Arduino Based Weather Station

[Gina]

I'm thinking of a considerable change in the design of the casing for the wind sensors.  The anemometer pillar was designed to bolt onto the aluminium saucepan lid and now that I'm not using that there is no reason to keep that design.  It would make more sense to have a proper combined design approach and a mainly one piece casing except for an access plate or cover.

[Gina]

Looks like the casing will have to be in several parts.  The rotation sensor chip and its breakout board can be housed in the pedestal that supports the anemometer making a rather fatter body than the original but permitting a good length for the axle for the wind vane.

[Gina]

Breakout boards from Amazon are 11.12mm x 10.38mm which gives a diagonal of sqr-rt(123.6544+107.7444) = 15.2mm.  That will easily fit inside a reasonable sized pedestal for the anemometer.

[Gina]

A "prettier" design

 

Edited December 8, 2017 by Gina

[Gina]

This diagram shows how the rotation sensor may be mounted together with the bearings and axle system with all the parts designed for 3D printing without needing support material.  The pedestal for the anemometer will go over this and the three wires for the Hall sensor in the anemometer can also go through the hole in the casing where the rotary sensor wires go.

[Gina]

Weather station wind sensors.  Cross-section view and general appearance.

Edited December 8, 2017 by Gina

[Stub Mandrel]

Took apart my anemometer, after two half-hour sessions in an ultrasonic cleaning tank with metal cleaning fluid to free up the two parts. Grim!

It was quite nicely made with a circular PCB and a hall-effect sensor standing on its legs, the magnet seems to have dropped off and disappeared under the brutal force needed to free everything up. One ball race is solid, the other grittier than Rooster Cogburn...

It would be nice to salvage it and fit a new anemometer cup assembly, 3D printed with a better 'skirt' to protect the bearings from rain - and taper the end so any dew rolls away from the bearing.

I opened up the 'control box' - covered in algae and several moss colonies. inside a few connectors, an orphaned humidity sensor, a nasty spider and a hibernating (or dead) harlequin ladybird...

 

So... my advice to you is think about moisture and insect ingress, and plan on a strip down for maintenance every year or three!

[Gina]

Oh yes, both those are well in the forefront of my mind, both for this and my all sky camera.  Nice long skirts to stop driving rain and I have a spray I use to keep moisture off the electronics (no not WD40).  Forget the name ATM.

[tekkydave]

I cheated  and bought a quality wind vane / anemometer from Hobby Boards in the U.S. Sadly they have since closed down.

It's been outside on a mast since 2012 and no signs of deterioration in operation or performance (maybe shouldn't have said that).

Edited December 9, 2017 by tekkydave

[tekkydave]

These are the parts on the manufacturers website

Wind vane

Anemometer

Not sure if the anemometer is exact model I have - mine has a hall effect sensor not reed switch.

They came with a 1-wire interface card which is sited in my external box containing the RPi.

Looking back at the order email total cost was $245 at the time (Feb 2012). I had a good job back in 2012 so money wasn't an issue. Couldn't afford it now!

 

[Gina]

Yes, Hall effect sensors are much better than reed switches as they don't influence the motion of the magnet.  I found a reed switch stopped the anemometer rotation below about 2mph or so.  I have three tiny Neodymium magnets in mine with a Hall effect sensor and it works down to 1mph wind speed and produces 1pps at that speed.  Apart from magnets, Hall device and the bearing it's all 3D printed from ABS.  I bought items from Hobby Boards a few years ago.

I see your wind direction sensor produces an analogue output, did you have any problem with the calibration around the 359/0 changeover?  I considered an analogue output Hall sensor chip but though of that problem and went for a digital output.

[tekkydave]

I've not had any noticeable issues. I've just followed the manufacturers quoted voltage range in their instructions. It claims an accuracy way above what is achievable in practice. After all it could be several degrees out in physical alignment to North.

At the time I bought it the cheap AAG weather station I had had just died due to water ingress. I wasn't happy with just having 16 point wind data and it was the only analogue one I could find on the market.

[Gina]

I started with a weather station from Maplin - the ubiquitous Fine Offset model.  Had all sorts of problems with spikes in the temperature graph and also with the mechanical reliability generally.  Tended to shed cups from the anemometer, seized up, etc.  I bought a second one but that gave up too so decided to design and build my own.  Initially based on the Arduino and 1-wire devices and several arrangements of optical Gray code wind direction sensor with 16 states.  This latest version is almost a complete rebuild though I may keep the Arduino data logging system as that seems to work well.  I'm hoping to get the weather data on my web site again eventually plus the ASC images.

[Gina]

SOIC-8 breakout boards, filament and spacers for the wind vane axle are out for delivery

[Gina]

Arrived

[Gina]

Here's a photo of the wind vane assembly with axle, bearings and damper disc balanced on top of a pot.  I'm working on the magnet holder ATM, trying to get the hole sizes right

[Gina]

I'm not going to attempt to solder the SOIC-8 chip onto the breakout board tonight - I'll leave that for daylight and a freshed head   May design the printed parts though and print them though.  Already started with the magnet holder actually

[Gina]

Magnet holder finished.

[Gina]

Here's a modified cross-sectional diagram showing the magnet attached to the bolt head of the axle with the sensor chip and breakout board above it plus a close-up photo of the breakout board sat on top of the magnet & holder.  For an idea of scale, the ball bearings are 10mm OD.

[Gina]

Looking into soldering the direction sensor chip onto the breakout board and the connections taking data from the MLX90316 Rotary Position Sensor IC datasheet.  Below is a photo of the chip on the breakout board (not soldered yet) as close and well focused as I could manage   followed by screenshots from the datasheet.

[Gina]

Chip soldered to breakout board  

[Gina]

Holder for top bearing printed satisfactorily.  Now printing the anemometer pillar (just for a change ).

[Gina]

This pillar is actually 20mm shorter that the earlier one but still looks higher than needed.  Very plain, just a cylinder but it doesn't really matter what it looks like at 6-7m above ground.  The inside of the pillar tube well clears the the wind vane sensor assembly and leaves plenty of room for the wires (multi-way flat cable).

 

Edited December 10, 2017 by Gina

[Gina]

Printed a new pillar half the height.  I think this looks alright.  Anyone see any problem with this?

[Gina]

Two possible colour schemes depending on which filament I use for the main casing.  I'm running a bit low on the green.

Edited December 10, 2017 by Gina