Originally Posted by
soma5
The only valid argument for placing the magnet (and sensor) closer to the hub is that the linear speed of the magnet relative to the sensor is lower so that there is less chance that a revolution will be missed (not counted).
I have seen it argued on these forums that it is better to put the magnet close to the rim in order to produce a higher "spike" on the sensor. i.e., they were claiming that with the sensor close to the hub, it could still faintly detect the magnet at 180 degrees away and thus the "low point" reading to the "high point" reading won't be as great, and thus the sensor might not count some revolutions.
Not saying I believe that theory... Just pointing out that it has been said. Personally, I subscribe to the theory that getting the magic tire circumference number correct is far more important then magnet placement.