Design Articles

Safe, fast charging of NiCad batteries is attractive in many applications. Short charge time requires high current. A potential difficulty with high current charging is battery heating. Excessive internal heating degrades the battery and can cause gas venting to the outside atmosphere. Fast charge schemes based on monitoring cell voltage during charge suffer because cell voltage is not necessarily indicative of the battery’s charge state. Additionally, the battery’s charge-voltage relationship may alter over life and temperature. Similarly, open loop charging techniques involving high charge rates for a fixed time do not account for battery charge state or characteristic shifts over life and temperature.

One way to charge batteries without abuse is to measure cell temperature and taper the charge accordingly. This method is based on the fact that a discharged battery converts charging current to stored electrochemical energy, with relatively little heat produced. When the battery arrives at full charge the cell is saturated and cannot hold and more energy. As such, heat is produced, raising battery temperature. One way to detect this point is to measure cell surface temperature referred to ambient. An absolute temperature measurement is undesirable because cell temperature represents the summation of excess charging energy and ambient temperature. Additionally, the ambient and battery temperatures must be measured in phase. The thermal time constant of a battery pack can easily exceed on hour. If battery temperature is referred to a quickly responding ambient temperature poor charging characteristics can result. Consider the case of a portable computer retrieved from a locked automobile on a summer day. Passenger compartment temperature can exceed 120 degree F. The computer is brought inside, where the ambient temperature sensor quickly settles to 73 degrees F. The battery pack temperature is sitting at 120 degree F looking through a one hour thermal time constant. Under these conditions the system is fooled into believing the battery has just received a full charge, and no charge is delivered. The opposite effect occurs if the computer is in a car parked overnight in Minneapolis in January. These effects are avoidable by lagging the ambient temperature information with a time constant similar to the battery packs.