lamp air flow sensor

They make/ sell hot wire air flow sensors. I made (copied) a pressure sensor that uses the self heating of a thermistor. You could probably make that work for air flow too.

George H.

fredag den 29. marts 2019 kl. 19.04.32 UTC+1 skrev bitrex:

it relies on self heating and the cooling from the flow, you could use resistor instead, in any case you have to compensate for air temperature to be real accurate. and it measures mass flow not velocity

sounds like something Rube Goldberg would be proud of ;)

What you describe is a Pirani vacuum gauge. Works well for vacuum, not so well for normal atmosphere. Old computers used self-heating thermistors to force a shutdown on fan failure.

For higher airflow, hot wire anemometers are commonly used, mass airflow sensors in cars often use this technique.

Jon

You can use a pair of TO92 IC temp sensors, like LM35s, as an air flow thing. Make one dissipate more power than the other and compare their reported temperatures.

Or apply a constant voltage to a PTC thermistor or a polyfuse and note the current. It will increase considerably with air flow. That might not need ambient temperature correction if there's a lot of flow. Fun rainy-day experiment, with a radial-lead polyfuse or one of those disc inrush limiters.

I designed a sensor like that about 10 years ago with thin film platinum sensors. Worked pretty well in air.

--Spehro Pefhany

If you want to do it on the cheap just use as high current low voltage lamp as practical. Robustness depends on current (wire thickness) and length (voltage), and support count to a lesser extent.

If you need enough robustness to deal with high speeds use a more solid thermistor, which is about what a heavily underrun lamp is.

NT

We get a lot of wind where I'm at I was thinking a fun weekend project would be a "solid state" wind-meter. not looking for single digit mph accuracy but with a readout more like a bargraph LED. A "VU meter" for wind, I could stick it outside my dining room window.

It reminds me of when I was kid my father's friend had an expensive car from the late 70s, maybe it was a Caddy, I recall it actually had some of the lamps like for the turn signals perhaps mounted on the _outside_ of the car, on the hood, rather than the dash. A strange design decision that never seemed to catch on but kinda cool

d.

m sensors. Worked pretty well in air.

Ultrasonic Anemometer sound more interesting, three transceivers measure speed of sound between them

all you need is 2 self heating temp sensors, one exposed to the wind one not. Plug em into an opamp & Robert's yer uncle. But I'm not sure who's your daddy. And a bargraph if you don't like needles.

NT

This is the sensor for a Kane-May 4107

Width of the channel is 0.3"

A bare lamp filament is downright ROBUST compared to the smaller thermistor.

IIRC, I used #382 lamps in the lamp sensor I built.

If I did it today, I'd use a microcontroller and just build a calibration table instead of trying to linearize it.

My dad had a 1969 Chrysler New Yorker that had both--the exterior ones were in the chrome front fender trim.

Cheers

Phil Hobbs

(on the train to DC)

Mhmm, mhmm, I see what you did there :) To rephrase the signal _indicators_ were on the hood. I found an example to confirm I didn't imagine it:

Most cars used to be invisible at night from the side, even with their lights on.

So that's a little thermistor hanging in the middle. Do you know the value?

George H.

Robustness also depends on whether the lamp has been used, as exposing the tungsten filament to high temperatures causes it to recrystallise and become brittle. A brand new filament is a LOT more robust than a used one.

John

I thought filament lamps were all tested before shipping.

NT

Sorry. I don't think it matters much as long as the temperature characteristic is the same as the bigger one below it. I'm not gonna stick a probe in there to try to measure it ;-)

My memory is fuzzy, but if you've got enough power to burn, it's easier to put a big and a little thermistor in a bridge and stuff current into the little one until the bridge balances. The required current (assuming constant voltage) represents the wind speed. Obviously, both thermistor temperatures have to be higher than the highest ambient temperature.

Running a tungsten lamp puts its filament in a fully annealed state, I believe. It's the evaporation/redeposition that leads to large crystals and weak spots.

Cheers

Phil Hobbs

Yes, but only for a few seconds.

Halogen lamps are designed to redeposit tungsten. Does that happen in general lighting service bulbs or is the tungsten lost to the glass envelope?

John