> I didn't see the Electronics Now project but I've seen this done using a
> small pilot bulb with the glass broken off as the sensing element. This
> wouldn't use too much current from a Ni-Cad pack.

    That sounds like a good idea. AFAIK the EN article used a solid state
device self heating characteristic to determine airspeed. I think it used
two transisters, one as a control device and one as a sensor. I remember
reading it, was interested in making it until I found out that calibration
wasn't that easy, it involved traveling in a car on a windless day, just
couldn't be bothered with that. I can look it up if someone is interested.
TTYL