Need very low airflow sensor idea

[Jim Basinger]

I need to measure the velocity of very low airflow.........less than 1/2
foot per second or so. When I say measure, I mean an analog signal or that
I can store in a data logger and calibrate against some known. This is not
a switch application.

I have an Extech "hot wire" anemometer but aside from the fact that it does
not measure airflow at the velocities I need, the sensor "head" is
directional.

I have seen some ideas for using self-heated transistors, but don't know if
this type of idea is sensitive enough.

I think I have seen a design using a piezo sensor of some sort that could
sense the wafting of a hand over the sensor, but I don't remember if it
output an analog (relative) signal or not.

The application for this device is to measure the air movement at a forest
floor to about 6 inches from the ground. The sensor and data logger would
be left in place to collect data, since a persons presence might have a
large effect.

Any ideas on how I might build this?

Thanks
Jim
[email protected]

 

[Robert Lacoste]

Ultrasounds ? Transmit a pulse and measure the frequency variation of the
received signal after a short travel in the air, say 10". Thanks to M.
Doppler a 1/2 foot per second air speed should change the frequency by
0.02%, no ? (could someone recalculate it please ?). Put two of them at 90°
and you get both wind direction and speed... I've seen wind anemometers
built on that principle.

Yours,
Robert Lacoste - ALCIOM : The mixed signals experts
http://www.alciom.com

 

[Pat Ford]

make your own hot wire sensor, with a larger gap and either 2 crossed wires
or 4 seperate wires.
That way you also get direction info. You could also do a single vertical
wire.
Pat

 

[mike]

A miniature sail attached to an axis with torsion spring return coupled
to rheostat....some kind of gizmoid with electronically adjustable
counter torque feeding into a nullification thingamajig...

Your biggest problem is going to be biology, not technology.

No matter what you do, spiders are gonna find it and set up
housekeeping. If it's warm, they'll find it faster.
Anything small that depends on air flowing over it will fail in short
order.

The ultrasonic suggestion sounds like it might be more
or less immune to spiders, but then bats will probably try to
mate with it ;-) A plant growing up in the path will kill it too.

mike

--
Bunch of stuff For Sale and Wanted at the link below.
laptops and parts Test Equipment
4in/400Wout ham linear amp.
Honda CB-125S
400cc Dirt Bike 2003 miles $550
Police Scanner, Color LCD overhead projector
Tek 2465 $800, ham radio, 30pS pulser
Tektronix Concept Books, spot welding head...
http://www.geocities.com/SiliconValley/Monitor/4710/

 

[Vladimir Vassilevsky]

I've tried a high tech mechanical method in my home to detect and display
air movement.........a down feather hanging by a cottom thread from a map
pin in the middle of a doorway. It has to be THAT sensitive. Any idea how
to couple that to a variable voltage device? (no joking!) without the
sensor detracting from the devices' (the feather) sensitivity?

Use an ultrasonic Doppler effect device. You can measure the spead below
1cm/sec directly from an airflow.

Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

 

[Dr. David Kirkby]

I need to measure the velocity of very low airflow.........less than 1/2
foot per second or so. When I say measure, I mean an analog signal or that
I can store in a data logger and calibrate against some known. This is not
a switch application.

I have an Extech "hot wire" anemometer but aside from the fact that it does
not measure airflow at the velocities I need, the sensor "head" is
directional.

I know some sketchy details of a way that has been successfully used to
measure air speed (not velocity) in and out of the lungs, which
admittedly is much faster, but I don't think that would stop the
principle being applicable to you.

This used a wire (not sure what, but platinum would probably be okay)
through which a large (I suspect a few Amps) short-duration (I suspect a
few ms) current was periodically pulsed. The temperature rise in the
wire heated the air. A second wire (probably platinum) some distance
downstream of the first was then used to sense the temperature rise. The
actual temperature rise is unimportant, but the velocity can be computed
from v = distance/delay, where delay is the time difference between the
current turning on and its effect being measured.

In your case, the speed is about 0.1 m/s (sorry I can't be bothered to
work in feet). So if there was a gap of say 5 mm between the wires, you
would expect the signal to be delayed by 50 ms.

Of course if the two wires are not orthogonal to the airflow, the time
delay experience will be longer. I suspect if you mounted 2 such
instruments at right angles you could determine the velocity, not just
speed - assuming airflow is parallel to the ground. If not you would
need a 3rd sensor to truly determine velocity.

If you think this might be useful, contact this person.
http://www.medphys.ucl.ac.uk/~ddelpy/
who will be able to tell you more about it.

--
"The day Microsoft makes something that doesn't suck is probably
the day they start making vacuum cleaners." -Ernst Jan Plugge.

Dr. David Kirkby,
Senior Research Fellow,
Department of Medical Physics,
University College London,
11-20 Capper St, London, WC1E 6JA.
Website: http://www.medphys.ucl.ac.uk/~davek
Author of 'atlc' http://atlc.sourceforge.net/

 

[Jim Basinger]

I've tried a high tech mechanical method in my home to detect and display
air movement.........a down feather hanging by a cottom thread from a map
pin in the middle of a doorway. It has to be THAT sensitive. Any idea how
to couple that to a variable voltage device? (no joking!) without the
sensor detracting from the devices' (the feather) sensitivity?

Jim

 

[Kevin McMurtrie]

I need to measure the velocity of very low airflow.........less than 1/2
foot per second or so. When I say measure, I mean an analog signal or that
I can store in a data logger and calibrate against some known. This is not
a switch application.

I have an Extech "hot wire" anemometer but aside from the fact that it does
not measure airflow at the velocities I need, the sensor "head" is
directional.

I have seen some ideas for using self-heated transistors, but don't know if
this type of idea is sensitive enough.

I think I have seen a design using a piezo sensor of some sort that could
sense the wafting of a hand over the sensor, but I don't remember if it
output an analog (relative) signal or not.

The application for this device is to measure the air movement at a forest
floor to about 6 inches from the ground. The sensor and data logger would
be left in place to collect data, since a persons presence might have a
large effect.

Any ideas on how I might build this?

Thanks
Jim
[email protected]

The ultrasound idea might work. A very high frequency will you get
bigger phase shifts. A phase detector between two microphones on either
side of the ultrasound source should show the air motion. Hetrodyne
conversion could drop the frequencies to enable use of a simple low
frequency phase detector.

Monitoring drift in a column of hot air would really get you top
sensitivity. I see the power source as a limiting problem.

 

[Michael]

You can find commercial units here http://www.gill.co.uk

They have to be working on 'time of flight'.

I did a search and had a quick look at some ultrasound sensors, and I
agree it's the best way to do it.

If you don't have any experience with ultrasonic transducers and its
in the forest then it wouldn't matter if you used audible frequencies.
You could send a short tone and measure the phase difference between
your source and microphone.
You would need to incorporate some temperature compensation, as this
changes the speed of the sound, and if you used audible frequencies
you would have to allow for the occasional bad reading caused by a
bird trying to have a conversation with it.

 

[John Fields]

I need to measure the velocity of very low airflow.........less than 1/2
foot per second or so. When I say measure, I mean an analog signal or that
I can store in a data logger and calibrate against some known. This is not
a switch application.

I have an Extech "hot wire" anemometer but aside from the fact that it does
not measure airflow at the velocities I need, the sensor "head" is
directional.

I have seen some ideas for using self-heated transistors, but don't know if
this type of idea is sensitive enough.

I think I have seen a design using a piezo sensor of some sort that could
sense the wafting of a hand over the sensor, but I don't remember if it
output an analog (relative) signal or not.

The application for this device is to measure the air movement at a forest
floor to about 6 inches from the ground. The sensor and data logger would
be left in place to collect data, since a persons presence might have a
large effect.

Any ideas on how I might build this?

---
You've got yourself a tricky little problem!

At the outset, whether you try to solve it mechanically or electrically
you not only have the wind velocity to contend with, you've also got the
temperature and humidity to consider.

Just for fun, considering a mechanical solution, (and neglecting
temperature and humidity) I found

P = 0.0042Va²

in my seventh-edition Reference Data for Engineers, where:

P = the pressure in pounds per square foot, and
Va = the actual velocity in miles per hour.

Since 15MPH = 22 FPS, your 6" per second will yield

15MPH xMPH
------- = ------ ~ 0.341MPH
22FPS 0.5FPS

and the force that wind will exert normal to 1 square foot will be

P = 0.0042*0.341² = 0.00049 lb. ~ 0.008 oz. ~ 0.22g, which comes out
to about 2.4mg/in², or 372µg/cm². Not much, and you've still got the
problem of directionality to contend with!

What I'd try would be a vertically mounted Savonius rotor on a taut-band
suspension with a d'Arsonval movement used to counterbalance the force
trying to move the rotor. That way, the current through the coil needed
to keep the rotor "parked" would be proportional to the force of the
wind. Kind of like if you blew gently on the pointer of a meter with a
d'Arsonval movement and moved it to half-scale, the current through the
coil required to return the pointer to zero would be proportional to the
velocity of the airstream; the harder you blew, the more current it
would take. The Savonius rotor is inherently non-directional and would
mostly take the angular error out of it, (I think) and with large enough
vanes would probably work.

For an electrical solution what I'd do would be to use two matched
thermistors. One in stagnant ambient air (the reference thermistor)and
one in the airstream. I'd use the reference to determine the heat
capacity of the air at that temperature by putting a pulse with a known
power into the reference and measuring the time needed to get back to
ambient (or some arbitrary temperature above ambient so the measuremoent
wouldn't take forever), then I'd put the same pulse into the one in the
airstream and measure the time it took for that one to get back to the
same temp as the reference. The difference in time would be due to the
heat transport away from the thermistor due to the moving air, the
shorter the time the faster the air.

 

[Phil Hobbs]

I need to measure the velocity of very low airflow.........less than 1/2
foot per second or so. When I say measure, I mean an analog signal or that
I can store in a data logger and calibrate against some known. This is not
a switch application.

One very sensitive method would be to generate a brief dc corona discharge once per
measurement cycle. Put the corona point inside a square box made of wire mesh, with
each wall connected to the summing junction of an op amp, with a small capacitor for
feedback. The wind will force the ion cloud to drift, which will cause a
differential displacement current in the screens. The screens can be wide-spaced
fine wires, so as not to disturb the air flow. The nice thing is that you see the
current the whole time the ion cloud is moving, not merely when the ions arrive at
the screen--it really is a displacement current sensor. It will also get more
sensitive the slower the air is going.

This is not nearly as small an effect as you might think--I once built a sensor for
testing air ionizers used in clean rooms, and it was very easy to see. You can reset
the integrator capacitors with the trick Win posted here some time ago--inverting the
power supplies (with suitable series resistors) to make the protection diodes
conduct.

The transient from the corona generation will die away much more rapidly than the
displacement current, so you can easily reject it, e.g. by keeping the amps reset
until after the transient is over.

This will give you the wind direction as well as its velocity, though the problem
with insects and animals is very much to the point in all these measurements.

Cheers,

Phil Hobbs

 

[Barry Lennox]

I need to measure the velocity of very low airflow.........less than 1/2
foot per second or so. When I say measure, I mean an analog signal or that
I can store in a data logger and calibrate against some known. This is not
a switch application.
snip

Any ideas on how I might build this?

There are some super-tiny thermistors available, that when heated to
something above ambient, will give a good indication of low velocity
airflow. They have been used in very sensitive variometers in gliders.
Fenwal make some and Newark used to carry them. No too cheap though,
IIRC, the tinest ones, about .05' dia, were about $11 each.

Barry Lennox