MAX15034
Configurable, Single-/Dual-Output, Synchronous
Buck Controller for High-Current Applications
18 ______________________________________________________________________________________
Design Procedures
Inductor Selection
The switching frequency per phase, peak-to-peak ripple
current in each phase, and allowable voltage ripple at
the output, determine the inductance value. Selecting
higher switching frequencies reduces the inductance
requirement, but at the cost of lower efficiency due to
the charge/discharge cycle of the gate and drain
capacitances in the switching MOSFETs. The situation
worsens at higher input voltages, since capacitive
switching losses are proportional to the square of the
input voltage. Lower switching frequencies on the other
hand increase the peak-to-peak inductor ripple current
(∆I
L
), and therefore, increase the MOSFET conduction
losses (see the
Power MOSFET Selection
section for a
detailed description of MOSFET power loss).
When using higher inductor ripple current, the ripple can-
cellation in the multiphase topology, reduces the input
and output capacitor RMS ripple current. Use the follow-
ing equation to determine the minimum inductance value:
Choose ∆I
L
to be equal to approximately 30% of the out-
put current per channel. Since ∆I
L
affects the output-rip-
ple voltage, the inductance value may need minor
adjustment after choosing the output capacitors for full-
rated efficiency. Choose inductors from the standard
high-current, surface-mount inductor series available
from various manufacturers. Particular applications may
require custom-made inductors. Use high-frequency core
material for custom inductors. High ∆I
L
causes large
peak-to-peak flux excursion increasing the core losses at
higher frequencies. The high-frequency operation cou-
pled with high ∆I
L
, reduces the required minimum induc-
tance and even makes the use of planar inductors
possible. The advantages of using planar magnetics
include low-profile design, excellent current sharing
between phases due to the tight control of parasitics, and
low cost. For example, the minimum inductance at V
IN
=
12V, V
OUT
= 0.8V, ∆I
L
= 3A, and f
SW
= 500kHz is 0.5µH.
The average current-mode control feature of the
MAX15034 limits the maximum inductor current, which
prevents the inductor from saturating. Choose an
inductor with a saturating current greater than the
worst-case peak inductor current:
where 24.75mV is the maximum average current-limit
threshold for the current-sense amplifier and R
SENSE
is
the sense resistor.
Power MOSFET Selection
When choosing the MOSFETs, consider the total gate
charge, R
DS(ON)
, power dissipation, the maximum
drain-to-source voltage, and package thermal imped-
ance. The product of the MOSFET gate charge and on-
resistance is a figure of merit, with a lower number
signifying better performance. Choose MOSFETs opti-
mized for high-frequency switching applications. The
average gate-drive current from the MAX15034’s output
is proportional to the total capacitance it drives at DH1,
DH2, DL1, and DL2. The power dissipated in the
MAX15034 is proportional to the input voltage and the
average drive current. See the
Supply Voltage
Connections (V
IN
/V
REG
)
and the
Low-Side MOSFET
Drives Supply (V
DD
)
sections to determine the maxi-
mum total gate charge allowed from all driver outputs
together.
The losses may be broken into four categories: conduc-
tion loss, gate drive loss, switching loss, and output loss.
The following simplified power loss equation is true for
both MOSFETs in the synchronous buck-converter:
For the low-side MOSFET, the P
SWITCH
term becomes
virtually zero because the body diode of the MOSFET is
conducting before the MOSFET is turned on.
Tables 1 and 2 describe the different losses and shows
an approximation of the losses during that period.
Input Capacitance
The discontinuous input-current waveform of the buck
converter causes large ripple currents in the input
capacitor. The switching frequency, peak inductor cur-
rent, and the allowable peak-to-peak voltage ripple
reflected back to the source, dictate the capacitance
requirement. Increasing the number of phases increas-
es the effective switching frequency and lowers the
peak-to-average current ratio, yielding lower input
capacitance requirement. It can be shown that the
worst-case RMS current occurs when only one con-
troller section is operating. The controller section with
the highest output power needs to be used in determin-
ing the maximum input RMS ripple current requirement.
Increasing the output current drawn from the other out-
of-phase controller section results in reducing the input
.
