Amidst the current explosion of portable electronic devices are important issues regarding battery maintenance. Learning the tricks and techniques related to the care and feeding of batteries is essential. Of particular interest are the dark secrets surrounding proper maintenance of high-energy rechargeables like highcapacity nickel-cadmium (NiCd) and nickel-metal-hydride (NiMH) cells.
When properly applied, batteries employing these popular chemistries offer energy densities of 150 to 300 Joules/gram (20 to 40 watt-hours/lb.), service life that spans hundreds of deep discharge/recharge cycles, and recharge times as short as one hour. Unfortunately, the critical proviso is “when properly applied,” because only a minority of these batteries escape abuse that can drastically shorten their life expectancy.
Chief among the traumas that doom NiCd and NiMH batteries to premature failure is improper termination of fast recharge cycles. In lead-acid and rechargeable-lithium batteries, cell voltage is an accurate indicator of charge level. Therefore, they can be safely charged using simple cell-voltage control criteria. On the other hand, nickel-based batteries display strong temperature effects that make voltagethreshold charge-termination criteria unreliable. Techniques for safe, highperformance recharge of nickel batteries must instead depend on monitoring battery behavior throughout the charging process. One such technique is the “delta-V” method (Fig. 1).
Delta-V depends on the fact that when NiCd or NiMH cells are subjected to a fast-recharge current (on the order of “1C” = 1 A per amperehour of cell capacity), the cell voltage behaves in a strange but consistent way. It rises slowly until the battery is roughly 90% recharged, then rapidly to a peak that occurs near the 115% point. At this stage, it reverses slope and declines at about 0.1% per 1% of overcharge thereafter.
Although the precise voltages associated with the Figure 1 curve depend upon details like battery construction, initial and ambient temperatures, and battery condition, the occurrence of the voltage reversal and its relationship to charge state are highly reproducible. It’s therefore used by the circuit in Figure 2 as the sole criterion for charge termination.
In the circuit’s operation, battery recharge begins with the closure of power switch S1. On power-up, the digitally controlled potentiometer P1 (Xicor XC9104) resets itself so that the wiper terminal is connected to VL, minimizing the voltage divider ratio between A1 and A2 and the battery. This state turns off A1, allowing its output to float high and enable the VR2 current regulator to apply −1.25 A to B1, placing it in rapid (>> 1C) recharge.
As B1 charges, its voltage rises along the trajectory illustrated in Figure 1, taking P1’s VW along with it. As VW rises, it periodically causes A2 to enable the A3 multivibrator, incrementing the ratio held by P1. P1 thus continuously tracks and stores E1’s peak voltage. This process continues until D1 reaches full charge (this takes about 1 hour) when the cell voltage maximizes, and then starts to decline.
Because P1 is arranged to increment unidirectionally, it now acts as a zerodroop analog peak-hold-memory, and thereby “remembers” the peak battery voltage. When B1’s voltage subsequently declines to the −10 mV/cell delta-V endof-charge criterion, A1’s output goes low and triggers A4 to disable VR2, terminating the recharge cycle.
If a visual charge-status indicator is desired, optional components R1, Q1, and LED D1 will handily provide one, with D1 glowing whenever the battery is charging. The circuit output can be scaled for simultaneous recharge of one to four series-connected cells by proper selection of R2. More than four cells can be accommodated if VIN and R2 are increased appropriately. Be sure to provide adequate heat sinking for VR2, which can dissipate up to 10 W.