Pic based Wind Speed meter

[Bill Legge]

I've always wanted to have a go at a hot wire anenometer. I presume they work by the cooling effect of the wind changing the resistance of the heated wire? Gets around all the mechanical and weather-proofing problems?
Apart from that, I'd:
1. If it's for the fun/learning experience, try the slotted/disk method.
2. If you want it to work properly, outside and for a long period - byte the bullet and purchase the Davis sensor at about $125.

Regards Bill Legge

[Bobbo_ZA]

Agreed, the weather proofing of a homemade "stick and cup" versionis an issue, however I have managed to source a commercial replacement set thatis UV treated etc., so that should solve the longevity issue.

I was thinking last night, would it not be possible to still use the magnetand read switch and then use pulsein. From what i was thinking pulsein shouldgive me a frequency, i don’t know if it’s possible to convert the frequency toa usable value that i can do some calculations on.

Also, in this I'm also not 100% sure on the wiring for getting it to give mea frequency, I know the read switch is normally open, so would i connect it topull the pin high (v+) or connect the switch to ground?

[kellyseye]

Commercially available wind speed and direction (WS&D) indicators are common-place onboard sailing craft of all sizes. The wind speed is invariably measured by 'cups' on a rotating shaft passing a magnet across a hall effect sensor. One device I recently repaired had the cup unit changed and the cups were rotating on a radius of some 50mm so if manufacturing one yourself then don't go 'too big'.
Similarly, from what I could gather from the internals of a WS&D I recognised a Bournes 'potentiometer' that (I think) operated using hall effect. I keep meaning to look this device up but I can only assume it outputs a voltage proportional to the angle of the potentiometer shaft (which rotates through 360 degrees) - perhaps a search at the Bournes website might reveal the actual devices used?

edit - a quick search at Bournes reveal rotary (360 degree) position sensors but nothing based on hall effect that I can see immediately. Anyone else seen/heard of devices like this?

[Dave]

Bobbo_ZA, Why not use some OLD bell's from a couple of OLD phones. I did this years ago and it is still working as far as I know.

Dave Purola,
N8NTA
ALLSUNTRONEX.COM

[Charles Linquis]

Please note that a potentiometer has a limited life, even if it is a good one.

A "hot wire" anemometer requires a Whetstone bridge and an amplifier. The hot wire itself is generally fragile. I assure you that the PTC method I outlined above will be easier to implement.

[cncmachineguy]

Charles, I for one really like the wire method- no moving parts is always good. Is there anything needed to know about the wire mounting? Ie position, length...

[Charles Linquis]

I'm advocating that you NOT use the traditional hot wire approach. Most of the hot wire sensors use either a (special) purchased wire, or have you break a small incandescent bulb to get the wire. Both are relatively fragile. When I did this project I found some small (approx 3/8" diameter X 1/8" thick) PTCs in the DigiKey catalog. I don't see them there anymore. You could perhaps use one from

http://www.specsensors.com/custom/co...er.asp?thm=srh

The idea is that they have a low resistance so that when you apply 5V you can get enough power into them to easily heat them up to their "knee" temperature. The ones that I chose had a 5 ohm resistance and a knee temperature of 110C. When you apply 5V across them, they initially draw 1A (5Watts), which quickly heats them to their knee temperature, at which point their resistance rapidly increases and the current falls. If you try to cool them (by moving air across them), the current draw increases. Because of their self-regulating nature, they draw as much current as needed to keep their temperature at 110C. If you put a small (.1ohm or so) resistor in series with the PTC thermistor, the voltage drop across that resistor is proportional to the cooling effect of the air - which is proportional to the velocity. You do have to amplify that voltage in order to get it p to a good range for the PICs A/D converter.

The airspeed to voltage drop across this resistor is not a linear function. Also, be aware that you must compensate for ambient temperature, since a 20 MPH wind at 0C will cool a hot thermistor more effectively than a 20 MPH wind at 30C, so you need to add a second temperature sensor to measure ambient and use that in a calibration routine. I can't really give you the calibration code that I used because I'm certain your thermistor will be different than the one I used.

Still, with a little work and some calibration, the method outlined performs very well. I have used it to measure air velocities in electronic cabinets.