Convenient use of portable equipment implies having short maintenance periods. This means the time period when the portable device needs to be connected to service stations for doing any kind of configuration work or up and downloading data. Handling data can be implemented by using any wireless transmission techniques such as WLAN, Bluetooth or Zigbee – depending on the amount of data that needs to be exchanged so it can be implemented in a way that would not greatly disturb using the application.
The other part of maintenance is supplying the device with energy. The simplest approach is to use batteries. One method is to use primary cells that can be replaced once discharged. This approach almost guarantees 100 percent availability (24h/7d) of the application. Another method is to use rechargeable batteries, and charge them with a wall adaptor at the AC line. But while charging, the devices can not be used, at least their portable function is restricted.
To lower the power maintenance cycles, approaches to energy harvesting or any kind of alternative energy sources becomes more interesting – for charging or even powering the complete application. Examples include low power wireless sensor networks with remote sensors in areas where they cannot be easily accessed for maintenance, and AC mains are not even close. This generates new challenges to the power supply design. Usually energy harvesting techniques rely on energy that is not available the whole time, and needs to be buffered in any kind of storage device such as a rechargeable battery or a large capacitor.
Since portable devices are continually getting smaller or are integrating more functionality in the same device size, available space for batteries also is getting smaller. For this reason, smaller batteries are going to be used. To lower the average power demand, power-saving modes of operation are essential. Typically, this means that parts of the circuits are turned on only when the respective feature is needed. This needs to be done very fast, since most users do not accept a long waiting time for an application to be ready. So turning on and off happens at the highest possible speed. At the battery side, this means a high pulse current loading although the average battery may have gone down. If battery impedance increases because a smaller battery is being used, the voltage drop at the battery caused by the current pulses becomes significant. This also generates new challenges to the power supply circuits. It even may require different topologies such as buck boost conversion for Li-Ion batteries, or very low voltage boost converters for alkaline or NiMH cells.
Another challenge is dealing with energy harvesting approaches. In these kinds of applications, power is not always available and must be stored until needed. Additionally, the power source usually is very weak and cannot deliver enough power to run the application directly. A weak energy source also challenges the connected converter. Output voltages available at the energy harvesting device may vary widely and may go very low. To get as much power as possible from the energy harvesting device, the connected power supply must be able to deal with this wide voltage range in all possible modes of operation. This includes startup at low voltages to reduce the risk of maintenance required, if the buffer battery was discharged completely.
Solar cells, for example, have a terminal voltage in the range of 0.5V per cell. If the light intensity is weak, the cell impedance becomes high, and the current that can be drawn out of the cell is low. This means an attempt to start the power circuit connected to the cell is causing a significant voltage drop of the cell’s output voltage – and further lowers the available supply voltage for startup of the power circuit.
Fuel cells behave similarly. If fuel and oxygen are available at the optimum ratio across the membrane, their output impedance is lowest. If not, cell impedance increases. A typical fuel cell operates with an output voltage in the range of 0.4V to 0.6V when current is drawn out of the cell. Without load, the cell voltage can go up to 1V. The capability of dealing with fast changing loads is limited by the capability of controlling the flow of fuel and oxygen. This means that a fuel cell can not react fast enough for typical load changes in mobile equipment, even though it could be designed to generate enough power to supply the application completely. In a reliable design, an energy storing device such as a large capacitor or a battery can be used to handle peak load demands of the applications.
To get rid of the low voltage operation problem, using multiple cells in series may be a solution. Basically, it means that higher voltages are available to have more margin for the voltage drops during load changes. But, if the total available size for the energy cells stays the same, less space will be available for the cells. Connections between the cells and protection of individual cells may be required. Since the cells are getting smaller, the individual cell impedance also becomes higher. A series connection of the cells also results in higher total cell impedance, which may make the voltage drop caused by load changes to be even higher.
Looking at batteries, this means the total reliability of the battery stack is getting lower. Since there are more components (such as the cells itself and their connections) that can fail, the probability of a complete failure of the battery stack rises.
Another reason for failures is having a combination of good-charged and bad-charged batteries in series. The higher cell impedance of the bad battery in this setup causes the energy of the good battery to not be used efficiently anymore.
To efficiently operate fuel cells in series, you must be careful that each individual cell is supplied with the same amount of fuel and oxygen to keep the cell impedance at a similar level. This can be handled efficiently in large setups like those used in power plants, since there are no space constraints for dealing with some overhead for controlling fuel and air flow. Portable equipment, on the other hand, has severe space constraints as well as requirements for being able to operate at any orientation. This may be easier to handle with one cell only.
Solar cell panels need to be protected against partial shading. Cells that can not see light have high impedance. These cells can be destroyed with the voltage generated by the other cells that do see more light. Due to the typical usage in portable equipment, the risk of partial shading is very high. This means that, in this case, the most simple and reliable setup of solar cells is a single-cell configuration as well.
The impact on the maximum output power generated by two cells connected either in series or in parallel is shown in Figure 1. In this analysis, both cells have the same voltage. The cell impedance of both cells is different. The mismatch ratio describes the difference. For example, at the Ratio 1, both cells are the same. At Ratio 2, one cell has double the impedance of the other.
It easily can be seen that in a parallel configuration shown in the upper curve, in worst case, the maximum available output power drops to 50 percent. The lower curve shows the results for a configuration in series. At high imbalance ratios this curve goes down to zero. This means if imbalance is expected to be significant, a parallel configuration is more reliable for providing at least some power.
In this case, the supply voltage of the power circuit can get very low. In the solar cell example used above, it could be anything between 0.5V and 0.3V. This means a DC/DC converter, which is capable of working with that input, must be able to start at 0.5V and operate down to at least 0.3V supply voltage. The lower the possible minimum input voltage of the converter, the longer reliable supply can be maintained.
For systems using a battery for buffering the energy, a low startup voltage is not critical if the battery can be used for powering the DC/DC converter control circuit. But this kind of configuration can never recover when the battery is deeply discharged, and a protection circuit has disconnected the cell. Low voltage startup capability increases the reliability of the circuit.
When dealing with energy storage devices in energy harvesting applications, often it is more useful to store the energy at a higher voltage level. Usually this is more efficient since lower current levels can be used in the application, and parasitic resistances in connections and headers are not that critical. Also, in capacitors the stored energy is proportional to the square of the voltage which may make it easier to achieve higher volumetric energy densities. This implies using a boost converter for generating the higher voltage. Compared to standard boost converter designs where the output voltage is regulated accurately, the output voltage here depends on the charge state of the storage device, for example a big output capacitor. The boost converter should be able to run stable under this condition.
Figure 2 shows a solar cell powered DC/DC converter module based on the TPS61200. In this example it is used to charge a Li-Ion battery. It is configured so that the TPS61200 will try to regulate the output voltage and draw as much current as possible from the input. The high cell impedance of the solar cell in this case is limiting the input current, and the voltage at the solar cell may even drop to zero. As soon as the battery is charged, the TPS61200 regulates the output voltage. This prevents the battery from being overcharged. If the implemented power save mode of the TPS61200 is enabled, it will also make sure that no energy is wasted in this condition. The device just stays idle and starts operating again if the battery voltage drops and solar cell voltage is available.
In this application the current drawn from the solar cell is limited by the solar cell impedance, or the current limit of the TPS61200. If the solar cell impedance gets very high, it is possible for the output voltage of the cell to drop to 0V. This means that no energy can be taken out of the cell since it is not operated at its optimum operating condition. Adding an operational amplifier to feed the inverted solar cell voltage into the TPS61200 current control loop, in addition to the output voltage, makes sure that the TPS61200 converter input current is regulated lower at weak lighting conditions. This enables the circuit to even generate output power at very weak lighting conditions and operate the solar cell close to its maximum output power operating condition. Figure 3 shows the circuit with these modifications.
Illustrations:
Figure 1Figure 2Figure 3
