Home Automation Domotica Forum Europe, Bwired Forum

Hello,

I just finished a prototype of an ultrasonic wind speed and direction sensor. A year ago I started the project, but after disappointing results I dropped it. Picked it up again early this year and have the prototype working now. The results are excellent: resolution (and theoretical accuracy) of about 0.1 m/sec. I'm transmitting the results by means of X10, so I only have to connect the system to 220 Volts.
The unit has a (programmable) threshold build in; if exceeded, an X10 OFF command for a (programmable) address is issued. It was meant to close a sun screen. But I replaced the sunscreen with a fixed roof last summer, so I have to find another item to control.
I have to put some more effort in the mechanical structure and the housing for the electronics to make it weather proof.

<b>Download Attachment:</b> [img]images/icon_paperclip.gif[/img] ultrasonic.jpg<br />120.15 KB

The sub project also finishes the main project "building weather sensors with X10 interface". I have working units now for sensing temperature, pressure, light intensity, humidity, wind speed and wind direction.
I have some other things I want to put effort in this year: Controlling LEDs by means of X10 and Zigbee communication.
If other forum visiters are working on these subjects, let me know; maybe we can exchnage information.


Han Alink

Han --

The project looks great. It definitely solves the standard problem with cups/vanes --> debris, snow, ice, birds getting in the way of data.

Instead of X10, I was thinking that a 1-wire solution might make sense. I have some idea how much time/cost it would take to read sensors and log results.

I am curious on the cost of the parts. The production versions of these are often $3,000+. Can you give us an approximate cost for the hardware, sensors, etc?

Thanks!
Politics

ps -- sorry 'bout the mail

Well, there are some things that I have to look at.
<ul><li> I'm not sure yet about the behaviour in heavy rain.
</li><li> If a bird is sitting in between the sensors, the unit will not work. </li></ul>The X10 interface has several advantages:
<ul><li> no additional wiring </li><li> autonomic transmit of data </li><li> direct interface with other X10 equipment, so direct control of a sunscreen </li><li> programming the threshold and the address to be transmitted by means of X10 commands </li></ul>A 1 wire solution is indeed something which I have been thinking of. It should not be to difficult to add.

The parts are not very expensive. The total part cost is less than € 100. That does not include the PCB. I did the PCBs myself (one layer).
The problem (costs) is the mechanical construction.



Han Alink

True, but conceivable the ultrasonic pillars can be mounted in a way that makes it unappealing for a bird. How much does accuracy decrease when you move the pillars closer together? What's the power draw? Could this conceivable be battery driven w/a wireless interface (wireless X10, bluetooth, wifi) for a remote installation?

Any plans to create a PCB schematic or parts list?

The influence of the sensors itself, blocking free flow of air, will be responsible for a decrease of accuracy. That's also the reason that I called it theoretical accuracy. It's a mathematical calculation and based on free airflow. If the real accuracy has to be determined, you probably need a wind tunnel to do measurements.
When moving the transducers closer to each other, the resolution will decrease.
I have not designed the unit with power consumption in mind (An X10 interface is always connected to 220Volts). One of the factors for power consumption is the firing frequency. The lower, the less power.
I have been thinking about a Zigbee interface; do not know enough yet to comment on that.
There is some interest for a commercial version. Which means I have to do a full blown PCB. I have not made any decision yet.


Han Alink