Drift issue: The software finds and centers the field okay. As time passes, the stars drift around (likely due to bad mechanics). At some point, the software decides to make a correction and moves to re-center the stars. The amount it allows the stars to move before it re-centers is too large.


From Lee:

Here is the file 'devices' and what it contains:
» cat /etc/rts2/devices
#RTS2 devices configuration
#device type    device_name     options
#example:

dome    ahe     D0      -f /dev/ttyS5
camd    alta    C0      -d C0 -c -35 --focdev F0 --wheeldev W0
teld    lx200gps   T0   -f /dev/ttyS0 --max-correction 30 -c 0.05
focusd  ez4axis F0      -f /dev/focuser
filterd alta    W0      -F "CLR:UU:BB:VV:RR:II:BG40:CLR2:CLR3" -t 1
sensor  bigng   S1      -f /dev/bigng
sensor bwcloudsensorii-weather S0 -f /dev/boltwood


You can get more info on the lx200gps driver by passing the --help arg.
Here is the details on the correction:
» rts2-teld-lx200gps --help
-c   minimal value for corrections. Corrections below that value will be rejected.
-g   minimal good separation. Correction above that number will be aplied immediately. Default to 180 deg
-l    separation limit (corrections above that number in degrees will be ignored)


Mike: So if I'm correct, to adjust how positional changes are done, edit the file /etc/rts2/devices and update the line
teld lx200gps T0 -f with different flags. It should probably look like
teld lx200gps T0 -f /dev/ttyS0 --max-correction 30 -c 0.02 -g 0.021 -l 30 (assuming all angular values are in degrees).
When the rts2 daemon is started, it will read these flags.



Tracking images:

FIXED!



The above image shows tracking after the adjustment described above.
This problem is solved.



The above image shows the position of the target star over the course of 400 images, spanning 5 hours.


The above image shows the position of the target star over the course of 800 images spanning 6.8 hours.