James - be sure to measure the rolling circumference of the wheel as the tyre distorts in contact with the road.
The Sigma BC1200 only seems to work if the magnet and sensor are in the same plane, ie parallel to each other, rather than forming a 'X' when they pass.
I inititially had some dodgey readings but traced them to a wheel balance weight knocking the sensor.
Failing that, check your maths?
I applied high beardage and measured 10 wheel revolutions to counter error; however, this proved unnecessary as the final reading was only a digit or so out. Subsequent calibration of the Sigma with a GPS showed it to be accurate to a percent or so, and then comparing that with a measured mile it seems accurate to 1/100th of a mile.
Richard - my wheel seemed OK without rebalancing it, but I have since moved the magnet, like Pierres, to the head of a brake disk bolt, so I can swap wheels around and retain the speedo. Although the epoxy glue is only guranateed to about 60 degrees C, it is still there after a couple of trackdays.
There is some detail of mounting the display on my site at www.confessions-of-a-beard.com.
Edited by - Nick on 12 Jun 2001 11:55:14