Sensors
As discussed in the Working Requirements we will need to measure 8 parameters. To match power self-sufficiency and cost requirements we will divide the sensors in two categories: low consumption and standard consumption:
Standard consumption: (we can accept a higher power consumption for these sensors since mounted on the master)
- rainfall
- solar radiation
Low consumption:
- ambient light
- wind intensity
- wind direction
- temperature
- humidity
- acoustic sound level
For each parameter we will here discuss what sensors have been considered, attempts, tests and results. It would be nice to include: paper citations (DOI or other if not doi not available), exact hardware, materials source and availability.
- tipping bucket:
- PROS: low power, available off the shelf (might require reverse-engineering?)
- CONS: mechanical (wear and tear), harder to DIY and calibrate
- attemps: ...
- piezoelectric disdrometer:
- PROS: high resolution, no mechanical parts, cheap, easy to DIY
- CONS: sensor rusting (for unenclosed design), not available off the shelf
- resources: M.T. Charact. and Design of Circuit for Piezoelectric Acoustic Disdrometer, Ravi Bagree, Delft
- Hardware: (?)
- attemps: Piezo Disdrometer, @Anze-, in progress
Measuring solar radiation is necessary to provide accurate estimates for solar power generation, independent of the solar cell of choice
- Solar panel
- pro: provides power
- con: does not provide a cell-independent sunlight radiation measurement, estimates will vary based on chosen solar cell
- Pyranomter
- pro: provides a cell-independent power measurement, industry standard
- con: can be expensive depending on which technology it's based on
- sensing approaches: thermopile, peltier cell, multiple photodiodes combination, single photodiode
Literature references:
Burgt, A.P. van der (2020) Designing a Low-Cost Autonomous Pyranometer.
based on available literature (papers and thesis research) an appropriate multiple photodiode combination can give better results when compared to Peltier cell systems, maintaining comparable hardware and simpler data processing.
The cited Thesis [Burgt] reports best performance is achieved ( RMSE of 16Watts ) when using a combination of the following sensors:
-
SI1145:
- digital sensor (I2C)
- High dynamic range 1 LUX to 128000 lux
- High resolution down to .1 lux
- Two Bandwidths: IR + VIS Dev. Info
-
GYML8511
- analog sensor
- 280-390nm [UV complementary Band]
the model and fitted parameters for irradiation calculation are provided in the thesis paper.
First test run on the 25/04/2021, sensors provide good data, but combining data for an improved output is challenging:
using "mymodel1" model lowpass(a*viss+(viss-irs)*uvs/b, c):
attempt by @jerzeek
We want to measure ambient light for many reasons, like knowing how much shade is provided by some trees, what are sunrise/sunset times given the conformation of the land and adjusting power readings for eventual solar panels placed at such location. Light intensity is measured in Lux, ranging from 100.000 lux (under direct sunlight), to 20.000 lux (shade on a bright day), to 400 lux (at sunset), down to 0.2 lux for a full moon at night
Options
-
LDR (light dependent resistor)
- pro: cheap
- con: resistance does not tell much, no linear response, requires analog pin
- price: €0.37
-
BH1750
- pro: i2c communication
- con: only up to 65000 lux
- price: €3.30
-
TSL2591
- pro: up to 88.000 lux [still not in range]
- con: more expensive
- price: €8.35
Conclusion: bh1750 works good
- 2 Axis Ultrasonic Anemometer
- PROS: measures both intensity and direction, inexpensive, DIY, high precision, high reliability, low power
- CONS: challenging electronics
- resources: Soldernerd.com,Digital Measurement (getting 81kS/s ++) , Analog phase shift measurement (flow)
- Hardware: (?)
- attempt by @Anze-, not yet started
attempt by @jerzeek, in progress We want to measure temperature over time to see hotspots and cold places on the camp.
-
DS18B20 waterproof
- pro: accurate, waterproof housing
- con: one-wire interface(?)
- price: €3.55
-
DHT11 temperature and humidity sensor
- pro: cheap, also humidity
- con: not accurate
- price: €2.80
-
BME280 temperature, humidity pressure sensor
- pro: also humidity and pressure, I2C communication
- con: sensor on a pcb
- price: €4.50
-
Pi Pico CPU temperature sensor
- pro: built into microcontroller, but is often a few degrees warmer than ambient
conclusion: except for the DHT11 All sensors seem to work good, the temperature and humidity sensor should be placed in a place where ambient air can easily get, but where moisture can not harm the electronics, this is easily done with the DS18B20, but more difficult with the BME280
attempt by @jerzeek, in progress as mentioned above, BME280 works well but should be outside the box.
- Microphone:
- PROS: cannot provide a full audio noise spectrum (know if loud sounds are low or high pitched)
- CONS: needs to run the whole time
- Hardware:MAX9814+ADS1015
- Microphone+Programmable Comparator:
- PROS: the programmable comparator circuit lets the microcontroller sleep -> low power
- CONS: more expensive, ADS1015 only goes up to 3 kHz, cannot provide a full audio noise spectrum( a mixed approach could get both)
- Hardware:MAX9814+ADS1015
- attempt by @Anze-, not yet started
system monitoring
- INA219 current and voltage sensor
- makes it possible to see how the battery and solar are doing
- price: €3.00
- SD card storage
- SPI connected SD card storage to save measurement data
- price: €1.95
- Real Time Clock
- DS1307, did not work on battery power since it had no 5v
Soil sensor
- used a capacitive soil sensor, and seems to work well, one with 2 rods instead of the pcb type
For the MVP I used a 6v1a solar cell from waveshare, with a 1n4001 diode inline to the TP4056 charge IC, this is connected to a 2200mah battery, the TP4056 makes sure it is charged correctly, and protects the battery.