Re: [ST] GPS as a Speedometer vs. Bicycle Computer

["Rod Brown" <rodneybrown@xxxxxxxxxxxxxxxx>]

> Sorry to correct you - the bicycle computer depends on the rpm of the
front
> wheel, which will change the diameter in relation to the speed. So the
error
> is - depends on your calibration - around 5 % in minimum.
>
> The system immanent error of GPS in speed measuring is around 1%.
Enrico, sorry but I don't have a clue what system immanent error is but I do
agree that GPS accuracy will be a lot better than what Coop was saying.
Coop, your numbers are based on point accuracy.  Speed is point to point.
The reason why WAAS works is that it is a form of differential GPS.  The
data being computed in the GPS unit is calculation from the position of
points and derived from measing the distance to individual satellites at
known locations.  Over a short period of time, the error in that computed
position can be computed by applying an error factor from a surveyed ground
position to the position the satellites say it is.  The closer to the
correction point to the survey point the more accurate the data.  WAAS has
some kind of averaging system which retransmits the correction value.
That's how you get better positions.   Survey grade GPS can use this type of
technique to get mm accuracy but using a local GPS survey point.  Mapping
grade can get to 10 cm or less than a meter depending on techniques used.
Speed does not require point accuracy because it is kind of self correcting.
Because the points are so close together, the error cancels out giving the
distance between points and the time between the points which is the basis
of speed.

If you want to know why the bicycle calibration is going to have errors in
it, you can start with the diameter of the tire changes with lean, it
changes with tire pressure and temperature, probably even with speed itself.
My final guess is that a good bike computer and GPS will probably be within
a MPH or so and it doesn't make any difference.

Rod


     *   *   *   *   *   *   *   *   *   *   *   *   *   *   *   *   *   *   *
      The ST/RS Mailing list is sponsored by Jack Lilley Ltd.
          http://www.TriumphNet.com/st/lilley for more info
   http://www.TriumphNet.com/st for ST, RS and Mailing List info

=-=-=-= Next Message =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=