MCP6271R Microchip Technology Inc., MCP6271R Datasheet - Page 32

• Manufacturer Part Number: MCP6271R
• Description: 170 ?a, 2 Mhz Rail-to-rail Op Amp
• Manufacturer: Microchip Technology Inc.

Optimizing the external-peripheral power trade offs is also
important. In addition, you will find real power savings when
using the external and internal peripherals in concert with
the microcontrollers programming capability. For instance, the
microcontroller controls the power-supply voltage by switching
a new configuration into the resistive feedback system of the
MCP1252-ADJ regulated, adjustable-output charge pump. The
charge pump generates a higher output voltage to insure that the
analog circuitry performs at its optimum level. Digital events from
the microcontroller can tolerate a lower power-supply voltage.
For instance, the power-supply specifications of the PIC18F1320
are from 2V to 5.5V. You can calculate the power savings for
this type of change as a direct ratio of the two voltages from the
charge pump. Power savings are further improved if the external
peripherals are powered down with the lower-power supply
voltage using the I/O ports.
Controlling Your Clocks
One issue that is often overlooked when designers are trying to
reduce the overall power consumption of an embedded-system
circuit is the management of the clock when the microcontroller
comes out of its sleep mode.
Figure 2: This diagram illustrates an example of three oscillator
sources for a microcontroller. Externally, there are two clock
connections (Timer 1 and Primary Oscillator). Internally, there is
an oscillator.
A microcontroller can have a variety of clock sources (Figure
2). The most obvious clock source is an external one. In
this instance, you would connect a crystal oscillator, ceramic
resonator, an internal controller clock, or a clock generator to the
appropriate device pin. Beyond these elements that generate the
clock signal the microcontroller can have a postscaler, prescaler
Frequency-lock-loop (FLL). The postscaler and prescaler divide the
input-clock frequency down. You would use the FLL to multiply the
input-clock frequency.
In a Real Time Operating System (RTOS), it is critical to have a
clock management strategy when the system wakes up for short
periods of time and sleeps for a long period after the wake-up.
If the wake-up time is typically <1 second and you are using a
crystal oscillator or ceramic resonator, you may find that there will
be a delay between pulling out of the sleep mode and beginning
to execute code. The microcontroller will not execute code during
this delay or start-up time. However, the application circuit will be
consuming power.
30
Analog and Interface Guide – Volume 2
Miscellaneous Articles
For example, Figure 3 shows the typical start-up time for a
4 MHz crystal oscillator. In Figure 3, this time is approximately
450 msec. If this crystal oscillator were the only clock connected
to the microcontroller and one second was allocated for the code
execution, the actual execution time of the code would be 45%
longer than expected. During the clock start-up time your circuit
is consuming power but not executing code.
Figure 3: The clock connected to the microcontroller is a 4 MHz
LP crystal. In this oscilloscope plot, Channel 1 is a microcontroller
output pin. The programming of the microcontroller toggles this
output pin every 12 instruction cycles. Channel 2 shows the
power-supply to the microcontroller, V
microcontroller becomes active approximately 450 msecs after
power-supply start-up.
With this type of application, it might be advisable to use an
internal clock to execute code. The internal clock will start up
nearly instantaneously. It is not unusual for a 4 MHz internal
clock to start in a few microseconds. Figure 4 illustrates the
start-up time of an internal clock.
Figure 4: The start-up time of the internal clock is approximately
1.15 microsec. This is considerably faster than the start-up time of
the oscillator (in Figure 3).
Internal Clock Waking-Up
˜ 1.15 μsec
DD
. The external clock to the