MP38873– 15A, 16V, HIGH FREQUENCY STEP-DOWN WITH SYNCHRONOUS GATE DRIVER
MP38873 Rev. 0.9 www.MonolithicPower.com 10
9/17/2009 MPS Proprietary Information. Unauthorized Photocopy and Duplication Prohibited.
© 2009 MPS. All Rights Reserved.
tantalum, or low ESR electrolytic capacitors are
recommended. Low ESR capacitors are
preferred to keep the output voltage ripple low.
The output voltage ripple can be estimated by:
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
××
+×
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
−×
×
=Δ
2Cf8
1
R
V
V
1
Lf
V
V
S
ESR
IN
OUT
S
OUT
OUT
Where L is the inductor value and R
ESR is the
equivalent series resistance (ESR) value of the
output capacitor.
In the case of ceramic capacitors, the
impedance at the switching frequency is
dominated by the capacitance. The output
voltage ripple is mainly caused by the
capacitance. For simplification, the output
voltage ripple can be estimated by:
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
−×
×××
=
IN
OUT
2
S
OUT
OUT
V
V
1
2CLf8
V
V
In the case of tantalum or electrolytic capacitors,
the ESR dominates the impedance at the
switching frequency. For simplification, the
output ripple can be approximated to:
ESR
IN
OUT
S
OUT
OUT
R
V
V
1
Lf
V
V ×
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
−×
×
=
The characteristics of the output capacitor also
affect the stability of the regulation system. The
MP38873 can be optimized for a wide range of
capacitance and ESR values.
Compensation Components
MP38873 employs current mode control for
easy compensation and fast transient response.
The system stability and transient response are
controlled through the COMP pin. COMP pin is
the output of the internal error amplifier. A
series capacitor-resistor combination sets a
pole-zero combination to control the
characteristics of the control system. The DC
gain of the voltage feedback loop is given by:
OUT
FB
VEACSLOADVDC
V
V
AGRA ×××=
Where A
VEA is the error amplifier voltage gain,
9600V/V; G
CS is the current sense
transconductance, 12.8A/V; R
LOAD is the load
resistor value.
The system has two poles of importance. One
is due to the compensation capacitor (C3), the
output resistor of error amplifier. The other is
due to the output capacitor and the load resistor.
These poles are located at:
VEA
EA
1P
A3C2
G
f
××π
=
LOAD
2P
R2C2
1
f
××π
=
Where, G
EA is the error amplifier
transconductance, 2.4mA/V.
The system has one zero of importance, due to
the compensation capacitor (C3) and the
compensation resistor (R3). This zero is located
at:
3R3C2
1
f
1Z
××π
=
The system may have another zero of
importance, if the output capacitor has a large
capacitance and/or a high ESR value. The zero,
due to the ESR and capacitance of the output
capacitor, is located at:
ESR
ESR
R2C2
1
f
××π
=
In this case (as shown in Figure 3), a third pole
set by the compensation capacitor (C6) and the
compensation resistor (R3) is used to
compensate the effect of the ESR zero on the
loop gain. This pole is located at:
3R6C2
1
f
3P
××π
=
The goal of compensation design is to shape
the converter transfer function to get a desired
loop gain. The system crossover frequency
where the feedback loop has the unity gain is
important. Lower crossover frequencies result
in slower line and load transient responses,
while higher crossover frequencies could cause
system unstable. A good rule of thumb is to set
the crossover frequency to approximately one-
tenth of the switching frequency. The Table 3
lists the typical values of compensation
components for some standard output voltages
with various output capacitors and inductors.
The values of the compensation components
have been optimized for fast transient.
