We did a quick range test of our new CC1120 radio based sensor nodes. We did this within Pruksa (community in east Bangalore). This was a non line of sight test in which one WSN1120L node was configured to send packets to a coordinator node (a WSN1120CL) once every second. The nodes were around 526 meters apart with lots of houses in between. The nodes were at a height of around 5 feet above ground. The walls in these houses are fully concrete (no bricks). Even at this distance the signal strength was pretty good (around -75 dBm). The nodes were configured to transmit at +14 dBm at a data rate of 1.2 kbps (GFSK modulation) with channel bandwidth configured to 25 kHz.
We tried to do a line of sight test along a nearby highway but we could not find a flat stretch longer than 1.1 kilometers close by. The two nodes were able to communicate up to a distance of 1.1 kilometers. The signal strength was again very good (around -80 dBm). We will be doing further tests to find out the max range under line of sight and non line of sight conditions.
Here is a pic of one of the nodes used in the range test. The node was powered by two AA batteries.
For more information, please visit wisense.in.
We will soon be introducing our first consumer IOT product – a wireless soil moisture sensor for your plants. The soil moisture sensor is based on the open source “Chirp” plant watering alarm. We added an NTC thermistor to the same PCB. The sensor PCB is interfaced to the WSN1101L sub-GHz wireless node. The soil moisture sensor is a capacitive sensor. When water is added to the soil (in the vicinity of the sensor probe), the capacitance of the probe increases. When the soil looses moisture, the probe’s capacitance decreases. The MSP430 on the WSN1101L is used to generate a square wave (around 330 kHz) which is fed to a low pass filter created by a 10 kilo-ohms resistor and the capacitance of the soil moisture probe. The output of this low pass filter is rectified using a diode and the average level of the filtered square wave is measured using one of the 10 bit ADC channels on the MSP430. As the capacitance of the probe increases (when water is added), the output voltage decreases. As the soil dries out, the probe capacitance decreases and the output voltage increases. Note that capacitance increases when water is added because water serves as a dielectric. Higher the dielectric constant, higher the capacitance.
The sensor PCB has a coin cell holder which can hold a 20 mm coin cell (CR-2032 for instance). This sensor node can run on the CR-2032. If the sensor measures and reports the sensor readings say once every 30 minutes (or more), the coin cell can last more than a year.
This product will be an open system. We intend to allow users to add more sensors to the product. We are going to make the source code available so that anyone can add sensors or modify any layer of the stack. The MSP430 on the WSN1101L can be programmed through the 2 pin spy-bi-wire interface using a low cost launchpad (MSP-EXP430F5529LP) from TI. Note that the MSP430 variant on the WSN1101L is the MSP430G2955. All the MSP430 pins will be exposed. Add new sensors, replace existing sensors, modify/add any part of the firmware or just completely re-purpose the system.
Here are some pics.
The WSN1101L has a U.FL connector.
We are prototyping a WiSense network for monitoring the level of standing water in paddy fields under the guidance of Dr. K. Palanisami (Agricultural Economist). This a first step towards Precision Farming. Paddy requires around 50 mm (or less in some cases) of standing water. We are trying to evaluate the feasibility of using a wireless network of water level sensors to make sure that paddy fields always have adequate amount of standing water. A cluster of sensors can share a single GSM gateway to keep costs down.
Pic courtesy: http://travel.paintedstork.com/blog/2009/07/monsoon-musings.html
We built a prototype which can measure water level up to 300 mm and send the measurements to farmers over SMS. In the first stage, farmers will be informed of the water level so that they can decide when to irrigate the paddy field. In the second stage, water pumps will be activated automatically to irrigate the fields.
The prototype is powered by a couple of AAA batteries. Since it will be deployed outdoors, these nodes can also be solar powered.
Here are some pics of the sensor node interfaced to a WiSense WSN1101L.
We built another version of a gateway box for a customer with multiple interfaces (wired ethernet, WiFi, GSM and of course a WiSense radio). Here are some pics. All these interfaces plug into a Raspberry PI. You can run the WiSense CLI and the WiSense network management system (WiSight) on the Raspberry PI. WiSight can be accessed through a web browser.
The gateway box includes a power supply with 12 V and 5 V output to power all the interfaces and the Raspberry PI.
Here are some pics.
The INA219 (from TI) is a high side bi-directional current sensor with I2C output. It measures the voltage across a shunt resistance. It can also measure the supply voltage.
We are using the INA219 to monitor the battery pack in a solar powered light. During the day , the solar panel charges the battery. At night, the battery powers the light (string of LEDs). This requires bi-directional current measurement.
The INA219 can measure both the current (into the battery or out of the battery) as well as the battery voltage. The INA219 allows battery current to be measured using high side sensing. High side sensing and low side sensing are two different concepts.
We have a customer who wants to monitor the battery on a group of street lights. These lights will be part of a WiSense mesh network. The voltage/current data from each light will be forwarded to a gateway node comprising a Raspberry Pi and a GSM modem (there is no internet in the area where the lights are going to be deployed). Software on the Pi will upload the data to the cloud.
In the figure above, the WiSense node and the INA219 are shown to be powered from a different power source (couple of AAA batteries for instance). The Battery pack ground and the INA219 ground need to be connected only of the INA219 is used to measure the battery voltage. It is not required to tie the two grounds together if only shunt voltage (across the shunt resistance) needs to be measured.
Here is a pic of the PCB with a 0.1 ohm shunt resistor mounted.
We tested the PCB at very low currents (around 20 mA). The INA219 is reporting the expected shunt voltage value. We will be testing the PCB with one of the solar street lights when we get the equipment. The battery current is expected to be around 300 mA to 600 mA.
We built a couple of gateway boxes for a customer. Each gateway has a WiSense node and a Raspberry Pi. The box can connect to a router through ethernet. The box is powered through a USB mini-B connector. The USB interface also provides serial connectivity to the Raspberry Pi. This gateway has a CC2D33S temperature and humidity sensor.
Here are some pics.
The low profile of our new offering (WSN1101L) allows it to be easily integrated with the Chirp soil moisture sensor. I glued the WSN1101L to the top of the moisture sensor and connected the two with 5 wires (Vcc, ground, I2C-SDA, I2C-SCL and RESET). The moisture sensor “stick” has a coin cell retainer which can hold a 20 mm CR-2032 coin cell. This coin cell powers both the moisture sensor and the WSN1101L. The WSN1101L has a flexible whip antenna connected to a U.FL connector on the radio board.
Here are some pics.