How to test batteries

The two most important reasons for doing your own battery testing are:

• To verify that the characteristics of a battery from a manufacturer comply with the manufacturer’s specifications.
• To make sure that the battery performance meets the needs of your application.

While a battery’s chemistry will have baseline characteristics, such as tolerance to temperature variations or low internal resistance, there may be differences in a battery’s chemistry specifications. Sometimes these are explicitly stated in the manufacturer’s datasheet, other times you must test the battery. It is important to fully investigate a battery’s suitability for your specific application. For example, the capacity of a battery may be adequate, and the temperature tolerances well within the designed range, but transient current spikes could cause capacity degradation.

In the following diagram, there are two different devices running the same firmware. The second of these two devices uses a super-capacitor which dramatically reduces the peak current. Most Lithium-ion and Lithium-polymer batteries can handle the higher peak current (135mA); but most other batteries cannot due to their chemistry. In the following example, a lower peak current allows you to choose from more batteries.

Figure 4: Difference in peak currents for two different circuits for the same application

To verify battery performance, determine the characteristics of the operational load placed on the battery. Determine the average, peak and pulsed current levels. Also, determine how much time an over-stressed battery needs to recover and get back to nominal levels. It is also important to understand these characteristics in their environmental conditions. If the battery can withstand certain cycle stresses at room temperature but has accelerated degradation at extreme temperature (high or low), the environmental conditions must be factored into the testing and validation. You can use several different devices to determine the device profile, including:

• Digital multimeter (DMM)
• Oscilloscope with a current probe
• DC power analyzer
• Precision source/measure unit (SMU)
• Device current waveform analyzer

To evaluate the power profile of your target electronics, you can use these devices individually or together.

DMM and oscilloscope devices are relatively inexpensive and common. However, they typically provide quick answers rather than detailed information. These devices are particularly poor for evaluating the sleep current, which can be in the nano-amps. Getting an inaccurate measurement of sleep can have a significant impact on estimating overall battery life. You can combine these devices with other devices such as the DC power analyzer, Precision source/measure unit (SMU), or a Device current waveform analyzer for a more precise evaluation.

After you understand the profile of a specific application, you can use that profile information to test your batteries. You can test batteries in several ways and use different devices to help your evaluation. One of the most illustrative devices to use is an Electronic Load Tester. A load tester can simulate the draw for the target circuit as well as perform accelerated testing. This can be done under static (constant draw) or dynamic conditions, based on the tester’s capabilities. Using an Electronic Load Tester instead of the actual circuit provides these advantages:

• More control over the settings and measurements
• Ability to run automated tests and gather data for extended periods of time
• Ability to perform the equivalent of days or weeks of loading within mere days of testing due to an increase in the rate of dynamic loading

You will need to balance accelerated testing against the overall capabilities of the battery under test. Overloading the battery during accelerated testing does not show how the device will perform in operation. Therefore, make sure that any deficiencies you find during accelerated testing are due to battery limitations related to the application, not related to the acceleration itself.

Example battery pulse test

The example case uses an Electronic Load Tester where a nominal draw test (static) has been extended to include a peak test. The 20mA draw test is good for equivalent battery life. We conducted these tests at a fixed maximum current draw of 20mA; 10 mA per battery for a dual-battery configuration. We ran this to 2.5V or 0.1V. The 0.1V seemed to get a slightly higher rate (although this may be due to the batteries sitting in test devices longer). For a pulsed test, the current was set at two different levels for a specific amount of time.

In continuous transient operation, the load is continuously switched between two load values. Here is an example:

          Continuous fixed draw= 20 mAsec/sec

• Transient draw:
• Peak limit: 150mA
• Low limit 0.1mA
• Peak time (20 bytes, DR1): 350 ms
• Low time = T
• 150mA*Tp + 0.1mA*Tl = 20 mAsec/sec
• Tp = 0.35/Tt
• Tl = X/Tt
• 150*0.35+ 0.1*X = 20*Tt
• Tt = 0.35+X
• 52.5+0.1*Tt-0.035 = 20*Tt
• 52.465 = 19.9*Tt
• Tt = 2.6364
• X = 2.286 seconds

                Electronic Load Tester Settings:

To activate this continuous transient condition, press Shift = Tran, then press On/Off. The load will begin switching between the two Tran values with the timing you entered.

• Level A: .15A
• Width A: 350ms
• Level B: 0.0001A
• Width B: 2286ms
• Total cycle = 2.6364 sec
•