Connect with us

Building an analogue wind direction meter

Discussion in 'Sensors and Actuators' started by JoeSmith, Sep 27, 2018.

  1. JoeSmith

    JoeSmith

    2
    0
    Sep 22, 2018
    Hello All

    I want to build a wind direction indicator. It needs to be more than a mechanical weather vane because our house is built in such a way that I can't see any points outside from inside where a vane can sit freely in the wind. So I'm envisaging a simple and relatively small weather vane outside, probably on top of the house, that transfers its horizontal rotational movement to a vertically mounted analogue dial on a wall mounted plaque inside the house. I understand that to do this I need a pair of syncro motors which makes sense but I don't know the best place to acquire these (I'm currently in the UK) or the spec. I've looked into creating a magnetic system but that doesn't seem to provide enough torque to move a decent sized needle on the wall mounted unit. Does anyone know of any existing plans for such a thing or somewhere I can refer to to build it? I'm a relative novice with electronics and this project is really just something to take my mind off work so it's meant to be fun and has no particular time scale but if anyone can offer me any advice on how to approach it and any other design input, I'd be very grateful.
    Many thanks
    Joe
     
  2. kellys_eye

    kellys_eye

    4,289
    1,143
    Jun 25, 2010
    Many of the amateur electronics magazines have had the run-of-the-mill wind speed/direction projects published in the in the past and there have been numerous iterations of the principle.

    There are the analogue scale devices, simple 'digital' devices and even non-moving parts devices (some sensed temperature differentials over two axes and integrated the result....others used ultrasonics).

    I don't know if you have any requirement for accuracy but a simple 16-cardinal-point device could be made using a magnet and reed switches with a corresponding LED display or, if you were adventurous, scaling resistors and an analogue readout.

    A simple Google search came up with:

    http://archive.siliconchip.com.au/cms/A_102593/article.html
    https://sites.google.com/site/weathermake2000/
     
  3. JoeSmith

    JoeSmith

    2
    0
    Sep 22, 2018
    Thank you for your reply. I've seen the project on the second link but the first link looks interesting. My only reservation is that I want the display to be indicated by a needle rather than LEDs. This is a link to a photo of the dial that inspired me and although I'm not going to create anything as ornate, I do like the way the needle sweeps round the dial. That's what I'm trying to create and it's driving the needle that seems to be the challenge. That said, I'll analyse both the links again and hopefully adjust where necessary. Many thanks.
     
  4. kellys_eye

    kellys_eye

    4,289
    1,143
    Jun 25, 2010
    The 'display' side is and always has been the difficult part of such a device and your original idea to use synchro's is the 'correct' way to go.

    It's possible to use a stepper motor to do this (more readily available) but they require complicated interfacing and most likely some programming too.

    You could get lucky and trawl eBay for a used marine setup to canabalise for bits.
     
  5. hevans1944

    hevans1944 Hop - AC8NS

    3,908
    1,867
    Jun 21, 2012
    A synchro (or selsyn) pair would be a really old-school, 20th Century, way to handle the problem of remote sensing and display of shaft position. Here in the 21st Century we have available "new and improved" ways of doing such things, some of which don't even require a shaft or a rotating weather vane!

    Try this link for some suggestions using hot-wire anemometer-based solutions. Advantage is no moving parts. A hot-wire anemometer uses a thin wire, heated with a constant-power source, to develop a voltage, across the length of the wind-exposed vertical wire, that is proportional to the cooling effect (hence wind speed and the anemometer name) of air flowing past the wire. Use four of these in a bridge or differential arrangement to derive a vector (x-y) representation of wind direction. It should also be possible to reduce the number of anemometer elements to just three, computing the wind direction from the three voltage drops for internal use to drive a wind direction display device. The only requirement is the hot-wire anemometers must all be exposed to the same wind, which should not be a problem. There are no moving parts and just a few small-gauge wires to connect to your remote wind-direction read-out.

    As for the wind direction read-out, this doesn't have to be an old-school analog dial indicator, unless that just happens to be your preference. Maybe you could explore digital alternatives, such as discrete LEDs arranged around a compass rose and driven by a small microprocessor or microcontroller. Dual-colored or multi-colored (RGB) LEDs could be programmed to show wind gusts from different directions, or maybe wind speed. The μP or μC would perform the anemometer signal conditioning function and provide the directional and wind speed output "signal" computations. Let your imagination roam free on this for a moment.

    If you really need and want an analog needle display, you can directly drive a pointer with a tiny stepper motor, controlled by the μP or μC, and equipped with an external index output (either a Hall sensor or an optical encoder works well) to indicate a specific "home" position, from which you count steps to point the needle in a particular direction. The initial finding of the "home" position need only be done once after initial power up. The stepper motor will mechanically "remember" where it is and change positions accordingly when a change in wind direction causes the μP or μC to again issue stepper commands.

    Is it your intention to learn electronics and pursue it as a hobby, beginning with this somewhat ambitious project? If so, we can help you. I am recommending some pretty advanced hardware in the form of microprocessors or microcontrollers, but very little real electronics knowledge is required to hook them up and learn how to program them. Hot-wire anemometers are analog rather than digital devices, but their analog signals are easily digitized and subsequently processed as digital data for use by a program you write.

    Please let us know how you intend to proceed. Sounds like a fun project, even though less expensive commercial off-the-shelf (COTS) products are available.
     
  6. Alec_t

    Alec_t

    2,487
    642
    Jul 7, 2015
    Then they will all be cooled the same amount. Ok if you are interested only in wind speed, but how would this arrangement enable wind direction to be determined? I think you'd have to shield them so that they are exposed differently.
     
  7. kellys_eye

    kellys_eye

    4,289
    1,143
    Jun 25, 2010
    Hot wire is used for wind speed not direction.
     
  8. hevans1944

    hevans1944 Hop - AC8NS

    3,908
    1,867
    Jun 21, 2012
    The rate of heat loss from each hot wire is dependent on how fast the wind flows past each wire, so @Alec_t is correct: if all the wires are in the same wind then they should all lose heat at the same rate, regardless of wind direction. My bad for not thinking this through.

    Perhaps each of four hot wires could be fitted with a small cylindrical shield having a vertical slot that would expose each wire differently, according to different wind directions impinging on each slot. Unfortunately, such a sensor (assuming it worked) would probably be highly non-linear and its sensitivity to wind direction would vary with wind velocity. There must be some very good reasons why the rotating-cup anemometer and rotating wind-direction vane are still used in weather stations all over the world today, despite their analog "limitations" and mechanical complexity.

    I have a spool of very thin nichrome wire laying around somewhere in my stash of "junque parts" that I could use to construct hot wire anemometers, so maybe I will look into how to make four of them, appropriately shielded, sense wind direction. Actually, with modern electronics to handle the excitation and signal conditioning, just about any wire whose resistance varies as a function of temperature would work. Maybe even carbon or metal-film low-wattage resistors could be used. I will look into that, too, as soldering up resistors is easier than making connections to thin wires.

    Thanks @Alec_t for setting me straight on this.

    Hop
     
  9. Alec_t

    Alec_t

    2,487
    642
    Jul 7, 2015
    Perhaps a bit of non-linearity of directional response when using an array of sensors would be no big deal for the TS's purpose. A bit of signal processing could probably correct it to some extent if required.
    I was thinking a bunch of thermistors in respective 'tunnels', differently oriented, might do the job as sensors, energised at sufficient constant current for self-warming to be significant.
     
  10. hevans1944

    hevans1944 Hop - AC8NS

    3,908
    1,867
    Jun 21, 2012
    Hmm. You might be able to "get by" with just four "tunnels" with two pairs, each pair at right angles to the other pair. You might have to restrict the air flow out one end to differentiate which direction the wind is blowing from. I will continue to "research" the Internet for already existing implementations before attempting to "roll my own" however. This ultrasonic version looks promising if someone can get the cost down to hobbiest levers.
     
  11. Alec_t

    Alec_t

    2,487
    642
    Jul 7, 2015
    Agreed. Otherwise there could be a 180 degree ambiguity. I wonder if a tunnel closed at one end but with a non-axial vent might do the job, like this :
    Tunnel.PNG
    Any aerodynamiciists out there to confirm if air-flow inside the tunnel would be different if air externally, of a given speed, were in direction B to A rather than A to B?
     
    Last edited: Sep 29, 2018
  12. hevans1944

    hevans1944 Hop - AC8NS

    3,908
    1,867
    Jun 21, 2012
    I found another sensor that looks promising and potentially inexpensive. This one is based on sensing the thermal gradient across a centrally heated surface to calculate both wind speed and wind direction:

    [​IMG]

    Accuracy and resolution appears mediocre compared to rotating-cup anemometer and wind vane, but it has no moving parts and, hence, no maintenance... except maybe periodic cleaning of spider webs and insect detritus.
     
Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.