How the right measurements can optimize battery run time - Embedded.com

How the right measurements can optimize battery run time

A version of this article appeared in the Dec. 2011/Jan. 2012 issue of Test & Measurement World. See the PDF.

Ifyou need to simply validate battery run time in a mobile device, youcan treat your device as a black box and either directly measure runtime or measure current drain for a prescribed period and extrapolaterun time based on the battery's amp-hour capacity. Many conformancetests only validate battery run time. Optimizing battery run time,however, usually requires you to use several test methods to gaindeeper insights into your device's power consumption. You need totest and characterize the device–its sub-circuits and thebattery–both independently and in combination.

Tocharacterize your device, you can capture long-term battery currentdrain data at high sample rates and over a wide dynamic range. Adetailed characterization and analysis of your device'sbattery-current-drain profiles lets you make informed tradeoffs foroptimizing run time. You can take a number of approaches measurepower consumption.

Youcan measure current consumption with an oscilloscope and currentprobes. Oscilloscopes provide high-speed waveform digitization, buttheir limited dynamic range, 8-bit resolution, and relatively highnoise floor add uncertainty to your measurements.

Anotheroption uses a high-sampling-speed, high-resolution, data-acquisitionsystem and a precision current shunt. The better resolution of adata-acquisition system (typically 12 bits or 16 bits) providesbetter accuracy and wider dynamic measurement range compared to acurrent probe and oscilloscope. You must, though, keep maximumtolerable current shunt peak voltage drop small so that it does notunduly affect the mobile device. Keeping the shunt voltage dropsufficiently small limits the measurement's dynamic range andlow-signal-level accuracy

DCsources that incorporate a high-speed digitizers wide-dynamic-rangealso let you accurately characterize a mobile device's currentdrain. This method eliminates the voltage drop issues associated withusing external shunt resistors.

 

Figure1Figure 1. A wireless device's current draw pulses during transmit operations, with small pulsed current during receive time.

Figure1 shows thecurrent-drain profile on a GSM/GPRS smart phone placed in a call. Theplot shows idle current base level value, idle period duration,current-drain values, and durations of activities. Detailedcurrent-drain profiles like this one let you see how the differentmodes of operation affect a mobile wireless device's battery runtime.

Whilemobile phones and many other mobile devices have high-power activemodes (which need to be optimized for battery run time), they oftenspend the majority of their time in standby or another similar typeof power-savings mode. Other wireless devices, like sensors, may haveonly power-savings operating modes. Although the power consumptionmay seem negligible, because the devices spend long periods in thesemodes, they can consume a major portion, or even all, of thebattery's capacity. Evaluating these power-savings operating modesis a top priority for optimizing battery run time. The nature of current drain, spanning several decades of amplitude, during power-savings operation makes them challenging to measure.

Wirelessdevices spend most of their time in a low-power sleep state.Periodically, the device wakes up and briefly enters a higher-poweractive state, often to transmit to and maintain contact with a basestation. The resulting current drain is pulsed and has the followingcharacteristics:

  • Long period of typically tenths to tens of seconds (even minutes or greater for wireless sensors, depending on their function),
  • Extremely low duty cycle of tenths of a percent to a few percent, and
  • Extremely high crest factor on the order of a few hundred or higher.

Thesleep state and the pulsed active-state currents are often bothsignificant portions of the overall average and hence both of theseextremes, and everything in between, need accurate measurements.

Figure1

Figure 2. Wireless temperature transmitter's current drain measurements show the difference between sleep current and transmit current.

Toillustrate this point, look at a representative example ofpower-savings operation. A wireless temperature transmitter has apulsed current drain with the following characteristics illustratedin Figure 2 .

  • Period of 4 s (5 divisions at 0.8 s/div),
  • Duty cycle of 0.17%, and
  • Approximately 21.8 mA peak and 53.7 µA average currents, for a crest factor of 400.

Thewireless temperature transmitter is an example of a very low powerdevice. While these characteristics are quite substantial, they areeven far more dramatic for cellular-based devices, certain medicaldevices, and other battery powered devices drawing peaks of hundredsof milliamps to amps, but having sleep currents of 10's to 100'sof microamps.

Optimizingbattery run time calls for much more than simply validating the runtime. You need to test and characterize your device, its subcircuits, and its battery, both independently and in combination as asystem. Capturing and analyzing detailed long-term battery currentdrain profiles provides deeper insights on the inner workings of thedevice for optimizing battery run time. The current draw of a wireless devices had a wide dynamic range between power-saving and operating modes. Thus, making the current measurements is particularly challenging.

EdBrorein holdsa BSEE from Villanova University and MSEE from New Jersey Instituteof Technology. Ed joined Agilent Technologies (at that time, HewlettPackard) in 1979 and worked as an R&D engineer, manufacturingengineer, and marketing engineer in many various roles and presentlyas an applications engineer in marketing.E-Mailed_brorein@agilent.com.

This article has also been published in EDN Magazine. 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.