How exactly does a GPS recieve accurate reading for speed?

[Guest davidp]

I have a gramin etrex base model. I have taken it on various trips and it has given me extremely accurate readings. However, compared to a car speedometer and a radar gun on a smaller moving object, an rc car, it was innacurate. How exactly does it pick up readings?

[Guest ClayJar]

I'll go out on a limb and say it does it by dead reckoning: it knows its current location and time, and it knows the previous fix's location and time, so (X-X')/(t-t')=v

[Guest makaio]

As I understand it (and I could be completely off-base), the GPSR receives signals from the satellites it has locks on every 1-2 seconds. Given the current position and the position it was in 1-2 seconds prior, it can determine the speed necessary to have moved this distance. The displayed speed isn't necessarily how fast you are going at a given instant, but just the 'average' speed as calculated from the time you turned the unit on until the present.

 

I'm sure there are folks here who can provide much more accurate details on this.

[Guest peter]

In addition to the position and time data mentioned above, the GPSR also has Doppler frequency shift data that directly gives relative velocity between each received satellite and the GPSR. It gets this data automatically since it needs to tune the receiver to match the received frequency.

 

AFAIK, the actual algorithms used to derive the position and velocity from the observed time and frequency shifts and the filtering that is used to provide a more stable solution are still considered proprietary information. You can frequently observe a slight time lag in the displayed speed if you come to a rapid stop.

 

Could you elaborate on the inaccuracy you reported relative to a car speedometer and radar gun used on an RC model? My experience has been that under good reception conditions the GPSR reported speed is better than most car speedometers - i.e. it agreed with my time vs. distance calculation on a measured course. Most car speedometers read a bit low.

[Guest Kerry]

If anything especially these days if there's a big difference then it's normally not the GPS. It's not perfect but it's not bad either and generally well less than 1km/hour with the faster the speed the less the % error will appear.

 

Also bearing and speed are "in the past" with the bearing not entirely correct at being called a heading in the true sense of the term.

 

Cheers, Kerry.

[Guest Gliderguy]

just going out on a guess about the rc car vs radar gun:

 

the GPS speed indicated usually lags a few seconds. if the RC car was only at a peak speed for an instant, the GPS may not have recorded it. I am presuming you re-set the maximum speed on the GPS, put it on the car and ran it, then compared the freshly made max speed with the radar gun reading. If you used average speed then all bets are off because you would have been comparing apples to oranges.

 

another possibility is if it were actually on an RC car, there may have been RF interference that degraded its accuracy.