Hello

I own a Altina SiRF III bt GPS with Firmware
GSW3.2.2_3.1.00.12-SDK003P1.01a

I once was one of those reporting that it crashes using gpsd which resulted
in the -b option being added in future releases.

Playing around with gpsmon I have now found a way to switch my BT gps from
NMEA to SiRF binary and back quite reliably.

I suppose the problem is the baudrate change (or the BT stack not reporting
the baudrate correctly).

So this is how I manualy switch my gps:

# gpsmon /dev/rfcommX

Now it is in NMEA Mode, I get a reading and it tells me it's running with some
baudrate. This seams to be the baudrate of last time I opened that serical
device but not the true one.

I cannot set baudrate (s) or binary (n1) while in dumb NMEA Mode.

Let's switch to SiRF binary:

=> tSiRF

The display changes to SIRF but the driver falls back to NMEA.

But now I can set the baudrate:

=> s4800

Very important, switching to any other baudrate than 4800 caused me to loose the BT connection.

Switch to binary mode and reinitialize:

=> n1
=> i

Voila, the drivers successfully switches to SiRF and I get the SIRF output.

But this would be one-way, as soon as I disconnect I can never reconnect without resetting my BT
so it's back to NMEA.

I found a way to switch it back to NMEA but that does not seam to be as reliably as from NMEA to
SIRF:

=> n0i

Sometimes then I'm back to NMEA Mode.

Now I suggest an baud option to be added to gpsd to tell gpsd to switch and lock the serial line
to a specific baudrate and not go around probing for the right speed. This would make it possible
to switch my BT gps to SiRF reliably.

Also if -n is not used gpsd would need to set the gps back to NMEA before disconnecting, else it
won't be able to connect anymore.

Hi,

could you please check if the issues are still there when you use gpsd 2.92-1
from unstable?

Thanks,

Bernd

Benoit Panizzon wrote: