Human Comfort Sensor
This project is an attempt in creating a low cost weather station sensors in order to measure the level of human comfort in that area. The sensors are compossed of wind speed sensors, black globe temperature sensors, temperature & humidity sensors
Wiring diagram and program code : Human_Comfort_Supplementary
Contents
Windspeed Sensor progress
DC motor prototype
We tried to build a windspeed sensor from scratch, the first sensor are make using DC motor as a base. The idea is we will have the wind to turn the propeller that are attached to the dc motor and have the motor act as a generator. Turns out this method is not viable to be use for windspeed sensor application as the motor have a stall torque that are preventing it to turn when the windspeed is below 1 m/s.
Hall effect protoype
For the second prototype we tried to use a more popular method to measure windspeed which is to use a hall effect sensor and magnets. The sensor works by having the hall effect sensors to trigger when there's magnet on top of them and the hall effect sensor measure count the amount of time the magnet triggers the sensors.
Early prototype shows promise as this method could handle with low wind speed measurement as the only thing that could cause kept it from turning is the smoothness of the bearing. Sadly we don't go forward with this method as our users prefer to go with premade one as they have concern with the build quality of the 3D printed material. Although there are other 3D printer filament that could go against sunlight degradation, we need a long term testing to be sure as such we are not going to continued with this setup.
Assembled windspeed prototype
Component inside the sensors
QS-FS Sensor
Since the 3d printed option is not viable for this application, we tried to use the one that are available in the market and found this windmeter sensor. The version that we use is the one that gave the output of 0-2 Volt but the input voltage is at 12V. After dissassembling the 12V apparently just ensure the voltage stability and if we remove that module we could make the sensor run in a lower voltage and even run in 3.3v albeit changing a bit in how the original analog readout calculation.
We setup this wind sensors running in the rooftop and we compared it to a reference sensors that are being used previously by Melbourne Uni and results are as follows.
Overall it looks that we could get trend of the wind speed like the ref sensors but there are some points in time where it got an error spike in its reading, if we check the correlation without filtering this high spike errors, If we remove this errors we could improve the correlation value to 84%
| Windspeed Correlation | Windspeed Correlation w/o error spikes |
|---|---|
Overall sensors value is still in acceptable ranges, might need to do some trial and error on finding the right value conversion from the received analog signals
Black globe temperature sensors
We tried to make this buy getting a copper ball coated in black paint that are used for floats and put in a heat sensor inside. Later we'll compare the data that are gotten from the readout to an actual black globe thermometer to ensure the quality of the data.
This sensor are also being setup outside of the building and we test them by comparing it to a reference sensors.
| Black Globe Results | Correlation Test Result |
|---|---|
Results from this test shows a quite promising results with the sensors capable of following the trend of the ref sensors data and have a 93% correlation to the ref sensors.
Temperature & humidity Sensors
In the market there's a lot of chips that could detect temperature and humidity at the same time. There are 3 candidates that seems to be the best in terms of quality and price ratio which is the BME280. This is based on testing that are done on this website. The main advantages of this sensor is that if we need to do reading adjustment for location that have a different atmospheric pressure it could compensate for that. Although for current application we don't use this method and just have the sensor to gave the reading directly to us.
Weather shield testing
Testing the weather shield that we bought and the previous one that we have gotten from melbourne uni to see if theres any major effect on the different material it have. We put all 3 types of temperature & humidity sensors and have it compared to a standard instrument that are used for actual weather station monitoring.
