BU-208: Cycling Performance

As part of ongoing research to examine performance degradation caused by cycling, Cadex tested a large volume of portable batteries for wireless communication devices. The population consists of nickel-cadmium, nickel-metal-hydride and lithium‑ion. The batteries were prepared by applying an initial charge, followed by a regime of full discharge/charge cycles. The internal resistance was measured with OhmTest™ and the self-discharge was obtained from time to time by reading the capacity loss incurred during a 48-hour rest period. The tests were carried out on the Cadex 7000 Seriesbattery analyzers.


In terms of life cycling, nickel-cadmium is the most enduring battery. Figure 1 illustrates the capacity, internal resistance and self-discharge of a 7.2V, 900mA pack with standard NiCd cells. Due to time constraints, the test was terminated after 2,300 cycles. The capacity remained steady; the internal resistance stayed low at 75mWand the self-discharge was stable. This battery receives a grade “A” rating for almost perfect performance.

Performance of standard NiCd

Figure 1: Performance of standard NiCd (7.2V, 900mAh)
This battery receives an “A” rating for a stable capacity, low internal resistance and moderate self-discharge over many cycles.

Courtesy of Cadex

The ultra-high-capacity nickel-cadmium offers up to 60 percent higher specific energy compared to the standard version, however, this comes at the expense of reduced cycle life. In Figure 2 we observe a steady drop of capacity during 2,000 cycles, a slight increase in internal resistance and a rise in self-discharge after 1,000 cycles.

Performance of ultra-high-capacity NiCd

Figure 2: Performance of ultra-high-capacity NiCd (6V, 700mAh)
This battery offers higher specific energy than the standard version at the expense of reducedcycle life.

Courtesy of Cadex


Figure 3 examines NiMH, a battery that offers high specific energy at a reasonably low cost. We observe good performance at first but past the 300-cycle mark, the capacity starts to drift downwards rapidly. One can detect a swift increase in internal resistance and self-discharge after cycle count 700. NiMH has a higher specific energy than nickel-cadmium and does not contain toxic metals. The test battery was an older generation; new NiMH performs better.

Performance of NiMH

Figure 3: Performance of NiMH (6V, 950mAh)
This battery offers good performance at first but past 300 cycles, the capacity, internal resistance and self-discharge start to increase rapidly. Newer NiMH has better results.

Courtesy of Cadex


Figure 4 examines the capacity and internal resistance of lithium-ion. We observe a gentle and predictable capacity drop over 1,000 cycles while the internal resistance increases only slightly. Because of low readings, we omit self-discharge. Lithium-ion offers the highest specific energy among the above-mentioned chemistries, contains little or no toxic metals, but needs protection circuits to ensure safe operation. Li-ion is also more expensive to manufacture than the nickel-based equivalent.

Batteries tested in a laboratory environment tend to give better results than when used in the field; elements of stress in everyday use do not transfer well into the laboratory. Aging plays a minimal role in a lab because the batteries are cycled over a period of a few months rather than the expected service life of a few years. The temperature is often moderate and the batteries are charged with proper charge equipment, an advantage that the field cannot always claim.

Performance of lithium-ion

Figure 4: Performance of lithium-ion (3.6V, 500mA)
Lithium-ion offers good capacity and steady internal resistance over 1,000 cycles.
Self-discharge was omitted because of low readings

Courtesy of Cadex

The load signature of the discharge plays an important role when testing batteries, and our laboratory batteries were discharged with an even DC load. Cellular phones and other digital devices draw pulsed loads that stress the battery more than with DC. One could argue, however, that the lab tests apply a full discharge whereas the field user discharges the battery to about 80 percent. The degradation of a battery receiving a 100 percent discharge with a DC load may not be the same as an 80 percent discharge on a pulsed load, and we keep this possible discrepancy in mind when studying the results.

The tests were done with batteries from an earlier generation. Newer models show improved results, and this is especially apparent with NiMH. The internal resistance of the modern NiMH is similar to NiCd, so is the cycle life. The Li-ion battery tested was Li-cobalt for cellular phones. We excluded lead acid from the test because this battery is seldom used for portable applications. Lead acid is heavy and does not cycle well, especially on full discharges.

The outcome of battery tests depends very much on the application for which the battery is designed, and we distinguish between consumer and industrial use. With the advent of the electric powertrain, a new category of batteries is emerging. Built for safety and longevity, these batteries have a specific energy that is typically only one-half that of consumer batteries. 


On May 12, 2011 at 6:14am
Mindaugas wrote:

I always feel a lack of the date when the article was published. Therefore, it is complicated to relate to the present situation and understand how the situation could be changed until now.

On August 12, 2011 at 9:00am
wayne robey wrote:

The article should more precisely indicate the battery under test. For example NiCd cell types include sintered plate, pocket plate and plated fiber. The sintered plate has the highest discharge rate and the plated fiber may the longest cycle life and a high discharge rate.

On December 14, 2011 at 1:40pm
David T. wrote:

I am currently cycling nicad battery packs at the rate of 1C/10C charge/discharge. My tests indicate that the Cycle Performance is effected by rate of discharging and charging. Can Cadex engineers advise under what charge/discharge rate the above Ni-cad curves were obtained. And is there any derating factor that can be applied to the curves for various charge/discharge rates. Look forward to your response. Thank you.

On November 25, 2012 at 5:13am
John Dubalja wrote:

Yeah I agree. NiCDs definitely are…