Electronics Q: Wind direction sensing - how does it work

[skyflyer]

The actual angle is not as important as the fact that the value remains constant and any variation left or right can be detected.

So far a mate of mine who is a bit of an Arduino whizz, has managed to come up with a program to read the three voltages off the three wires, (fed via ballasting resistors to keep peak voltage below 5v). The sequence of wiree with the highest voltage determines which 60 degree sector is being looked ta (0-60deg, 61-120deg, 121-180 deg etc) and then the ration of the highest to second highest voltage determines how far into that sector you are - i.e. the angle.
The Arduino will then happily transmit the ASCII sentence for MWV and an available routine calculates the CRC checksum. The arduino can be set to give 4800,baud parity, stop and start bits as required.
Its an easy subroutine to take an average of say 100 or 500 readings for each angle and sentence transmission intervals can be set to whatever you want.
Windspeed isn't needed for what i want but could easily be added being essentially a 5v pulse every time the anemometer makes one revolution.
Arduino board available for under £5. a couple of resistors, a a box etc - should have it all done for under £10!
(but then need to convert NMEA0183 to N2K so another £100+!)

 

[Lon nan Gruagach]

Any links? All the breakout boards I can find some to be in double figures so it makes more sense just to use an arduino.

No links, just in my head.
Buy 1 op amp comparator and a diode for each input, use diode mixing into 1 op amp input to select the channel and PWM on the other op amp input to measure the value.

 

[nigelmercier]

.. a mate of mine who is a bit of an Arduino whizz, has managed to come up with a program to read the three voltages off the three wires, (fed via ballasting resistors to keep peak voltage below 5v)...

Surely he is feeding them with the stabilised 5V feed, rather than a nominal 12V?

You lose the MSB doing it this way, but it is far more accurate in the long run. I've just done the calculations to find the optimum feed resistor for a 30Ω to 240Ω sender unit. Result is 91Ω, but 100Ω is close enough.

 

[GHA]

No links, just in my head.
Buy 1 op amp comparator and a diode for each input, use diode mixing into 1 op amp input to select the channel and PWM on the other op amp input to measure the value.

'fraid you lost me there... maybe you could expand if you've time on your hands, always good to learn.
Though it seems a long way round when you can get an arduino mega less than a tenner with 16 x analogue adc inputs, add a few more quid for a sd/rtc card and you can save the time stamped data locally of send if to a pi serially, probably why arduino./pi is such a common partnership - it's really cheap, powerful and easy

Sorry for the thread drift - one thing, the internal voltage reference on arduinos is known to sometimes drift a little if the external power isn't solid, using an LM4040 as an external reference you can get better accuracy....
http://www.skillbank.co.uk/arduino/measure.htm

 

[Hoolie]

We have Stowe wind instruments and the windspeed only works when powered up (what a surprise!). But the wind direction instrument shows even when it's unpowered. How does that work?

 

[Lon nan Gruagach]

'fraid you lost me there... maybe you could expand if you've time on your hands, always good to learn.
Though it seems a long way round when you can get an arduino mega less than a tenner with 16 x analogue adc inputs, add a few more quid for a sd/rtc card and you can save the time stamped data locally of send if to a pi serially, probably why arduino./pi is such a common partnership - it's really cheap, powerful and easy

Sorry for the thread drift - one thing, the internal voltage reference on arduinos is known to sometimes drift a little if the external power isn't solid, using an LM4040 as an external reference you can get better accuracy....
http://www.skillbank.co.uk/arduino/measure.htm

Always got time for systems.

1 op amp comparator and a diode for each input, use diode mixing into 1 op amp input to select the channel and PWM on the other op amp input to measure the value.

Sorry, no diagrams....
Lets start with diode mixing.
Simple case:
You have some low impedance voltage sources, connect these each with a resistor and a diode to a common mix point. The mix point will be the highest source minus diode drop. Now, if you ground the junction between resistor and diode for any input you can take that input out of the running. So it is possible with just diodes and resistors plus the grounding circuit to select any of the voltage sources.

Op amp ADC
Apply the input voltage to one op amp input and a generated reference voltage to the other. Ramp the reference voltage and note its value when the op amp switches.

PWM generate voltage.
Use variable mark/space ratio on a digital output, smooth it to produce a voltage.

 

[Lon nan Gruagach]

We have Stowe wind instruments and the windspeed only works when powered up (what a surprise!). But the wind direction instrument shows even when it's unpowered. How does that work?

Only way I can think of is that it uses 2 stepper motors connected pin for pin, turn one by hand and the other will follow exactly with no power applied.

 

[ghostlymoron]

'Currently I would never have wind speed and direction instruments on my boat due to unreliability'. I agree Will.
If you can't tell where the winds coming from, you should get a motorboat.