Product how-to: Optimize battery life for BPMDs and wireless medical devices
The medical device industry is seeing an explosion in the number and type of mobile and wireless medical devices1, and is ripe for a new approach to analyzing the energy requirements of your battery powered medical devices (BPMDs).
Traditional method & Test challenges
The common method to analyze the energy requirements of your BPMD is for the design engineer to assemble multiple instruments and miscellaneous external circuitry. Typically a scope, DMM, or digitizer is used for making measurements (usually two channels, one to measure voltage and the other current), a power supply and/or battery to power the medical device, and some shunt resistors for current measurements. A method to control the medical device (to test its different states of operation) as well as control the instrumentation to collect and analyze the desired data (current, voltage, power) will need to be developed. This can be manual, or semi-automated, by connecting the instruments to a computer and writing software to programmatically control the test. The result is typically multiple files with voltage and current waveforms for the different operating states of the device. It is left up to the designer to manipulate the data to determine power consumption for each of the operating states of the BPMD.
Figure 1 Traditional setup for measuring sleep current
The information available from this approach is limited. Because of the finite dynamic range of scopes, digitizers, and most DMMs (8-21 bits), different shunt resistors are required to measure the peak values (100mA to Amps) and the sleep current values (low microamps). What is missed is the critical understanding of the transient behavior from sleep mode to peak demand of your design. Additionally, many of the larger energy demands of the BPMD are dynamic and more challenging to characterize. In summary, this simple power consumption information doesn’t provide much insight into how you would change your BPMD design to improve reliability and optimize battery life.
An integrated solution
The value of an integrated solution goes beyond saving the time and cost of integrating your own system. With a solution focused on the test challenges mentioned above, you are removed from low value-added tasks like gathering and integrating the instrumentation and writing test programs. Instead, you can spend time analyzing more insightful results, giving you confidence to make design changes that lead to a more reliable and energy efficient BPMD.
The new approach uses a different instrument for measuring and analyzing your BPMD’s power consumption. A source measurement unit (SMU) is a standard instrument, available today, providing an ideal foundation for measuring power consumption. An SMU allows one to source a voltage/current and measure a current/voltage. Since the SMU knows what voltage it applies and the current that it measures, it is capable of making voltage, current, and power measurements without additional equipment. Assuming the SMU’s power rating is sufficient, it can also replace the power supply needed in the traditional method. Additionally, an SMU makes current measurements without the need for external shunt resistors.
Combining this measurement capability with a data-logger function allows the capture of voltage, current, and power over time. Not only can multiple operating states of the BPMD be obtained during the datalog record, the BPMD’s use of battery capacity (Ah) can also be calculated. Using an SMU based architecture significantly simplifies your setup to measure and analyze the power consumption of your BPMD. For our example, the world’s first Wi-Fi blood pressure monitor system (blip) was characterized.
Figure 2 blip wireless blood pressure cycle: 1) Initial sleep mode, 2) Pressing ‘User 1’ button to measure blood pressure (BP), 3) Inflating BP cuff, 4) Measuring BP, 5) Passing information for Wi-Fi___33 communication, 6) Wi-Fi transmitting BP measurement information to internet, 7) Display of BP measurement, 8) Return to sleep mode. Markers are set to capture information within the complete measurement cycle. Peak current (during pump operation) was 619mA. Drain on battery capacity for one cycle is 2.37mAh.
A state-of-the-art SMU has recently been developed with an effective resolution of 28 bits when measuring current. This dynamic range allows measurement of current spikes up to 3A, and measurement of sleep currents of tens of microamps with resolution of tens to hundreds of nanoamps, within the same digitization pass (Figure 3a). This dynamic range is obtained by seamless ranging between three different SMU current measurement ranges. The traditional AUTORANGE function found in DMMs involves switching in different front-end attenuation for different ranges, which can glitch the measurement system. This obviously distorts the real power consumption by the BPMD. Seamless ranging does not glitch the measurement. Additionally, most AUTORANGE functions aren’t fast enough to detect quick pulses in power and therefore may overload, or completely miss the pulse event. If only one range is used (e.g., 1V range with 1Ω shunt resistor for the blip blood pressure monitor) to capture the peak current and measure sleep current, the measured value of the sleep current is imperceptible in the A/D noise floor (~30µA) for this range (Figure 3b). Unlike the traditional method, this new approach provides accurate measurement of dynamic power transients (turning on actuator, RF transmission), enabling clearer understanding of the impact on battery life.
Figure 3b Traditional method using multiple instruments and Excel (Average = 40.3µA, Peak-Peak = 87µA)
Figure 3 Sleep mode current measurement
For a different analysis perspective on the power consumption of your BPMD, a complementary cumulative distribution function (CCDF) can be used. The CCDF is very useful to determine how much current was drawn during a specific percentage of one’s datalog record (Figure 4). The x-axis is a log scale of current, and the y-axis is a log scale of percent-of-time during the datalog record.
So, if the CCDF line near 690mA moves horizontally, this would suggest changes in the peak current value. Likewise, if the CCDF line near 30% moved vertically, this would indicate changes in the percentage of time the blip was displaying its measured result. The CCDF view provides new insight into which functions of your design are drawing the most current and for how long, enabling you to focus design improvements on the parts of your design that drain the battery most quickly. By comparing the CCDF graphs of your designs iterations, one can quickly verify the efficiencies gained. The CCDF tool can also be used to compare different hardware and firmware releases, so you can document the impact of these changes on power consumption of your BPMD. The CCDF provides invaluable insight and confidence to reliably optimize the power consumption of your BPMD.
Figure 4 Complementary Cumulative Distribution Function: 1) Sleep mode current, 2) Display result current, 3) BP measurement current, 4) BP cuff pump peak current. Markers show that 7.7% of the datalog record was spent drawing more than 100mA.
With the explosion of medical device types and volumes comes the need to better understand and design reliably optimized battery life for your BPMD. The traditional method leaves test challenges for the designer to measure the impact of dynamic and transient power needs, as well as in providing guidance where the designer should focus their effort to optimize the battery life of their BPMD. With the new integrated approach, the designer need not spend weeks developing a measurements system to measure energy consumption of their BPMD. Because of advancements in SMU technology like seamless ranging, and new ways to analyze one’s design (i.e., CCDF), one can more easily and confidently optimize the battery life of one’s BPMD. Agilent’s Battery Drain Solution (composed of an N6705B DC power Analyzer, an N6781A Source Measurement Unit (SMU) with seamless ranging, and the 14585A Control and Analysis software – Figure 5) was used to make measurements in Figures 2, 3, and 4.
Figure 5 Agilent’s Battery Drain Solution, and blip, the world’s first Wi-Fi Blood Pressure Monitor System
1 Patient monitoring applications reportedly grew at 23% compounded annual growth rate (CAGR) (2007-2011), with projections to continue at about that rate, reaching $20B by 2016 (Kalorama Information; July 2012). At the same time, the Association for the Advancement of Medical Instrumentation (AAMI) has identified battery management as one of the top 10 challenges for hospital’s biomedical departments. To address battery management concerns, the FDA recently (July 2013) held a workshop, Battery-Powered Medical Devices Workshop: Challenges and Opportunities, to raise awareness of the challenges and to collaboratively develop ways to ensure continued reliability from these devices. One of the goals from this workshop is to "Promote better design, manufacturing, testing, system integration, maintenance and standardization of battery-powered medical devices."
- Scripts automate source-measure units
- Race to the sun
- Solve MOSFET characteristic variation and reliability degradation issues
- Choosing the Optimal Source Measurement Unit Instrument for Your Test and Measurement Application
- Keithley SMU boasts touchscreen