GPS lag

[bray]

Hello

I have sirf III (modulstek) gps receiver module which is very compact and works very well but i found relatively big delay/lag compared to real time situation. For example....current speed data is somehow around 5 seconds behind real speed.
How is this in other gps modules?

[Unterhausen]

some of the ublox have a good reputation for update rate. You should be able to find something in this forum about lag and GPS.. GPS modules have a processor board on there that is estimating parameters like velocity. I think that is better done in a Kalman filter that you run yourself, so ignoring the velocity is probably not a bad idea.

I thought I saw something recently about some Sirf III chips having different modes; in effect telling the on-board Kalman Filter what kind of dynamics the system has so it can filter out changes more effectively.

[Jack Crossfire]

1s is the delay on the Sirf III. Climb rate is delayed slightly more. Horizontal velocity comes in at the same time as position.

[bray]

One more "problem"....speed value from $GPVTG NMEA sentence which is lower then if i calculate it manually or in any program that calculates speed from points and time.

example:
Point1: Lat: 46.199100 Lon: 14.315973 Alt: 398m Time: 14:49:31
Point2: Lat: 46.198992 Lon: 14.315673 Alt: 401m Time: 14:49:32
Distance: 26.3 m
Time diff: 1 s

speed = 26.3 m/s = 94.86 km/h

But gps module gives me a 86km/h on this track segment....

Can someone explain this?

[dnile]

I believe that good GPS modules calculate speed from doppler shift, rather than integrating successive position measurements. This would explain the discrepancy. What you've done there is calculated your average speed over 1 second by using two position measurements one second apart, as opposed to calculating your instantaneous speed.

[Jack Crossfire]

The VTG sentence contains only the horizontal component of speed & it is smoothed. But why talk about it when U can just plot it?