While a GPS can be a good check for steady-state speedo error, it doesn't do nearly as well as an odometer accuracy check due to the way it manages position jitter and averaging. I'm still pending to get mine on the interstate and use the mile markers to calibrate.
Yes this is correct. It only knows instantaneous position (with some error) and extrapolates distance between samples. It has to assume a straight line in between. The position error will average out but the extra travel between points is lost. However, I have found that it is VERY accurate if you are traveling in a straight line.
I just got back from a 2,600 mile run that was mostly interstate
and it was boring but provided the perfect opportunity to do the definitive speedometer and odometer accurate study.
The speedo was simple, I took it to 120 indicated and it read 110 GPS. There is some inaccuracy because both the speedo and the GPS read out in even MPH, no decimal points, but that is minimized at the highest speed. So I calculate for my bike at least, the speedo reads about 9% higher than true, or to look at it the other way, true is 8.3% less than what the speedo reads.
The odometer took a bit more work. I wasn't sure at the start how accurate the interstate mile markers are or the GPS. So I compared both of them and the odometer. I did 4 separate 100 mile runs on straight sections of interstate in two states. I was surprised at how close the GPS and mile markers matched. On straight interstate they matched within a few tenths over 100 miles! ???. If less straight, the GPS fell a bit more behind, but not as much as I expected. So I concluded the mile markers are very accurate over long intervals. So how did it come out?
All 4 readings were very close and averaged 103.4 miles for the actual 100 miles, or 3.4% high.
As always, YMMV (Pun intended
)
John