Hi, I have a question, I am working on a weatherstation (wind and speed), but struggling with the mechanical parts. (moisture ,rust etc) So what do you think of the following idea. Use 4 pitot tubes (90 degrees) with air pressure sensors, and calculate with the four readings the windspeed and direction. (perhaps even possible with three and 120 degree angle). Use PVC tube with one end closed.
Anemometer with pitot tubes and differential pressure sensors
(10 posts) (6 voices)-
Posted 02/06/2010 17:42:58
-
Interesting. Should be doable, although calibration might be tricky. You mean 3 or 4 BMP085's? They are a bit expensive, but no moving parts would indeed be a nice benefit.
I've been toying with the idea of using Paul Badger's wind sensor. Works by cooling of a heated part, i.e. convection cooling, with airflow affecting the rate of cooling. Same calibration issue, though.
I bought a KS300 a while back - the data can be picked up as OOK @ 868 MHz. Works well, but it doesn't report wind direction. So another idea was to create an add-on: a wind-vane, mounted on a pole, using solar cells as energy source (as the tail end of the vane?), and perhaps a compass for direction sensing, then the whole unit could be mounted with no exposed electrical parts.
And since you're also in Houten - check out this weather site - I have to admit that with all that info being collected (for many years on end!) and available in public, my urge to build something dropped a bit...
Posted 02/06/2010 19:11:53 -
Could you not mount the BMP085 horizontally and have a vane steered collector pipe with a 90 degree feed pipe to track the wind and direct the air down onto the single BMP085?
Pitot tubes do have mechanical problems too... They can freeze. Aircraft ones have electric heaters in them (pretty powerful ones too). Not to mention at least 3 pitot tubes for redundancy.
As for the old fashioned mechanical methods, old ball mice are a great source of plastic parts, I'm specifically thinking about the optical ball movement encoders for air speed measurement.
Posted 02/06/2010 20:19:28 -
good ideas, I didn't know the weatherstation in houten. I tried years and years ago (+/- 35) a circuit from elektuur with two NTC's and resistors (cooling method). It did not work properly at that time.
Posted 02/06/2010 20:48:46 -
Very nice visualization for the weatherstation in houten!
Using html5 canvas and Javascript I wrote some mittelung.html?GlaettungWind=0.3&GlaettungBoeen=1.0>weather data visualization with interactive zoom. Works quite well with Firefox, a little slow with Internet Explorer.
anemometers>Sonic anemometers for measuring wind direction in 3D (horizontal AND vertical components) got my interests for Paragliding. Any clues which two-way ultrasound transducers to use and were to obtain them ? I am dreaming of a small portable unit with 3D wind sensor, inclination meter, electronic compass, wireless data transfer and solar panel.
Posted 03/06/2010 12:18:39 -
Some details:
And a cheap Kit (1D):
AAAND DIY-Information for a 2D:
Und hier noch auf Deutsch:
Posted 03/06/2010 12:36:35 -
WOW! Thanks a lot. I knew about the other links, but not the cheap kit. The link provides for schematics as well! It uses a dual 4-channel analog multiplexer, demultiplexer ic.
This would be sufficient for a 3D anemometer using 4 transducers in a tetraeder configuration, with each transducer sending to all of the other 3 transducers listening, which would require for a large (non/little-focused) sound field.
With the 6 transducers configuration, which is shown in the other links and seems to be a kind of "gold standard", a 6-channel multiplexer is needed and the sound field could be focused. Any suggestions for a 6-channel multiplexer ?
Any suggestions for which transducers to use ? They should be waterproof. Extra bonus for small. And cheap. But not too cheap. ;-)
Posted 03/06/2010 14:05:09 -
Wolfram, excellent website, special the zooming part. I am not that good with javascript, so I was working with cacti for a few evenings, but without success. I create the graphs with rrdtool but without zooming.:-( regards Jeroen
Posted 03/06/2010 15:59:06 -
Indeed the website is great wolfram! I have been thinking of doing something like that for my visualisation, but I have not idea how :)
Any chance we can get the hands on the code?
Thanks!
Posted 04/06/2010 06:33:43 -
I am not quite sure, if you really want to use this very source code, it is not very pretty. It is experimentally grown code (this really calls for a major rewrite) and my first major attempt to javascript. And it is tailored exactly to the data format of this weather station: Example $DPTAW,2010/06/04,09:34,BDSchriesh,512,2,4,0,0,4,7,,295,1,,15.8,13.1,16.5,,,,,11.5,76,68,90,B,12.8,3D
Only very little is done server side: getweatherbydate.php?weatherdate=... This retreives the original weather data ASCII strings from the server. Everything else is done client side: simply save the page and look at the source code. It is javascript, html 5 canvas using the flot library http://code.google.com/p/flot/ .
Using flot made it easy in the beginning, but then became a real drag since it is very slow and a little hard to customize the look. Next time I would probably use html 5 canvas and do all the painting functions there myself.
I have made some experiments with a virtual trail of smoke. It uses only canvas. Look at the page source. It shows how to do the mouse event handling. Ignore old function variants like drawBig2, drawBig3 etc, focus on function drawBig.
Posted 04/06/2010 07:53:45