> - The problem with the reed switch you describe is unlikely to happen in
> real life, and can be easily remedied in software. Just check for minimum
> time b/w pulses, or implement an even more clever algorithm. I have no
idea
> why Olin thinks it is not a good choice (it's certainly good enough for
> commercial speedos).


As you say, to get the problem described is practically impossible.  The
wheel needs to move a reasonable distance to toggle the reed switch.

There's a reason reed switches are used, well, several actually.  They're
cheap, simple, work, and use no power.  Even the cheap bike lights have gone
to reed switch & magnet assemblies for the on/off switches these days, it
makes waterproofing easier.

Calculation time between pulses is pretty easy.  A bike wheel is a bit over
2000mm in diameter, say 2100mm.  Assume a top speed of 100km/h.

100km = 100,000,000mm, so to travel that distance the wheel rotates
100,000,000 / 2100, or 47,619 times.  Divide by 60 to get RPM = 793, divide
by 60 again to get revolutions per second = 13.2.

So at 100km/h the sensor is sending out a 13Hz signal, giving you 0.075
seconds between each pulse.  A second pulse arriving under that time can be
ignored (especially since 50km/h is a more realistic figure for most
people).

Tony

(The metric challenged are welcome to use miles, feet & cubits to check my
figures)


-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist