Arduino Based Weather Station

[Gina]

Update

[Gina]

I don't get a chance to read much if the forum these days but your threads are always fantastic. I love your designs and the way you think things through.

Keep up the good work gina your an inspiration.

How's your roof automated? I've forgotton what you did on that, I'm about to put a gate ooener on mine.

Sent from my GT-N5110 using Tapatalk

Thank you Neil

Roof isn't automated yet.  I need to get my 3D printer working better to make some brackets.

[Gina]

The replacement control board for my printer has blown again    

[Gina]

I shall be printing the parts for my weather station shortly with my new UP Plus 2 printer when the black ABS filament I've ordered arrives - Monday or Tuesday they say

I shall also be printing the brackets for my roll off roof

[Gina]

I said I'd get back to this in my CCTV thread so here goes...

I have the wind direction encoder base printed.

[Gina]

I think there are parts of this project I can advance further without having to go outdoors so should fit in with my current health problems.

I should be able to finish the wind vane position encoder and the anemometer either needs repairing or rebuilding having fared badly in the storms (lost a bucket).  I think a rebuild using 3D printed parts rather than bottle tops and hot melt glue would produced a rather more robust unit

[Gina]

I'm back to thinking about this project with my All Sky Camera nearly finished.  The work on the latter has provided information regarding the mounting of the wind instruments mast.  I've decided that attaching this to the observatory is not really practical for two reasons - firstly, the attachment of the mast to the building with the clearance needed for the roll off roof and secondly, I would prefer it to be much further away from the all sky camera than it is at present and if attached to the obsy, it would be nearer!

This would indicate moving the weather station mast away from the obsy.  The problem is where else to put it with the trees around and height of sheds.  I think I must face the fact that there is nowhere on the property that is a good place for the wind instruments.  The best place was where the obsy is and astrophotography takes precedence over weather recording

I have had a mast on a corner of the goat shed in the past when I was running two experimental weather stations and I think this must be second choice after the observatory.  I'll post some photos later to show what I mean.  1-wire has a maximum range of around 100m using CAT5E cable so the wind instruments don't need to be near the rest of it.

Edited May 16, 2015 by Gina

[Gina]

Here's some photos showing where I'm thinking of putting the mast.  As you can see, there's a fair bit of storm damage to be repaired apart from playing with weather stations etc.    Plastic wall panels and a whole door and frame.  There is also a hole in the garage roof and several other jobs around the place so it won't be all astronomy related this summer!

Edited May 16, 2015 by Gina

[Gina]

One thing the current mast does is place the wind instruments where thay can be seen while I'm sitting on the settee in front of my desktop PC through the smaller east facing window.  So even without the electronics working I can see the wind direction and strength   This means I want to get the new station up and working plus a local display before removing the existing mast.

[Gina]

The present weather station wind instruments and mast are getting really annoying in the view from my All Sky Camera and I know of at least one other member here who doesn't like it.  I must admit, it certainly spoils the view   To cure this problem I need to get the wind instruments and mast moved away and a suitable display produced.  So to advance the All Sky Camera project I need to add better dew heating to the thing itself and work on the weather station as well.

Next job for the weather station is to locate all the bits I have made already and see where I am with this project.  As I recall and from re-reading this thread, I don't think it should take too long to get something working to the point where I can move the mast.

Reading the wind speed is simply a matter of counting the pulses from the anemometer (or measuring the period between pulses, depending on which works best) and direction is to take the Gray code and converting it to binary.  Displaying the results is simple using the Arduino IDE Serial Display in text, more complicated producing a graphical display and probably more difficult to provide an physical wall mounted display but a nice thing to have - not relying on using a PC.

I have had several ideas for a wall mounted display.  Probably the simplest would be LEDs - 16 for the wind direction and probably more for the speed (or a large digital display).  The disadvantage with this is that it wouldn't show up very well in bright daylight though I could use ultra-bright LEDs and dim them for evenings.  I would prefer a nice big analogue display rather like a large clock for wind speed and direction.

Edited May 20, 2015 by Gina

[Gina]

Thoughts on a large analogue display for wind speed and direction.

This form of display is rather like a derivative of my large wall clock project (currently on hold).  I'm thinking of a large hand for wind speed and small hand for direction.  Temperature and humidity displays would be separate. 

I'm considering a micro servo motor for the wind speed and whilst 180 degrees would be adequate, I would prefer something like 270 or 300 degrees.  I think this could be achieved using gearing up - these little servos are very powerful for their size and the load of a 3D printed plastic hand would be pretty light. The resolution would appear to be 180 steps using the standard Arduino servo library - viz. 180 degrees in one degree steps, which would be quite adequate for this application.

The direction is another matter and requires a full 360 degrees.  It also requires the capability to turn more than the 360 degrees. eg. if the wind direction goes from NNW to NNE it wants to turn 45 degrees clockwise rather than turning 315 degrees anti-clockwise.  This indicates using a stepper motor such as the little 28BYJ-48.  (Or maybe a motor and encoder or even a purpose built 16 pole induction device similar to a stepper motor.)

Edited May 20, 2015 by Gina

[Gina]

I have the base part made for the wind vane with compartments for the photo-sensors and I have these too.  I shall print a new Grey code disc and the parts that take the vane and its bearings.  The electronics will be what I used before, so that's done   Just needs the photo-transistors connected up.

The anemometer needs a proper re-build using 3D printed parts instead of bottle tops and hot-melt glue.  I'm wondering whether to replace the reed switch with a Hall effect device.  Reed switches are magnetic and produce a small but significant drag on the anemometer as the magnet goes past.  I might also use two tiny magnets instead of one to improve the balance.  One reason for using reed switches originally was that they take no power but since I will be providing a separate 5v supply, this is no longer important.

[Gina]

I was going to use pin pong balls cut in half for the buckets of the anemometer but then thought - hey, why not 3D print them instead?   Ping pong (aka table tennis) balls do not hold their shape well when cut in half.  So I have been experimenting.  Started with designing a single bucket in SketchUp and printing it out.  I tried 0.5mm thickness to start but I think that's pushing it  Two problems in fact - SketchUp refused to make a full hemisphere using FollowMe - left a hole in the bottom.  Then the 3D printer couldn't cope with 0.5mm wall thickness.

Hole filled with discs.  Probably wouldn't make much difference from a perfect hemisphere.

[JamesF]

My anemometer uses conical "buckets", so I'm not sure shape makes much difference.  I guess as long as they're identical then it shouldn't be a problem.  They're more tennis ball sized than ping pong ball sized too.  I wonder if you couldn't just use closed-off cylinders.

James

[JamesF]

Closed off cylinders might be easier to mount too, come to think of it.

James

[Gina]

I think you're probably right James   Though the shape might alter the calibration but calibration is virtually impossible anyway - there are far too many variables.  You could calibrate the anemometer by fitting it to a vehicle and travelling at fixed speed on an empty straight road with no wind - ha ha   No wind occasionally happens but empty straight roads... unlikely. 

In a domestic environment there will be all sorts of things that upset wind speed and direction at any particular location.  And if you measure the wind speed in one place, chances are the speed elsewhere will be different.  So I think an anemometer is mainly for interest and just to get a rough idea of the wind force.  So trying to make an accurate model based on a commercial instrument is a. pretty hopeless and b. little point in reality.

The upshot of it all is that I agree that virtually any shape will probably do that has more wind resistance from one side that the other. 

I would like to print the rotor in one piece but I'm not sure if this is practical.

Edited May 23, 2015 by Gina

[Gina]

On Googling anemometers, I've found Maplin do a replacement anemometer for their weather station (which I used in the past) for £2.49 plus P&P.  Makes me wonder if it's worth making one   Is it really important that I can say "I built all of this myself from scratch"???  I doubt it.  The time and effort to design and make an anemometer (including severable unworkable attempts) is not inconsiderable.

Edited May 23, 2015 by Gina

[Gina]

Maplin total is £5.58 = £2.49 + 3.99 P&P.  I'll think about it  

One reason for making my own was to produce a more robust device.  The Maplin anemometers tend to shed their buckets in a bad storm.  I also found I needed to add a skirt to the rotor to stop wet and hence ice getting in and causing it to sieze up in winter.  I have had 2 complete FineOffset make weather stations from Maplin in the past both of which are defunct in various ways.  The latest anemometer is still on the mast but with one cup gone.  The other has 2 cups gone and several repairs.  Of course, I don't expect much from these cheapo Chinese weather stations

Edited May 23, 2015 by Gina

[Gina]

Just modelled and printed a conical cup/bucket successfully   Much easier than a hemisphere.  A strong connection to the hub should be much easier too

[Gina]

Here is a screenshot of a complete anemometer rotor model except for the hub part.  It is too big for my present working 3D printer, unfortunately, though could be scaled down.  I have crafted the spokes to offer less wind resistance where the cup is back to the wind though I doubt this is actually necessary.  This rotor is intrinsically far more robust than the FO version

As designed it won't fit the printer.

Scaling by 80% and it fits fine.  The yellow pattern on the base shows where the support material will start.

[JamesF]

I'd have thought you could print the cups and rotor separately and screw them together.

James

[Gina]

Yes, I think that's what I'll do

Edited May 23, 2015 by Gina

[Gina]

Single bucket and spoke.

[Gina]

The single bucket has printed well and it's very strong - I can't break it with my fingers!   At least this project has turned out well tonight unlike the all sky camera

[Gina]

Modified anemometer design including hub.