Hi All,

I'm working on hobby project to build a weather station from 3d printed parts + diy electronics. I already got the wind speed, temp and humidity sensors working, but now I'm looking into how to implement the wind vane. I come across the 8 reed relay or some kind of magnetometer to determine direction. Any other clever idea?

Thanks
Zsolt

 

I saw one of these in Wyoming. Four feet of heavy chain. It points out wind direction and speed.

 

I saw one of these in Wyoming. Four feet of heavy chain. It points out wind direction and speed.
View attachment 328974

Is that windy over there?

 

I use a La Crosse Technology weather vane.

 

Hi All,

I'm working on hobby project to build a weather station from 3d printed parts + diy electronics. I already got the wind speed, temp and humidity sensors working, but now I'm looking into how to implement the wind vane. I come across the 8 reed relay or some kind of magnetometer to determine direction. Any other clever idea?

Thanks
Zsolt

Mount your windvane of a quadrature encoder with sufficient steps for your needs.

 

Is that windy over there?

Yep! When the wind does die up there, people fall down because there's nothing to hold them up.

It's a great place to get good at crosswind landings!

 

Melexis offers a range of ICs for measuring the angle of a magnetic field

 

Mount your windvane of a quadrature encoder with sufficient steps for your needs.

A quadrature encoder generally only gives relative position.
He needs an absolute encoder.

 

A pair of small selsyn motors, one outside with a vane to rotate it with the wind, the other inside with a pointer. You get an exact reading to the smallest resolution. . The down sides are needing a 4 or five wire cable, and needing an AC supply to update.
Another is four thermistors arranged to be self heating and cooled by the wind flow. The downside is complex math to calculate the direction.

 

A quadrature encoder generally only gives relative position.
He needs an absolute encoder.

There are several ways of dealing with that without using the usually much more expensive absolute encoders (though I haven't checked prices for many years). Many encoders have a reference output at one position, so you only have to manually turn it through one revolution for the firmware to index it.

 

There are several ways of dealing with that without using the usually much more expensive absolute encoders (though I haven't checked prices for many years). Many encoders have a reference output at one position, so you only have to manually turn it through one revolution for the firmware to index it.

Okay, so if you lose power, you need to wait until the wind moves it past the reference point to reset the index.

 

Okay, so if you lose power, you need to wait until the wind moves it past the reference point to reset the index.

I said that there were several ways of dealing with the indexing and gave just one example. The TS hasn't said where the unit will be employed, and that plays a role in a number of decisions. If it is going to be remotely located and inaccessible for long periods of time (e.g., mounted on a roof), the decisions will be different than if it is going to be a portable unit that is easily accessible (e.g., set up as part of a model rocket launch site).

 

Is that windy over there?

After 4 days of 60mph winds you go nuts and make thinks in the barn. More likely the wife sends you to the barn. The chain is welded together so it stands out like that even indoors. (joke) People from California believe anything.

 

Hi,
something like this ?

 

Hi,
something like this ?

It would be still required to turn it around once to get accurate position, right?

 

It would be still required to turn it around once to get accurate position, right?

No. Such device can provide absolute position.
How many bits (resolution) depends of number of 'rings' on optical plate and opto detectors.

 

For wind direction and speed you can use three ultrasonic sensors in an equilateral triangle with an ultrasonic sender. I recall seeing an instructables project that shows how to do this

 

Another simpler scheme doe not involve an encoder, but rather a surplus potentiometer that allows continuous rotation. But like I posted before, a pair of surplus selsyn motors is the best scheme.
If you can get an encoder like the ones from industrial robots that can work very well. They are incremental with a battery backup and report position in a digital format, so just a few wires needed. They have very good resolution, and are extremely rugged. But you will need some logic to translate the digital output to a useful display.

 

I would opt for four reflective photol sensors.

One to differentiate N-S orientation.
One for E-W orientation.
And relative movement detection (two pairs of sensors).

 

I come across the 8 reed relay or some kind of magnetometer to determine direction.

He needs an absolute encoder.

He never said that.

Any other clever idea?

Depending on the level of accuracy you wish you can easily go with a series of photo sensors (IR receivers) arranged in a circular pattern below a single IR LED. To get power to the LED you can implement induction similar to wireless cell phone chargers. Each receiver can represent a given direction. Suppose the wind is coming out of the North West - and suppose you've arranged 24 receivers in a circular pattern beneath the movable LED. With one receiver at a time active you can get a resolution of 15˚. If there's room between the receivers and the transmitter it may be possible to get three indications. Suppose with that North West wind sensor #9 would be indicating a wind approaching from the North West (315˚). If #9 AND #10 sensors are indicating you can split the difference between 135˚ & 150˚ and come up with an estimate of 142.5˚. That may not be as accurate as you wish, but it should give you a fairly decent idea of the direction from where the wind is coming. No need to reset or zero out every time you power it up.

With 24 inputs you should be able to program a microprocessor at the sensors to transmit the information to your monitoring station. I think the same sort of thing may have been the approach using reed switches. The magnet position activates one or more reed switches. You're not going to get high accuracy unless you spend the bucks for something far more precise. I take it this is a hobby project; something you would like to accomplish more so for the accomplishment than the practical use of a weathervane.

That's my take on the TS post.