Quiescent current has become an important specification, both when a regulator is being loaded and when it's in standby. Historically, quiescent current was not a concern. As electronic content in automobiles has increased, the limitations of present use batteries and alternators have been reached. Semiconductor fabrication processes can have an effect on the amount of quiescent current, an effect that we can see in the typical performance characteristics for products manufactured on two different types of fabrication. Figure 6 shows a device manufactured on a bipolar process while Figure 7 shows a device manufactured on a BCD process. Note the flat line nature of the device fabricated on the BCD process.
The bipolar process results in a device that requires increasing quiescent current at higher loads. The device manufactured on the BCD process will preserve the low quiescent current at low loads when operated at the higher loads. The result is a lower contribution to the module quiescent current limitation.
Current savers You can use a watchdog regulator in a current-saving maneuver. Watchdog regulators save current by sending a wakeup signal to a microprocessor. While microprocessor instructions are being set in motion, a concurrent signal is sent back to the voltage regulator by the microprocessor, notifying the regulator that it must maintain regulation. Once the microprocessor completes its commands and directives, the feedback signal back to the regulator is removed. The watchdog regulator recognizes this event and sends a reset signal back to the microprocessor, shutting it down, shown in Figure 8. The end result is less current draw until the microprocessor's efforts are needed again.
Another current-saving scheme new to the IC-regulator world is to momentarily power-down unneeded circuitry. Any part of the regulator that's not immediately needed can be powered down and operated in a pulsed on/off mode.
This scheme works best for lightly loaded conditions at cold or room temperature. Increased leakage current at higher temperatures (reached through an ambient temperature rise or die temperature rise caused by on-chip power) complicates proper operation of the device.
Interest in dual regulators (two independent output regulators on one chip) is increasing. Several microprocessors now require a dual supply voltage. One supply (usually the lower voltage) powers the core and the second powers the I/O. Lowering the core voltage allows more transistors to be squeezed on chip without melting the device or exceeding the thermal limitations of its package.
While dual linear regulator use is not intended as a quiescent current savings tool (it's more for convenience, space savings, and cost), they contribute to the power savings and distribution of power in the system. Current savings is a result of the common use of circuitry (such as the bandgap reference voltage and the current source bias strings) within the dual regulator.
Integrating multiple regulators on a single IC is good for convenience, space savings, and cost but is limited by the amount of allowable power in the IC.
Packaging is another area where improvements have allowed more power to be dissipated in a single package. Improvements in thermal resistance have been achieved by accessing the metal lead frame material (exposed pad). Heat can be dissipated more efficiently through the metal connection compared with its plastic counterpart. Figure 9 shows a typical exposed pad (epad) package. This device is a 300-mil, 16 lead SOW epad package, with an epad dimension of 150 mil x 184 mil.
Exceeding the manufacturer's limits on temperature (usually around 150°C/302°F at the junction) can either damage the regulator instantly or lead to early failures due to stress caused by the different thermal expansion coefficients of the silicon, the wire bonds, and the plastic package. Failure rates go up exponentially as temperature increases. Research is under way to increase the acceptable operating temperatures of these electrical components.
|