This article was first published in 2009.
If you are a sailor, fly a kite or model aeroplane, or just like knowing the weather around you, this anemometer project is for you. The design would also be very useful in the geography or science departments of a school. For those who don’t know, an anemometer is simply a device that measures wind speed. Our battery-operated anemometer has a digital screen that shows wind speeds up to 99 km/h (higher if you wish to spend a little more), can display wind speed in either km/h or mph, has an in-built service indicator, and is very durable. But the most amazing thing is that building the complete anemometer should cost well under AUD$35!
The design uses a cup-type spinning assembly that is rotated by the wind. A magnet positioned on one of the four cup arms triggers a reed switch as it passes, with the output of the reed switch monitored by a combined LCD display/processor unit. The LCD/processor unit is available for incredibly little – you just buy a digital bicycle speedo. These can now be picked up for under $10!
While that’s the electronics out of the way in one fell swoop, the mechanical design is very important if the anemometer is to be both durable and reliable. A lot of effort was put into devising a rotor that would last a long time while being exposed full-time to the elements. Our final design uses stainless steel cups, polypropylene arms, and a dual ball bearing axle.
Sounds expensive and difficult to source? Not really – the axle assembly is the front hub of a bicycle, the stainless steel cups are soup ladles, and the polypropylene is a plastic kitchen chopping board!
The first step is to select the bicycle hub. A visit to any bike shop will reveal a multitude of front hubs – including some very nice alloy ones! Often the shop will have second hand hubs available, and for our anemometer we selected what appeared to be a brand new steel hub from the second hand selection offered to us. It cost AUD$7.
When picking a hub make sure that the axle is able to spin freely but without end-float. If it turns with a “cogging” motion or the grease in the ball bearing area is old and coagulated, don’t buy it. If you live in an area very prone to corrosion (for example, near to the beach), you may wish to splash out and buy an anodised alloy hub or the like.
Either way, make sure also that you get the nuts that go on the axle.
Next, you need to cut out the plastic rotating arm assembly. This must be done very carefully so that the rotor retains good balance – more on this later. The first step is to select a polypropylene cutting board. It should be at least 28.5 x 28.5 x 1.0 cm thick. We purchased a board a little larger than this for $7 from a discount store. The board should be cut so that there are four arms extending symmetrically from a middle hub.
The plastic material “works” beautifully and can be cut with an electric jigsaw or even a coping saw. When cutting out the rotor, used curved corners as these reduce the chance of a later arm fracture. Once cut, the edges can be filed and/or sandpapered smooth.
Next carefully mark and drill the centre hole, starting off with a small drill and then increasing the hole diameter until it matches that of the axle. You can then place the arm assembly on the bicycle axle and temporarily tighten the nut. Spin the assembly and see how good the balance and run-out are. If you have made a terrible mistake and the assembly is way of balance (perhaps because you drilled the hole in the wrong place!), buy another chopping board and start again. If the assembly is only a little out of balance or is perfect, keep going!
The next step is to detach the soup ladle cups from their handles. When buying the ladles make sure of two things – that the cups are actually stainless steel (it’s usually stamped on the ladle) and that the cups can be easily detached from their handles. The ones we used were spot welded to the handles and they broke off with just some wriggling. Rivets (or stronger spot welds) can be drilled out. Our ladles cost AUD$3 each from a discount store but note that you can pay much, much more than this if you buy branded, fashionable ladles. The trick is to look in downmarket stores – not trendy kitchenware places!
The cups are attached to the rotating assembly by self-tapping screws which are about 20mm long. These pass through the cups near to their edges and then screw down into the length of the arm. If you hold a cup up near to the end of the arm, you’ll see that the arm needs to be cut at a slight curved angle if the cup is to nestle comfortably into position. Use a hack-saw and a half-round file to make this curved end. Do exactly the same to each of the four arms.
The holes can then be drilled through the cups to allow the screws to pass through. On the cups we selected, the spot welds used when the anemometer cups had a previous life as soup ladles were still clearly visible. We drilled through one remnant spot weld on each cup. The cup can then be held against the end of the arm, the hole position marked, and a small diameter pilot hole drilled lengthwise into the arm to take the self-tapping screw.
Before selecting this size of drill bit for the pilot hole, experiment with different drill sizes on a scrap offcut of the plastic. The size of pilot hole that works well in the plastic is not the same as you would use in other materials and depends very much on the coarseness of the thread on the screw. Experiment until you find a hole size that best suits the self-tapping screws that you are using. Note that over-tightening the screws will cause the plastic to “strip”, so be careful. For durability, the best bet is to use stainless steel for all of the fasteners used on the anemometer.
If you live in a very windy area and want the rotating assembly to be super heavy duty, you could make the rotor out of thick marine-grade ply. In this case, mount the cups using nuts and bolts, with the bolt passing through the centre of the cup and then through a hole drilled tangentially in the arm. This heavier assembly will be less sensitive to light winds, though.
With the cups mounted and the rotating assembly temporarily on the axle, you can blow on it and make it go round and round. Once you get bored with doing this, hold the axle horizontally and see if the assembly turns until one cup always points downwards. If the assembly is perfectly balanced, it won’t turn – well done! However, if one arm is noticeably heavier than the others, temporarily mark it. This information will be useful in a moment.
To prevent water flowing down into the bearing from above, a shield needs to be mounted just below the rotor. This extends down over the hub without fouling it. A plastic screw-on cap from an old oil container (or similar) can be used to form the shield. When the right diameter cap is found, drill a hole through the middle of it and mount it on the axle under the rotor.
Remove the rotor from the axle before performing the next step. Incidentally, note that dropping the rotor can dent the cups, so care should be taken of it during the rest of the manufacturing process.
Next the hub is mounted to a polypropylene mast stand-off using either saddle clamps or a clamp fashioned from scrap aluminium. We used an offcut of the chopping board for the stand-off, again selecting this material to prevent any corrosion occurring.
The magnet and its pick-up need to be mounted next. Remember how you marked the heaviest cup? On the arm directly opposite mount the magnet. Doing this will help balance the rotor. The magnet can be attached to the arm using two small self-tapping screws, and should be placed with its centre about 55mm from the rotor axle. Using a self-tapping screw, mount the sensor on the stand-off so that the magnet passes directly over it. Leave a gap of a few millimetres between the magnet and the sensor.
You should now be able to spin the rotor and read off a speed on the display. Of course, the speed will be wrong, but the instructions in the next section will fix that! If the assembly is out of balance once this stage has been reached, balance it by screwing small weights to the outer edge of the arm that is opposite the heavy one. Using a cable tie to hold the weight in place can be useful while doing the balancing, but make sure it doesn’t fly off when the rotor is being test spun!
Calibrating the anemometer is best done using a car and a still day. You will need a willing assistant to drive the car while you hold the anemometer out of the window and read off the wind speed on the digital display. The bicycle speedo can be programmed for different wheel diameters and this is the function that is used to calibrate the anemometer.
If the speed shown by the instrument is low, you need to set the wheel diameter to a higher number. If the speed shown by the instrument is high, set the wheel diameter to a smaller number. With the prototype, setting the wheel diameter to its maximum (2999) gave correct measurements.
If you find that you run out of calibration settings at the “large wheel” end, add a second magnet directly opposite the first. The LCD display module will then think that the rotor is spinning twice as fast as it actually is! As a result you will be able to use a reduced calibration number to set the instrument accurately. You will need to re-balance the rotor with the extra magnet in place though.
Note that you should be careful when completing this calibration procedure. At 100km/h the anemometer is spinning very quickly indeed – fast enough to cause injury if your arm were to come into contact with it, for example. You should also hold the anemometer far enough away from the car that the vehicle’s aerodynamics don’t affect the measured wind reading – 50cm should be enough.
Final Setting Up
The prototype was mounted on a 1 metre length of square aluminium tube. Incidentally, if you’re wondering how expensive materials such as aluminium can be used on a budget project like this, I’ll let you into a secret. If you go along to a large non-ferrous scrap metal dealer you’ll find that you can buy (by the kilogram) offcuts of aluminium angle, plate and tube for next to nothing. The metre of tube used here cost about 30 cents!
The figure 8 cable that connects the sensor to the display can be lengthened beyond the metre or so provided. Quite how long you can go with this cable we’re not quite sure, but certainly 10 metres doesn’t cause a problem. If a very long battery life is required, the 3 volt button cell in the display can be easily replaced by an external pair of AA cells and the new power supply wires soldered to the original battery clips.
If you want to read higher wind speeds than the 99.9 km/h available on many cheap bike speedos, select a more expensive unit. Some can measure speeds of up to 200km/h, which should be sufficient for all but tropical cyclone conditions. Incidentally, the prototype was tested at speeds of up to 120km/h without any mechanical problems. For absolute maximum durability, paint the complete anemometer. Some stainless steels will rust if they are of low grade, and all plastics will last better if they are not exposed to UV radiation.
Finally, what about that “service indicator” mentioned in the first paragraph? That’s the odometer part of the display. When it gets to 5000km (or whatever figure you decide is appropriate), it’s time to re-grease the bearings in the axle and check their clearances!