## Re: [Gpsbabel-code] GPS Help

 Re: [Gpsbabel-code] GPS Help From: Pierpaolo Bernardi - 2010-08-25 12:41:20 ```On Wed, Aug 25, 2010 at 14:22, Bill Hobson wrote: > I have noticed that, even when moving in an obviously very > straight line, the reported Track field can vary by 15-20 degrees left or > right of the real track being traveled. (In other words, if I'm actually > traveling say, 310 and start the program, it might show 301 then 320, then > 325, then 290, then 315, etc, etc - it's very unstable and couldn't, for > example, be used to provide input to an automated guidance system.)  Is this > because, in order to compute the current Track, the units are calculating > the Track they think has been traversed between the current position and the > position remembered from the previous sentence (both of which are only > accurate to within 40-50 feet in these non-WAAS receivers)? Yes, I think so. > And, is the > "swing" that is being reported just a natural by-product of the inherent > inaccuracies of GPS? Yes, I think so. > The slower the movement the larger the left/right error > seems to be.  Finally, are the algorithms necessary to compute Track from > two specific lat/lon fixes available in the public domain? Of course. It's just basic trigonometry. OK, maybe not so basic, but still centuries old math. Some of these links may be what you need: http://www.google.com/search?q=compute+bearing+from+two+positions Cheers P. ```

 [Gpsbabel-code] GPS Help From: Bill Hobson - 2010-08-25 12:23:23 Attachments: Message as HTML ```To be honest about it (since I may not actually be using GPSBabel), I don't know if it is appropriate that I join this site/list and seek your help. If, after reading my question, someone should determine that I shouldn't be here just let me know and I'll go quietly. I am trying to develop some navigation software and I am trying to find people who are knowledgeable about the NMEA sentence output feature of GPS devices. I have been doing some experimenting with USB GPS "dongles" and have come to some conclusions I'd like to bounce off of someone who understands the inner workings of these doodads. I have been focusing on the RMC sentence (which appears to be sent out once every second) and, specifically, the Track field that is within that sentence. I have written a program that displays this field each time a RMC sentence arrives and I have noticed that, even when moving in an obviously very straight line, the reported Track field can vary by 15-20 degrees left or right of the real track being traveled. (In other words, if I'm actually traveling say, 310 and start the program, it might show 301 then 320, then 325, then 290, then 315, etc, etc - it's very unstable and couldn't, for example, be used to provide input to an automated guidance system.) Is this because, in order to compute the current Track, the units are calculating the Track they think has been traversed between the current position and the position remembered from the previous sentence (both of which are only accurate to within 40-50 feet in these non-WAAS receivers)? And, is the "swing" that is being reported just a natural by-product of the inherent inaccuracies of GPS? The slower the movement the larger the left/right error seems to be. Finally, are the algorithms necessary to compute Track from two specific lat/lon fixes available in the public domain? (Or, perhaps, even better, is there a place where the actual BASIC code is published?) Thanks to all who reply. Bill Hobson Lowcountry Storm Panels (843)227-8887 http://www.lowcountrystormpanels.com ```
 Re: [Gpsbabel-code] GPS Help From: Pierpaolo Bernardi - 2010-08-25 12:41:20 ```On Wed, Aug 25, 2010 at 14:22, Bill Hobson wrote: > I have noticed that, even when moving in an obviously very > straight line, the reported Track field can vary by 15-20 degrees left or > right of the real track being traveled. (In other words, if I'm actually > traveling say, 310 and start the program, it might show 301 then 320, then > 325, then 290, then 315, etc, etc - it's very unstable and couldn't, for > example, be used to provide input to an automated guidance system.)  Is this > because, in order to compute the current Track, the units are calculating > the Track they think has been traversed between the current position and the > position remembered from the previous sentence (both of which are only > accurate to within 40-50 feet in these non-WAAS receivers)? Yes, I think so. > And, is the > "swing" that is being reported just a natural by-product of the inherent > inaccuracies of GPS? Yes, I think so. > The slower the movement the larger the left/right error > seems to be.  Finally, are the algorithms necessary to compute Track from > two specific lat/lon fixes available in the public domain? Of course. It's just basic trigonometry. OK, maybe not so basic, but still centuries old math. Some of these links may be what you need: http://www.google.com/search?q=compute+bearing+from+two+positions Cheers P. ```
 Re: [Gpsbabel-code] GPS Help From: Bernd Zeimetz - 2010-08-25 15:25:24 ```On 08/25/2010 02:22 PM, Bill Hobson wrote: > To be honest about it (since I may not actually be using GPSBabel), I don't know if it is appropriate that I join this site/list and seek your help. If, after reading my question, someone should determine that I shouldn't be here just let me know and I'll go quietly. I am trying to develop some navigation software and I am trying to find people who are knowledgeable about the NMEA sentence output feature of GPS devices. I have been doing some experimenting with USB GPS "dongles" and have come to some conclusions I'd like to bounce off of someone who understands the inner workings of these doodads. I have been focusing on the RMC sentence (which appears to be sent out once every second) and, specifically, the Track field that is within that sentence. I have written a program that displays this field each time a RMC sentence arrives and I have noticed that, even when moving in an obviously very straight line, the reported Track field can vary by 15-20 degrees left or right of the real track being traveled. (In other words, if I'm actually traveling say, 310 and start the program, it might show 301 then 320, then 325, then 290, then 315, etc, etc - it's very unstable and couldn't, for example, be used to provide input to an automated guidance system.) Is this because, in order to compute the current Track, the units are calculating the Track they think has been traversed between the current position and the position remembered from the previous sentence (both of which are only accurate to within 40-50 feet in these non-WAAS receivers)? And, is the "swing" that is being reported just a natural by-product of the inherent inaccuracies of GPS? The slower the movement the larger the left/right error seems to be. Finally, are the algorithms necessary to compute Track from two specific lat/lon fixes available in the public domain? (Or, perhaps, even better, is there a place where the actual BASIC code is published?) Thanks to all who reply. You might want to look at the gpsd documentation, there is a lot of info about various protocols. -- Bernd Zeimetz Debian GNU/Linux Developer http://bzed.de http://www.debian.org GPG Fingerprints: 06C8 C9A2 EAAD E37E 5B2C BE93 067A AD04 C93B FF79 ECA1 E3F2 8E11 2432 D485 DD95 EB36 171A 6FF9 435F ```