Analyze Distribution Profiles to Quickly Optimize Battery Run Time
Technical Overviews
Introduction
When optimizing battery run time, you need a way to quickly and easily visualize and quantify impact of design changes on a wireless mobile device’s long-term current drain.
The activities of various sub circuits in a mobile device are by nature random over time, depending on user behavior, the network environment it operates in and complexity of the device itself. The battery current drain associated with these sub circuits and activities is correspondingly random over time as a result. Validating improvements from design changes for optimizing battery run time requires you to log the current drain over a suficiently long timeframe to average out random behavior. The difference in average current before and after design changes, is the net impact of the changes. However, you need more detail about the impact of design changes when you are optimizing battery run time. Did you get the expected improvement? How do you determine which sub circuits and activities within the device were impacted? One approach you can take is to manually scroll through the data logs to estimate levels and durations of current bursts relating to various sub circuits and associated activities.
While useful, this approach has several drawbacks:
- It is extremely time consuming.
- Many values are estimates at best, due to the long-term random nature.
- It is easy to reach incorrect conclusions because of the dificulty of examining and quantifying countless millisecond-duration activities in up to hours-long data logs.
While long-term logging of a mobile device’s current drain is necessary, direct visual inspection to quantify the data log’s details is problematic. You need to resort to alternate methods to quickly and effectively analyze long-term current drain logs when optimizing battery run time.