Re: [Gpsbabel-code] [Fwd: Skytraq Venus based GPS datalogger]
Brought to you by:
robertl
From: Martin M. <mar...@gm...> - 2009-02-10 19:49:21
|
Hi Mathias, I removed the lines as suggested. My logger is connected via USB. a funny thing happened: ----------------------- Variant d: ./gpsbabel -D4 -i skytraq -f /dev/ttyUSB0 -o gpx -F /tmp/output_2.gpx > /tmp/output_2.log WORKS! ----------------------- Variant e: ./gpsbabel -D4 -i skytraq,read-at-once=1,dlbaud=9600 -f /dev/ttyUSB0 -o gpx -F /tmp/output_3.gpx > /tmp/output_3.log does not work ----------------------- Variant f: ./gpsbabel -D4 -i skytraq,read-at-once=1 -f /dev/ttyUSB0 -o gpx -F /tmp/output_4.gpx > /tmp/output_4.log WORKS ! ----------------------- See logs at: http://martin.maurer.selfip.org/martin/files/gps2.tgz So it looks like I will have to do some further tests, as soon as my logger has a longer track again. quite promising. regards, Martin On Tue, 2009-02-10 at 20:06 +0100, Mathias Adam wrote: > Hi Martin, > > Martin MAURER schrieb: > > all these settings failed, yet the last one has maybe just a simple > > reason: it tried to read data from sector 1 where most likely only > > sector 0 will contain data. > > > > I stored the data temporarily on: > > http://martin.maurer.selfip.org/martin/files/gps.tgz > > thanks for your logs! I just looked at them and found the following: > > * it seems that sometimes single bytes get lost within sector transfers, > hence the checksum and "Unknown item type" errors. I remember that we > had already observed this in the logs you provided back in November... > Perhaps reducing the baud rate could help here (you could try dlbaud=9600)? > How do you connect the logger (bluetooth or USB or ...)? > > * after reading a single sector my logger resets the baud rate so I put > a workaround in skytraq_read_single_sector(). However your device seems > not to have this bug thus the workaround breaks things here: that's why > reading sector #1 in your last log fails. > You can try commenting out lines 699-701 (the 3 lines after the > commented out "rd_char" block) and recompile. > > * for the moment it seems to be necessary to force read-at-once=1. I'll > have to check why the automatic detection currently fails. > > Please let me know if this makes any changes. > > Regards > Mathias > |