"Harry K" wrote in message
ups.com...
On Jul 29, 6:33 pm, (Scott) wrote:
On Sun, 29 Jul 2007 10:04:39 -0400, in rec.aviation.homebuilt, "Peter
Dohm"

wrote:
Have you 1) verified you odometer against at least 10 miles of highway
mile
markers, 2) verified your speedometer reading at 60MPH as a result of
the
same test, and 3) checked the cumulative fuel mileage, as shown on the
computer, against your fuel mileage calculated in the usual way?

Seconded. I've never seen an automotive trip computer that was worth
more
than the recyclable metals in it.

My Grand Prix has a trip computer, and what it tells me is scarcely
better
than a wild guess. It reports better fuel mileage than I really get,
and
underreports the amount of fuel I've used. When the tank's full it
tells me
I have a range of 430 miles (best I've ever gotten was ~350), and it
raises
a fuel alarm when I've still got 80 miles in the tank. Now and then, it
will raise a low fuel alarm when the tank is completely full.

It probably kicks puppies and steals candy from babies, too.

-Scott

Probably quite true but...

It is a useable instrument to detect what mode of driving is most
efficient even it it is not accurate. That is if it reports, for
example, 19 mpg over a stretch of road with tailgate down and then you
repeat the run in the same direction, same speed, same conditions,
etc. the reported mpg is then a useable bit of data. Yes, it will be
inaccurate but the comparison is useable and meaningful.

Harry K

Probably true. However, the coastdown test mentioned elsewhere in this
tread is probably the most consistently reliable method--when properly
controlled. The most obvious requirements a state the coast down from
the same place and speed each time, record the remaining speed at the
seconde marker, continue testing in the same direction, warm up the tires
before the first test, and choose a day (or days) with constant temperature
and wind conditions. There are certainly more, but those are enough to give
a far more reliable result than any test in traffic that I can think of--and
even then, since we are discussing the drag of detached flow, the test may
only be valid for the speed(s) at which the test was run.

Peter