On Mon, 22 Dec 2003 07:07:55 -0800, Grant Erwin
wrote:

I have been thinking about the issue of model equivalency between an air
regulator and a transformer again.

Hot wire aenomometers: check out
http://www.efunda.com/designstandard...res_theory.cfm

A simple filament driven by constant current (monitor voltage) would
be a start. I've used automotive taillight bulbs with the envelope
carefully broken away. Smaller bulbs are more sensitive, also more
fragile.

A bridge setup using two bulbs, one in the flow and one shielded from
flow, both driven by the same constant current, provides some
compensation for ambient temperature.

Both of these schemes assume constant resistance of the filament which
isn't so, but may be close enough for foolin' around -- particularly
with a calibrated device used at about the same ambient temp most of
the time. Ya don't run the filaments anywhere near red hot, just
warm. Delta -T may be small enough that delta R with delta-T may be
negligable compared to effects of air velocity.

For higher accuracy, consider small resistors (maybe surface mount)
with thermistor beads glued to them in close thermal contact. Use
fine wires or traces so as not to heat-sink them. Drive both
resistors with current controlled by opamps controlled by the
respective thermistors so both resistors are held at the same
temperature. The voltage difference on the resistors would then be a
function of air velocity -- perhaps about proportional to the square
root of velocity. This should make a pretty decent instrument,
though a bit slower than just a filament.

The mass flow meters in some fuel-injection systems work by hot-wire
aenomometry, as Bosch. (At least they used to)

Honeywell made a device called the "microbridge" which was essentially
a silicon hot-wire aenomometer bridge on a chip. Those things were so
fast they'd respond to sound waves. We used them with small
orifices to measure small pressures; if someone was talking loudly
near the duct I'd see audio on the pressure signal!