The instructions provide approximate values for a voltage divider based on a 5V Arduino. These days there are probably more 3.3V devices out there and the values are different. I found it was better to simply work off of the analog values after sampling at every possible direction (N, NNW, NW, WNW, W........). Takes a bit of time, but then a simple table allows a more reliable reading. I've included mine (using a 10k resistor. The value read must be LESS THAN the table value to represent the direction. Hope this is useful to someone. struct wind_Dir_Def { char dir_ASC[4]; uint16_t dir_Val;
uint16_t dir_Cnt;
}; // Uses a 10K resistor for VD wind_Dir_Def windDirTable3V[16] = { {"ESE",84,0}, {"ENE",92,0}, {"E",126,0}, {"SSE",185,0}, {"SE",245,0}, {"SSW",287,0}, {"S",408,0}, {"NNE",463,0}, {"NE",599,0}, {"WSW",629,0}, {"SW",702,0}, {"NNW",784,0}, {"N",824,0}, {"WNW",883,0}, {"NW",939,0}, {"W",1024,0}}; // Check all the possible values to determine the closest to what we have for(int ix = 0; ix < numDirs; ix++) { if(rawWindDir < windDirTable3V[ix].dir_Val) { // We tally up the number of hits for a particular direction and then use the one with the most for reporting windDirTable3V[ix].dir_Cnt++; prevWindIdx = ix; break; } }

We live on a windy coastal headland, and I was pleased to see good data flowing as soon as I plugged the unit into the WeatherShield. The anemometer and wind direction devices look robust -- they will be tested! I replaced half the metal mast, as it wouldn't survive the salty air. Having minute-to-minute wind speed and direction is a treat! No rain here yet, and some spurious hundredths associated with wind gusts are registering, but the wiring for the rain gauge is simple, and I can sub in a different unit if I can't get this one stabilized. I am running this unit in parallel with a reliable older unit (that gives instantaneous data but doesn't play nicely with my RPi system), and the new data is acceptably close. Good unit; easy assembly and installation.

The weather station is connected to the ublox cellular Expresslink and an RP2040 for awesome visualization in aws grafana.

I have 3 stations based on these sensors. I just replaced the oldest sensors at 6 years. I bumped a cup and it snapped off. The plastic of the cup arm had mostly deteriorated and was very brittle. At around $14 per year for sensors - Good Value

The vane is dependent on the length of wire between the sensor and the micro. Too long - won't work. Too short - only 4 directions solid.

The light cups are very sensitive to light wind and velocity changes. I have recorded 86 mph wind without issue. The cup arms are the limiting factor for age. A little more material would push service life a lot.

The rain bucket works well. It can give a false toggle from high gusty wind.

First, this is missing temperature, humidity, and barometric pressure sensors. If it had these sensors it would be significantly more complete. Next, there is a lack of interface units to connect it cleanly to your processing unit. I want to use it with a Raspberry Pi Zero, 3, or 4, but there does not seem to be a HAT for it yet/anymore. Once I built the required adapters to hook stuff up, I found that the wind vane switches do not seem evenly spaced and some of the 16 directions it can detect are larger than others, no matter how slowly I turned I could only barely get some directions to show up. Overall I think it's a good start, but be prepared spend a fair amount of effort to get it working.

Can anyone tell me the size and length of the screws needed to assemble this kit? I don't have any that small, so I'll need to run down to Ace Hdwr or Lowe's for some small machine screws. I'm not looking at it now, but I got couple pieces out of the box when it arrived to see how they looked, as I recall they looked like #4 or possibly #6 machine screws. Can anyone advise? Thanks in advance! BTW, I'll be using an ESP32 Modbus board (whatever it's called) as the processor on mine, and I'm also using the weather breakout board that has temp/humidity/pressure sensor, a lightning detector, and UV detector for a full weather station.. I'll use either wifi or BT to get the data into the house & onto my LAN. Not sure what platform I'll use for visualization of the data, and historical aspects. So far, it all looks good! My only concern is regarding how long it will last. Also, it would be really great if Sparkfun would develop a wind sensor using ultrasonic tech. I was looking for one but couldn't find one at a reasonable price. I'm sure Sparkfun could do it! Hint, hint! Hear the call, Sparkfun?!

I am still waiting for an ADC chip to be used with the wind direction element.

The weather meter kit was super easy to assemble - thank you for detailed instructions, and very well thought out assembly technique (unlike IKEA stuff). The connections are dead-easy to use. Used the sample sketch to check everything out; next step will be to prop it outside (and all the other fun stuff that will entail).

Yes its really good, but I think to make it log and a few other AWS things you will need a Arduino Mega as the code and libs max out the Uno.

We receive the order in perfect condition and in a fairly good time.