MAX15021
Dual, 4A/2A, 4MHz, Step-Down
DC-DC Regulator withTracking/
Sequencing Capability
Maxim Integrated | 14www.maximintegrated.com
Use the following equation to calculate the input ripple
when only one regulator is enabled:
The MAX15021 includes UVLO hysteresis to avoid possi-
ble unintentional chattering during turn-on. Use additional
bulk capacitance if the input source impedance is high. If
using a lower input voltage, additional input capacitance
helps to avoid possible undershoot below the undervolt-
age lockout threshold during transient loading.
Output-Capacitor Selection
The allowed output-voltage ripple and the maximum
deviation of the output voltage during load steps deter-
mine the required output capacitance and its ESR. The
output ripple is mainly composed of ΔV
Q
(caused by
the capacitor discharge) and ΔV
ESR
(caused by the
voltage drop across the equivalent series resistance of
the output capacitor). The equations for calculating the
output capacitance and its ESR are:
where ΔI
P-P
is the peak-to-peak inductor current, and
f
SW
is the switching frequency.
ΔV
ESR
and ΔV
Q
are not directly additive since they are
out of phase from each other. If using ceramic capaci-
tors, which generally have low ESR, ΔV
Q
dominates. If
using electrolytic capacitors, ΔV
ESR
dominates.
The allowable deviation of the output voltage during
fast load transients also affects the output capacitance,
its ESR, and its equivalent series inductance (ESL). The
output capacitor supplies the load current during a
load step until the controller responds with an
increased duty cycle. The response time (t
RESPONSE
)
depends on the gain bandwidth of the controller (see
the
Compensation-Design Guidelines
section). The
resistive drop across the output capacitor’s ESR
(ΔV
ESR
), the drop across the capacitor’s ESL (ΔV
ESL
),
and the capacitor discharge (ΔV
Q
) cause a voltage
droop during the load-step (I
STEP
). Use a combination
of low-ESR tantalum/aluminum electrolyte and ceramic
capacitors for better load transient and voltage ripple
performance. Nonleaded capacitors and capacitors in
parallel help reduce the ESL. Keep the maximum out-
put voltage deviation below the tolerable limits of the
electronics being powered.
Use the following equations to calculate the required
output capacitance, ESR, and ESL for minimal output
deviation during a load step:
where I
STEP
is the load step, t
STEP
is the rise time of the
load step, and t
RESPONSE
is the response time of the
controller.
Compensation-Design Guidelines
The MAX15021 uses a fixed-frequency, voltage-mode
control scheme that regulates the output voltage by
comparing the output voltage against a fixed reference.
The subsequent “error” voltage that appears at the
error-amplifier output (COMP_) is compared against an
internal ramp voltage to generate the required duty
cycle of the pulse-width modulator. A second-order
lowpass LC filter removes the switching harmonics and
passes the DC component of the pulse-width-modulat-
ed signal to the output. The LC filter has an attenuation
slope of -40dB/decade and introduces 180° of phase
shift at frequencies above the LC resonant frequency.
This phase shift in addition to the inherent 180° of
phase shift of the regulator’s negative feedback system
turns the feedback into unstable positive feedback. The
error amplifier and its associated circuitry must be
designed to achieve a stable closed-loop system.
The basic controller loop consists of a power modulator
(comprised of the regulator’s pulse-width modulator,
associated circuitry, and LC filter), an output feedback
divider, and an error amplifier. The power modulator has
a DC gain set by V
AVIN
/V
RAMP
where the ramp voltage
(V
RAMP
) is a function of the V
AVIN
and results in a fixed
DC gain of 4V/V, providing effective feed-forward com-
pensation of input-voltage supply DC variations. The
feed-forward compensation eliminates the dependency
of the power modulator’s gain on the input voltage such
that the feedback compensation of the error amplifier
requires no modifications for nominal input-voltage
changes. The output filter is effectively modeled as a
double-pole and a single zero set by the output induc-
tance (L), the DC resistance of the inductor (DCR), the
output capacitance (C
OUT
) and its equivalent series
resistance (ESR).