Wind Speed calculator

eagletalontim

Trying to get a few bugs worked out on my solar tracker that runs from a PROP and realized I need a way to calculate wind speed so I can move the panels to a safe position to avoid damage. I already have the solar panels "parking" at night but need to trigger this same "park mode" when the wind speed gets too high. To start, I am thinking of a simple spinning cup design with a magnet at the base and a hall effect transistor to detect each rotation. The problem is, I have never worked with a hall effect sensor and have no idea which one to get. There are thousands listed on digikey and without knowing which would be best for my application, I am lost

From common sense, I am thinking this would be the one I need :
http://www.digikey.com/product-detail/en/US5881LUA-AAA-000-BU/US5881LUA-AAA-000-BU-ND/431876

It is able to handle 5V and is a switch not a latch. I am guessing this connected to the 5V rail and a 4.7K resistor on the OUT to a PROP pin would be all I need. Is this correct? Would this be a good sensor to use?

LoopyByteloose

Jameco has a pretty good selection of Hall Effect sensors to shop from. They have ones that are not just surface mount and can be used in a context similar to what you are doing.

Mostly, the things that become awkward are to get the right kind... some latch, some do not. some only recognize one end of a magnet, some toggle between passed at both ends.

Simply put, you should read up on Hall Effect in general, so you get the right ones.

Also, you may find it a bit more difficult to locate 3.3V output devices for the Propeller. Yes, 4.7K on a 5volt output will protect the Propeller. (I generally use 3.9K)

eagletalontim

I use jameco, digikey, and allelectronics quite often, but I already have a cart over $20 with digikey and would like to ensure I get everything I need in one pass (Save on shipping). I could just go with a reed switch instead?

Sapphire

Yes, a reed switch would work. Which anemometer are you using?

eagletalontim

I am going to build my own. Probably going to use some of my 2 part plastic mix and make the cups and use all thread or a metal rod for the shaft. the base can also be a disk made from the same 2 part plastic which will allow me to enclose the magnet in the mixture.

JonnyMac

If you've got an unused ADC channel you could go solid-state:
-- http://moderndevice.com/product/wind-sensor/

I've never used this; found it while searching "low cost anemometer" on Google.

eagletalontim

That is pretty cool JonnyMac! A little worried about something like that being out in the elements for years though.

JonnyMac

The point is that there are other than mechanical options.

LA6WNA

Hi. I used 3 plastic ice-spoons to make anemometer at my weather station. They had the right shape, almost the way commercial anemometers have. Worked for almost 3 years now.
To make pulses, I used a slotted opto-switch like this one: http://www.ebay.com/itm/Slot-Type-Optocoupler-Module-3-3V-5V-LM393-Comparator-Slot-Type-for-Arduino-DE-/251557847295?pt=Wissenschaftliche_Ger%C3%A4te&hash=item3a920420ff#ht_6185wt_1097
Made a 10-pin wheel to spin through the slot, so I got 10 pulses pr turnaround at the anemometer.
To test and calibrate, i mounted it on my picup truck and had it hooked up with a Prop-card and a display and then went for a drive...
Just a tip. Have fun with your projects.

W9GFO

In the interest of simplicity may I suggest that you do not need to know what the wind speed is, just that it is enough to warrant moving the panels to a safe position. When the wind picks up to that point it will cause vibrations, an accelerometer can sense those vibrations - no moving parts.

eagletalontim

I did think about something like that but was worried the sensor picking up invalid readings.... Bumping into the panels while mowing, low flying plane or helicopter, etc. With the wind speed calculated, I can log this data and use it with other projects.

W9GFO

Of course it would need some programming to filter out the odd event, helicopters and airplanes should not present a problem - unless they fly within a few yards. Even then, the program could account for temporary disturbances.

kbash

other ideas:


National Digital Forecast Database (NDFD)
Simple Object Access Protocol (SOAP)
Web Service create an automatic lookup for forcasted windspeeds in your area ( if it's in the database... I haven't looked )



Wind chimes. ( listen for the volume and frequency )


Piezo tab force flex ( add a leaf-shape, determine the frequency and bend force. )

eagletalontim

When it comes to electronics, I like to KISS.......very very simple. With a magnet on a spinning disk and a reed or hall effect sensor, it seems pretty simple to me . I can build most of that from scrap I already have. Less money for me to have to spend.

kwinn

A small stepper motor would also work as a sensor for an anemometer.

eagletalontim

How would I be able to detect the rotation? If I were to pick up the generated electricity on one of the leads, I would have to have even more circuitry to prevent high speeds from outputting too much power to a prop pin. That would mean more money, research, time and possibly another circuit board

jmg

eagletalontim wrote: »

How would I be able to detect the rotation? If I were to pick up the generated electricity on one of the leads, I would have to have even more circuitry to prevent high speeds from outputting too much power to a prop pin. That would mean more money, research, time and possibly another circuit board

Limiting energy is as simple as a series resistor.

The classic pickup for widest speed range would be a zero crossing comparator, but you do not care about very low speeds, so you can try one and see what speed crosses the prop threshold.
ie plot AC out or Peak, vs Wind speed.
In the most basic form, a divider can be used to set the speed, needing almost no code in the Prop.

I've wired LEDs directly onto Stepper coils as a simple demo, and 2 R/G LEDs in series will both clamp the voltage, and give a useful "it's working" visual check (or a single blue one ?)

A simple series R + LED Clamp + series R 'T' circuit, could gate a counter which you could use to average over as many seconds as you like, for gust filtering.
When the INC rate is above some set point, you have a sustained average wind speed.

If you wanted to used the counter edge based, then a comparator like MCP6542 will give clean edges.

jmg

eagletalontim wrote: »

When it comes to electronics, I like to KISS.......very very simple. With a magnet on a spinning disk and a reed or hall effect sensor, it seems pretty simple to me . I can build most of that from scrap I already have. Less money for me to have to spend.

A reed is simple, & needs just 2 wires, but has a finite life cycle.
The lowest sensor-cost combination is likely to be a magnet and an inductor.
At very low speeds, the inductor value can be sensed, (ferrite cores shift on the curve with external magnetic fields), and for higher speed sensing the magnetic induction effect could be used.

kwinn

eagletalontim wrote: »

How would I be able to detect the rotation? If I were to pick up the generated electricity on one of the leads, I would have to have even more circuitry to prevent high speeds from outputting too much power to a prop pin. That would mean more money, research, time and possibly another circuit board

A series resistor and 2 clamping diodes is all you need to limit the voltage in to the prop pin. You can limit the speed by putting a load resistor on the other windings or using them to charge a battery or provide power to the monitoring circuit. Not very good for a weather station where you want accurate wind speed readings but fine for detecting a threshold speed to signal it's time to park the solar panels.

eagletalontim

I may have to try this ferrite method. Sounds simple enough to wire up and it does not have any components that may give out like a reed switch can. I just ordered 10 reed switches bit it cost about $3 for all of them. Could be used as a backup if other options don't work correctly for what I need.

jmg

When we worked on Magnet-Inductance shifting some years ago, I used 330uH SMD inductors as the price knee, but I see today 1mH and above are cheap.
Possible examples
http://www.digikey.com/product-detail/en/LB2518T102K/587-2509-6-ND/2230521
http://www.digikey.com/product-detail/en/CM453232-102KL/CM453232-102KLTR-ND/1970088



IIRC ~50% changes in L were seen as the magnet saturated the ferrite.

We made a LR oscillator ( very similar to a RC oscillator - same waveforms)
Or, you could make a LC osc, for sine wave, and accept the reduced dF

A single magnet anemometer has quite low spin speeds, so a COG that measures LR osc and looks for the magnet-crossing may work better than a magnet-coil which could suffer AC mains noise suppression issues.

LoopyByteloose

Reed switching might bounce for up to 100ms (1/10th of a second), whereas a Hall Effect sensor just won't.

To acquire zero bounce with a reed switch, current of 100ma may be required. Google "reed switches contact bounce" to verify what I am asserting.

Read switches are cheap, but may not be worth the bother. I have always avoided dealing with them.

www.hamlin.com/technical-detail-glossary.cfm

cavelamb

All good suggestions, but they all give RPM, not wind speed.
That's gonna take a bit of calibrating...

LoopyByteloose

Actually, the calibration seems not to critical in this case. The OP desires to have the solar panels go into a safe position when wind speeds reach a hazardous level.

Wind begins to cause trouble when it reaches higher velocity. Have you ever noticed how automotive gas consumption goes up drastically when above 50 MPH. Air as higher velocity changes to behave as if a more dense material, such a water or mud were flowing over the object.

So I suspect locating one point is all that is required. Watching a windsock or a flag in a breeze can pretty much get a ball park reading. Behavior of tree branches is also rated on the NOHA scale.

Try searching "wind speed visual clues"

JohnR2010

My company makes exactly what you are talking about for a wattsun AZ-225 tracker. Check it out you can see a propeller installed inside the Solar Array CPU. I have several of these units installed in Alaska been working fine for years. If you are still looking for an Anemometer let me know I would be happy to help in any way I can. Here is a link to our web site if your interested. http://solar-current.com/how_the_furl-kit_works.html

eagletalontim

Sorry JohnR2010, I can't pay almost $1000 for something I can build for $50 or less Looks nice though!

The calibration is for sure not going to be exact as my first thought was a hanging flap that when it move far enough to any side, it would contact a metal ring which would trigger a park mode. Me thinking outside the box and wanting data from it, I have decided to try for the MPH of the wind. Calibration should be close enough for what I am doing. +/- 4 MPH should be more than enough accurate but I may be able to get it even closer with the right setup.

Alexander (Sandy) Hapgood

Sounds like you don't need the actual wind speed, you just need to know when it gets above a certain point. Imagine a wind vane mounted horizontally with the pivot point place such that, with no wind, the vane points straight up. As wind speed increases the vane moves toward horizontal and when it goes beyond a certain point a mercury switch sends a high signal to the prop. A certain amount of movement dampening might be required but it's worth looking into.

Sandy

eagletalontim

I truly don't need the wind speed in MPH, but it may come in handy for other projects I may work on in the future. The solar tracker will be getting an upgrade eventually which will include a WIFI connection to my spinneret which communicates with my offsite server. I can log the data sent back from the solar tracker for viewing later or possible some adjustments in the code. The Xbee Wifi USB board should be in tomorrow so I get to start tinkering with that. Once I have it all figured out, I plan on buying some circuit boards to replace the heavily modified one that should not even be in service right now :P Total project cost is not that much of a concern as is reliability and simplicity. If something does break, I can easily replace or repair that item with readily available local items.

Alexander (Sandy) Hapgood

This idea might work for you. You would just have to change the spring arrangement somehow to allow measuring the displacement. Fairly simple and ( almost ) no moving parts.

Sandy

homosapien

For something as simple as possible, I was going to suggest something similar to the link posted by Sandy, but a homebrew:

This device would only detect wind over a certain speed, it would not show actual wind speed. Simply mount a small flag on a stiff wire (such as the flags used for training dogs to avoid 'invisible fences', or those used by landscapers in plan layouts). Mount a metal washer around the wire flagstaff, so when the flagstaff is deflected by wind it touches the washer. You would need to do some testing (probably using a car-mounted test device) to determine what size washer to use and how high on the flagstaff to mount it. Anyhow, then put 3.3v to the washer and wire the flagstaff to a prop pin, pulled low by a pull-down resistor. If 3.3v is seen at the prop pin, the flagstaff is being deflected enough to touch the washer and the wind speed threshold has been seen. Of course you would need to institute de-bouncing and some sort of time-out protocol in your software.

Simple and cheap, pretty much nothing to break. You might have to replace the flag with something more heavy-duty (but I have regularly seen even the 'invisible fence' type flags last for well over a year).


Nate

jmg

That would have contact corrosion issues, but another variant on that idea, is a magnet on the end of a deflecting rod, and a reed (probably mounted on same axis as rod) to give less direction variance.
In still conditions the reed is always closed, and strong winds and gusts start to make it chatter.
Calibration is via rod stiffness/length and reed displacement.
Normally closed gives reasonable fail-safe checking against open-wire failures.