http://www.georgedillon.com/web/html_email_is_evil_still.shtml
http://www.gerstbach.at/2004/ascii/ (German)


signature.asc
Description: This is a digitally signed message part.
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community