With the rising trend of the Internet of Things, commonly known as IoT, remote sensors that are capable of tracking the environment and relaying the data back to an Internet connect host system are becoming increasingly common.
However, designing such types of sensors to be powered from an energy-harvesting power source (such as a thermo- or piezo-electric generator, solar panel, and so on) involves more planning and design time when compared to designing a system with a typical power source (for instance, a wall adapter or battery).
In this article, a step-by-step process is provided on how to optimally create an energy harvester-powered remote sensor including a transceiver, with a Li-ion rechargeable battery and, so that the battery and energy harvester are correctly sized to fulfill the end application load without overdesigning or resulting in additional cost. A cost-efficient system that has optimally sized components can be developed by beginning at the system load and working back to the input source.
Step 1: Harvester Selection
Selecting a suitable type, but not size, of energy harvester(s) for the application setting is the first step to increase the amount of energy that is actually harvested during non-dark times. Here, the example system used for illustration purposes is a minimally sized, low-power outdoor sensor driven by a solar panel. If the suitable excitation source, such as mechanical vibration or thermal gradients, is available, other types of harvesters can also be employed.
Step 2: Load Minimizing
Consider the outdoor remote, solar-powered temperature sensor as depicted in the simplified block diagram of Figure 1. When it comes to designing an energy-harvesting system, the most critical step is to reduce the load profile of the system by (1) selecting ICs that consume the least amount of power, and (2) operating such ICs in a low duty cycle burst mode.
Figure 1. Remote sensor application.
In order to simplify the calculations, it can be assumed that the sensor itself, RF transceiver, and I/O peripherals are directly powered from a Li-Ion battery with 3.6 V average “HI” rail, whereas the processor core is powered from an individual 1.8-V “LO” rail.
The system includes two periodic active modes and a sleep mode to reduce the overall power consumption. In measure mode, temperature measurements are taken for 0.5 seconds, and immediately after the measure mode, transmit and receive mode (TXRX mode) transmits and receives data to/from the host for 100 ms every 60 seconds. As the processor enters a low-power sleep-mode, the measurement and transceiver devices are switched off when they are not in use. Figure 2 shows the load current profile for the system.
Figure 2. Load current profiles.
Step 3: Power Management IC Selection
From the load profile, the DC/DC converter and/or battery charger with specifications capable of working within these current and voltage levels and also some other vital features were selected. Preferably, the power ICs quiescent current is much below the current consumption of the system in sleep mode.
Compared to a standard lower impedance wall adapter or battery, an energy-harvesting source has a much higher impedance output and thus acts more similar to a current source rather than a voltage source. This causes its output voltage to collapse at considerably lower output currents when compared to the voltage of a low-impedance (Low-Z) source. Hence, the DC/DC converter that instantly follows the harvester should manage (limit) its own input current draw so that the output voltage of the harvester no longer collapses.
A compact system, including solar panel area, will output 0.6–0.7 V each, in parallel or series, and will have as few solar panel cells as possible. More than seven to eight cells in series are needed to fully charge a 4.2-V Li-ion battery from a buck-based battery charger. A boost-based charger enables more flexibility in solar panel configuration and, possibly, less number of solar panels. The bq25570 boost-based battery charger with built-in buck converter was selected due to its input voltage regulation and maximum power point tracking, or MPPT, features and also because of its power levels.
The MPPT feature samples the open circuit (unloaded) voltage of the solar panel and stores a user-selected fraction of that voltage on a capacitor every 16 seconds, while the input voltage regulation circuit enables the boost converter to pull current from the source until the source voltage reaches that sampled voltage.
Step 4: Compute Minimum Battery Size from Load Currents
The capacity of a battery is calculated in milliamp-hours (mAh). Sizing the battery to guarantee complete operation for around two days of extended darkness (like cloud cover), the overall current sourced by the battery (IHI1 + IHI2) is initially calculated. By using the quantified efficiency values for the buck converter from the bq25570 datasheet, the buck converter load current (ILO) is reflected back to the high rail current (IHI2) by solving the efficiency balance equation ï¨=POUT/PIN = (VOUT x IOUT) / (VIN x IIN). Since the operating time of the solar panel is calculated in days, multiplication is done by each mode’s duty cycle and then 24 hours/day to obtain mAh/day consumed. These computations are summarized in Table 1.
Table 1. Computation of mAh/Day (1Includes battery leakage current and bq25570 quiescent current).
Given its combination of boost charger and buck converter, the TI - bq25570 is considered a good fit. Moreover, ILO = 5 mA is seen to be well below the 100 mA maximum buck converter output current, and 3.6 V (IHI1 + IHI2) TXRX / 0.80 = 86.9 mW is found to be well below the 510 mW maximum input power for the boost converter.
With two dark days, 2 days x 2.10 mAh/day = 4.20 mAh is the absolute minimum battery capacity needed for the remote sensor to constantly operate. It is recommended to use a battery that has a slightly higher capacity so that it does not fully discharge toward the end of the dark time to offer some amount of safety margin.
In case a supercapacitor that is fully charged to 4.2 V had to be used, but not permitted to drain below 2.5 V, then the following equation is solved for CSUPER:
3.57 mAh/1000 x 3600 seconds/hour x 3.6 V = 1/2 x CSUPER x (4.2 V2 - 2.5 V2)
And get CSUPER = 11.2 F.
Step 5: Determine Solar Panel Size
If the 2.10 mAh/day is multiplied by the 3.6-V average battery voltage, it gives a system need of 7.56 mWh/day. If the solar panel supplies power to the bq25570 boost charger 24/7, then the system that has 80% average efficiency may run from the boost charger output without a battery — provided the solar cell gives 7.56 mAh/day/0.8 = 9.45 mWh/day.
Conversely, five out of seven operating days a week are only charged for just 4 hours/day, with the battery supplying power during the non-charging hours and during the two dark days. This implies that the solar panel should supply more power to charge the battery for a total of 9.45 mWh/day x 7 days/5 days/4 hours/ day = 2.65 mW. If a small solar panel providing 0.025 W/cm2 at minimal lux is available, then 2.65 mW/0.025 mW/cm2 = 132 cm2 of solar panel area will only be required.
SIDENOTE: Remembering secondary school’s “cancel the unit’s” trick actually helps.
Five easy steps can be used to optimize a rechargeable battery-powered system with solar re-charging:
- Selecting a harvester for the application environment
- Reducing the load by running the system at a reduced duty cycle
- Choosing an optimal power management IC
- Sizing the battery to supply power during the dark times
- Establishing the minimum solar panel size required
By beginning with the system load and reflecting the output system power backward to the harvester output through efficiency power balances, the optimal size for the solar panel and batteries can be established, thus lowering the overall cost and solution size.
This information has been sourced, reviewed and adapted from materials provided by Mouser Electronics.
For more information on this source, please visit Mouser Electronics.