MIC4721 Datasheet, PDF (11/19 Pages) Micrel Semiconductor – 1.5A 2MHz Integrated Switch Buck Regulator

Micrel, Inc.
Efficiency Considerations
Calculating the efficiency is as simple as measuring
power out and dividing it by the power in.
Efficiency = ^POUT × 100
P_IN
Where input power (P_IN) is:
P_IN = V_IN × I_IN
and output power (P_OUT) is calculated as:
P_OUT = V_OUT × I_OUT
The Efficiency of the MIC2207 is determined by several
factors.
• R_DSON (Internal P-channel Resistance)
• Diode conduction losses
• Inductor Conduction losses
• Switching losses
R_DSON losses are caused by the current flowing through
the high side P-channel MOSFET. The amount of power
loss can be approximated by:
P_SW = R_DSON × I_OUT2 × D
Where D is the duty cycle.
Since the MIC4721 uses an internal P-channel
MOSFET, R_DSON losses are inversely proportional to
supply voltage. Higher supply voltage yields a higher
gate to source voltage, reducing the R_DSON, thus
reducing the MOSFET conduction losses. A graph
showing typical R_DSON vs. input supply voltage can be
found in the typical characteristics section of this
datasheet.
Diode conduction losses occur due to the forward
voltage drop (VF) and the output current. Diode power
losses can be approximated as follows:
P_D = V_F × I_OUT × (1 – D)
For this reason, the low forward voltage drop Schottky
diode is the rectifier of choice. The low forward voltage
drop will help reduce diode conduction losses, and
improve efficiency. Duty cycle, or the ratio of output
voltage to input voltage, determines whether the
dominant factor in conduction losses will be the internal
MOSFET or the Schottky diode. Higher duty cycles
place the power losses on the high side switch, and
lower duty cycles place the majority of power loss on the
Schottky diode.
Inductor conduction losses (PL) can be calculated by
multiplying the DC resistance (DCR) times the square of
the output current:
P_L = DCR × I_OUT2
Also, be aware that there are additional core losses
associated with switching current in an inductor. Since
most inductor manufacturers do not give data on the
May 2007
MIC4721
type of material used, approximating core losses
becomes very difficult, so verify inductor temperature
rise.
Switching losses occur twice each cycle, when the
switch turns on and when the switch turns off. This is
caused by a non-ideal world where switching transitions
are not instantaneous, and neither are current
transitions. Figure 6 demonstrates (or exaggerates…)
how switching losses due to the transitions dissipate
power in the switch.
Figure 6. Switching Transition Losses
Normally, when the switch is on, the voltage across the
switch is low (virtually zero) and the current through the
switch is high. This equates to low power dissipation.
When the switch is off, voltage across the switch is high
and the current is zero, again with power dissipation
being low. During the transitions, the voltage across the
switch (VS-D) and the current through the switch (IS-D)
are at midpoint of their excursions and cause the
transition to be the highest instantaneous power point.
During continuous mode, these losses are the highest.
Also, with higher load currents, these losses are higher.
For discontinuous operation, the transition losses only
occur during the “off” transition since the “on” transitions
there is no current flow through the inductor.
Component Selection
Input Capacitor
A 10µF ceramic is recommended on each VIN pin for
bypassing. X5R or X7R dielectrics are recommended for
the input capacitor. Y5V dielectrics lose most of their
capacitance over temperature and voltage and are
therefore not recommended. Also, tantalum and
electrolytic capacitors alone are not recommended
because of their reduced RMS current handling,
reliability, and higher ESR. Smaller case size capacitors
are recommended due to their lower ESL (equivalent
series inductance). Please refer to layout
recommendations for proper layout of the input
capacitors.
11
M9999-052907-A