Where is my battery charge going?

The craft of design engineering, especially in areas undergoing rapid innovation, is ultimately about decision making. In the Internet of Things (IoT), there are overlapping objectives, that generally fall under the heading of the 5 C’s of IoT: Continuity, Connectivity, Compliance, Coexistence, and Cybersecurity. In some cases, improving one of the five C’s will help another. Robust connectivity will make a device better able to withstand certain types of cybersecurity attacks. Better coexistence behavior may well lead to fewer failed transmissions that must be retried, which in turn leads to improved continuity (battery runtime).

In other cases, it is possible that overemphasizing one of the 5 C’s could negatively impact the others. For example, while you can always add more and more cybersecurity features, many may come at the cost of an edge device’s continuity.

To make proper decisions quickly, engineers need quantitative data. What is the “cost” in terms of battery life to run this encryption processor on a data payload of a given size? How much charge is consumed if I measure this physical characteristic over 800 ms instead of 100 ms? Is the extra accuracy worth the extra charge consumption? The only way to know is to measure the cost of various events, and this event-based power analysis can give engineers very fast insights as to how they are spending their device battery’s charge and how long the battery may last.

Figure 1: Battery life estimate from Keysight KS833A1B event-based power analysis software

In the image below, event-based power analysis software has time-sliced a current waveform (yellow) according to what is happening in the device.

Figure 2: Waveform time-sliced by X8712A IoT device battery optimization solution

The green waveform indicates the RF power in dBm, and not surprisingly, the yellow current waveform spikes quickly during the RF activity. The cyan (light blue) waveform is the voltage line that powers an LED, and there is a current increase associated with the LED turning on at the right side of the image. The colorful bars at the bottom of the image reflect the analysis according to user-defined parameters: Standby, MCU, RF, and LED1. This information is then automatically converted into a graphical form that lets the design engineer see exactly how the charge is consumed.

Figure 3: Results from Keysight KS833A1B event-based power analysis software

Furthermore, because event-based power analysis software can also control the measurement instrument, validation engineers can test their devices under different conditions to find interesting or perhaps unexpected behaviors quickly. As an example, consider the following two waveforms, taken under very similar, but not identical, conditions.

Figure 4: Charge consumption and current consumption waveforms captured by Keysight X8712A IoT device battery optimization solution

Why do the waveforms look so different? What are the implications for the device’s battery life? And what choices does the device designer have to increase battery life? Check out the answers to these questions in my next blog post on this topic. In the meantime, for more information on IoT device battery life and Keysight’s event-based power analysis software, refer to the links below.