I didn’t expect to pick an energy-harvesting device for the best power announcement of 2011. The whole energy-harvesting scene, which once held a lot of promise for new and useful applications, seemed to have gone dormant.
But then Texas Instruments announced the bq25504 (Fig. 1). It’s the missing part of the energy-harvesting jigsaw puzzle, and I didn’t even realize it was missing.
The bq25504 is a boost charger IC for nano-power energy harvesting and management. It makes sense that TI would develop it, because the company already has a great deal of expertise in battery charging, gas gauging, and security.
More than that, TI more or less dominates turnkey solutions for the RF data-out function in energy-harvesting applications. When I see a demo of an energy-harvesting product at a tradeshow, I can almost guarantee that the RF element is the TI CC2530ZNP Mini kit (Fig. 2).
A Closer Look
To get really excited about the bq25504, you have to have looked at a lot of datasheets for energy-harvesting ICs and modules. If you thumb back past to the applications section, you’ll notice that they all show the output of the device simply connected across a battery or a capacitor. There is no kind of monitoring, energy management, or charge/discharge control.
The implication is that the source of the energy being harvested is constant, and all the output of the harvester needs to do is keep the storage device topped up. Another implication might be that it’s up to the system designer to come up with the circuitry to keep the energy-storage device happy.
If you look at the typical energy-harvesting application, however, managing load demand is not so simple. Take the classic example of a system intended to monitor a highway or railroad bridge by monitoring the stresses at various points as traffic crosses and re-crosses. The energy to run the data collection from each strain gauge is harvested from piezoelectric transducers co-located with those strain gauges.
Within the system, the data is collected over relatively long time spans and periodically burst to a receiver at a collection point. Thus, the load while the data is being collected and stored is relatively low, with periodic peaks when a transmit event occurs. Meanwhile, the piezoelectrically generated energy comes in packets as traffic crosses the bridge.
A system designer might overdesign each node to store much more energy than needed during periods of heavy use to have sufficient energy to keep reports coming in during low periods, but a more conservative designer might want to code the system to skip reporting cycles during slow parts of the day. Those are the designers who need a way to really manage the energy inflow and outflow from the battery, and TI has aimed the bq25504 at them.
Specifically, TI says this is the first of a new family of intelligent integrated energy-harvesting Nano-Power management devices for ultra-low-power applications. That means the bq25504 isn’t destined for piezo transducers. It’s designed to efficiently acquire and manage the microwatts to milliwatts of power generated from of dc sources like photovoltaic or thermal electric generators (Fig. 3).
Its efficient dc-dc boost converter/charger, which requires only microwatts of power to begin operating, is customized for wireless sensor networks like the CC2500 family that have stringent power and operational demands.
Specifically, as long as the bq25504 can see something over 330 mV to start with, it will continue to operate from any input source with an output greater than 80 mV. (The 330 V need not come from the energy source. It can be supplied by a starting battery. When operating, input voltage regulation prevents the collapse of the input source.) In the steady state, the device’s quiescent current demand is less than 330 nA.
The device has built-in dynamic maximum power point tracking (MPPT) for optimal energy extraction. MPPT is achieved by periodically sampling a ratio of the open-circuit voltage of the energy harvester and using that as the reference voltage to the boost converter. The sampling ratio is set by external resistors. For photovoltaic harvesters, the ratio would normally be set between 0.7 and 0.8. For thermoelectric harvesters, the ratio is usually around 0.5.
Internally, the boost converter modulates the effective impedance of the energy transfer circuitry to regulate the input voltage to the sampled reference voltage. A new reference voltage is then obtained every 16 seconds by periodically disabling the charger for 256 ms and sampling a ratio of the open-circuit voltage.
Alternatively, the internal MPPT circuitry and the periodic sampling of the input voltage can be disabled, in which case an external reference voltage can be fed to the sampling pin. The boost converter would then be regulated to the external reference.
Batteries or Supercaps
Thanks to TI’s background in battery management, it’s easy to design an energy storage approach that uses rechargeable lithium-ion batteries, thin-film batteries, supercapacitors, or conventional capacitors. Whatever approach the system designer chooses, there are programmable undervoltage and overvoltage levels, plus an on-chip temperature sensor with programmable over-temperature shutoff. To monitor the battery condition and warn attached microcontrollers of pending loss of power, there is a battery good output pin with programmable threshold and hysteresis.
Complementary RF Speeds Development
The ubiquity of the CC2530ZNP ZigBee reference design in demos at energy-harvesting conferences highlights the synergy of the products in the TI portfolio. The development kit comes with three target boards that allow users to set up a small ZigBee network. Two boards can be mobile using AAA batteries. From there, it’s a simple matter to adapt the power scheme to use an energy-harvesting module such as the bq25504.
The kit sensor boards include an accelerometer, temperature sensor, and light sensor that can be used in conjunction with LED lights and pushbuttons to develop simple demo applications. It’s a quick learning experience. The target board connected to the USB stick is pre-programmed with a coordinator sample application. The coordinator sets up the network and configures the ZigBee network parameters.
The sensors, which could be powered by the bq25504, periodically report their key data to the coordinator. The devices can also be programmed as routers, which would be used in a system like a bridge-monitoring application to extend the ZigBee network beyond its normal range.