Power consumption is a key design specification in autonomous sensor solutions but information about the real power consumption of devices is most of the times hard to get. Knowing exactly how the components you will include in your design behave will save you lots of headaches along the road.

Following our “How to” Series, in today’s post we will write on how to solve some of the problems of high power consumption in devices connected to passive energy harvesters. We briefly discussed about key things to take into account when designing passive RFID sensors in this post.

First of all, I want to stress that the electronics market is increasingly moving towards smaller devices with lower power consumption. The power consumption of these devices is in the order of micro or nano amps. This is great news for us designers and manufacturers because it increases the possibilities for new designs for battery-free solutions.

However, one has to carefully read the information provided by manufacturers regarding these “ultra-low power” consumption devices. In some cases the information about current consumptions can be a little bit confusing. Logically, manufacturers have to announce and highlight the best characteristics of their devices to make them desirable for designers. But beware when you read these dazzling data in datasheets, because generally the information provided is only true under specific conditions.

As a general rule, devices have different power consumption profiles depending on the state they are working at each moment. Some cases in which current consumption could be altered are:

  • Operation mode
  • State and time of communication
  • Sleep modes
  • Wake up

Often times the current consumption information given in the datasheets of these particular cases is not shown in the main charts. In other instances, the information is hidden in a little tiny line or you have to search a little bit in charts and diagrams. There’s even a chance that this specific information you are looking for is not specified. So you have to be careful when designing because a good architecture may not work as you expected if you don’t pay attention to each of the energy states of these “ultra-low power” consumption devices.

Choosing an appropriate operation mode

The first thing to take into account is choosing the appropriate operation mode, in case the device has it. You will have to choose the low power modes. In these states, devices reduce their internal operations so that current consumption is lower than in normal modes. Probably losing resolution, accuracy, speed or any other characteristics.

There are devices without low-power modes but with sleep or stand-by modes. In these cases, an easy solution to reduce power consumption is to turn on the device during the data acquisition and then switch to sleep mode once the data acquisition is over. This way the device is on during a very short period of time.

Next charts are an example of the current consumptions related to the different operation modes. The first image shows part of the electrical characteristics of the ST LIS331DLH accelerometer.

It is one of the main charts of the datasheet where we can appreciate in red that there are only three bits of data related with current consumption. Not too much information. For more information we have to search a little bit deeper in the huge amount of data of the applications notes, as we can see in the next chart.

This second image shows different current consumptions according to the different output data rates of this sensor. The information contained in these charts is more vital than the information shown in the main chart. It is very important to have a good look at all the documentation provided by the manufacturer in order to avoid undesired problems.

Considering appropriate activity in your system

Another consideration to keep in mind related with current consumption is the energy need of transmission lines. Current consumption always increases when the communication port (SPI, I2C…) is active. We cannot eliminate this consumption but we can reduce it by reducing the data acquisition. So, by minimizing the active state of the communication port we will reduce the average current consumption.

Using a capacitor to cope with average power consumption

An easy way to solve some of the current consumption problems is to place a capacitor in the main power line so that the current requirements induced by the on-state or the communication port will be supplied by this capacitor. Energy will be stored in this capacitor during the short time the system is inactive while their harvesting properties are enabled.

Power consumption of the sensors when idle is very low so the system is actually charging the capacitor to be able to supply the sensors during their data acquisition time.

Dealing with peak current consumptions

However, the worst unexpected cases of current consumption happen when the devices are turned on, just before the device is enabled. Paradoxically, that’s when the device is still off.

While current consumption data for normal modes or on-states are more or less defined in the data sheets, it’s unusual to find peak current consumption information of the switch on phase when turning on these devices. And this is remarkable, because in most low power devices, the highest current consumption appears during the transient states in which the different blocks of the devices are activated.

As an example, the next image shows the current consumption behavior of the Linear Technology LT6656 voltage reference. We can appreciate an increase in current consumption before the device starts operating.

As you can see, peak current consumption can be 10 times higher than expected in your design, putting your design at risk.

In battery-assisted devices, this over-current consumption effect is very short and could be considered negligible because we have sufficient energy and a stable voltage. You will experience some difficulties when dealing with this peak current consumption when designing for energy harvester or battery free devices in which the energy available is limited and the time to attain stable voltage is long.

Harvested energy vs. time requirements

The key when designing is to find a good balance between the power consumption required by the application and the time to charge the capacitor. These two are very related:

  • The capacitor must be chosen so that it allows for the perfect performance of the sensor device. It has to be able to store enough energy so that the voltage never drops below the lowest supply voltage allowed.
  • At the same time, the bigger the capacitor, the longer the charging time, limiting the amount of times we can actually perform measurements.

It is a generally complicated balance to find and is a consideration need to be taken in the early stages of the design process. Selecting the right components – both the sensor and the capacitor – will make the whole design process easier while making bad choices may put the development at risk.

In the coming posts we will discuss about the solution we have given to our wireless battery free sensor tags. There is one thing I can share you in advance: we learned by doing. We know what it is to find unexpected facts about sensor performance not stated in their datasheets.

Have you been there? Share your experiences with the rest of us mortals. We might help somebody not to make the same mistakes we made.

Stay tuned to see what our solution is!