Datasheet

Data Sheet ADP5033
Rev. F | Page 21 of 28
POWER DISSIPATION AND THERMAL CONSIDERATIONS
The ADP5033 is a highly efficient micropower management
unit (µPMU), and, in most cases, the power dissipated in the
device is not a concern. However, if the device operates at high
ambient temperatures and maximum loading condition, the
junction temperature can reach the maximum allowable
operating limit (125°C).
When the temperature exceeds 150°C, the ADP5033 turns off
all the regulators, allowing the device to cool down. When the
die temperature falls below 130°C, the ADP5033 resumes
normal operation.
This section provides guidelines to calculate the power dissi-
pated in the device and ensure that the ADP5033 operates
below the maximum allowable junction temperature.
The efficiency for each regulator on the ADP5033 is given by
100%×=
IN
OUT
P
P
η
(1)
where:
η is the efficiency.
P
IN
is the input power.
P
OUT
is the output power.
Power loss is given by
P
LOSS
= P
IN
P
OUT
(2a)
or
P
LOSS
= P
OUT
(1− η)/η (2b)
Power dissipation can be calculated in several ways. The most
intuitive and practical is to measure the power dissipated at the
input and all the outputs. Perform the measurements at the
worst-case conditions (voltages, currents, and temperature).
The difference between input and output power is dissipated in
the device and the inductor. Use Equation 4 to derive the power
lost in the inductor and, from this, use Equation 3 to calculate
the power dissipation in the ADP5033 buck converter.
A second method to estimate the power dissipation uses the
efficiency curves provided for the buck regulator, and the power
lost on each LDO can be calculated using Equation 12. When
the buck efficiency is known, use Equation 2b to derive the total
power lost in the buck regulator and inductor, use Equation 4 to
derive the power lost in the inductor, and then calculate the
power dissipation in the buck converter using Equation 3. Add
the power dissipated in the buck and in the two LDOs to find
the total dissipated power.
Note that the buck efficiency curves are typical values and may
not be provided for all possible combinations of V
IN
, V
OUT
, and
I
OUT
. To account for these variations, it is necessary to include a
safety margin when calculating the power dissipated in the buck.
A third way to estimate the power dissipation is analytical and
involves modeling the losses in the buck circuit provided by
Equation 8 to Equation 11 and the losses in the LDO provided
by Equation 12.
BUCK REGULATOR POWER DISSIPATION
The power loss of the buck regulator is approximated by
P
LOSS
= P
DBUCK
+ P
L
(3)
where:
P
DBUCK
is the power dissipation on one of the ADP5033 buck
regulators.
P
L
is the inductor power losses.
The inductor losses are external to the device, and they do not
have any effect on the die temperature.
The inductor losses are estimated (without core losses) by
P
L
I
OUT1(RMS)
2
× DCR
L
(4)
where:
DCR
L
is the inductor series resistance.
I
OUT1(RMS)
is the rms load current of the buck regulator.
12
+1
)(1
r
II
OUT1
RMSOUT
×=
(5)
where r is the normalized inductor ripple current
r = V
OUT1
× (1 − D)/(I
OUT1
× L × f
SW
) (6)
where:
L is the inductance.
f
SW
is the switching frequency.
D is the duty cycle.
D = V
OUT1
/V
IN1
(7)
The ADP5033 buck regulator power dissipation, P
DBUCK
, includes
the power switch conductive losses, the switch losses, and the
transition losses of each channel. There are other sources of
loss, but these are generally less significant at high output load
currents, where the thermal limit of the application is. Equation 8
captures the calculation that must be made to estimate the
power dissipation in the buck regulator.
P
DBUCK
= P
COND
+ P
SW
+ P
TRAN
(8)
The power switch conductive losses are due to the output current,
I
OUT1
, flowing through the P-MOSFET and the N-MOSFET
power switches that have internal resistance, RDS
ON-P
and
RDS
ON-N
. The amount of conductive power loss is found by
P
COND
= [RDS
ON-P
× D + RDS
ON-N
× (1 − D)] × I
OUT1(RMS)
2
(9)
where RDS
ON-P
is approximately 0.2 Ω, and RDS
ON-N
is approxi-
mately 0.16at 25°C junction temperature and VIN1 = VIN2 =
3.6 V. At VIN1 = VIN2 = 2.3 V, these values change to 0.31 Ω and
0.21 Ω, respectively, and at VIN1 = VIN2 = 5.5 V, the values are
0.16 Ω and 0.14 Ω, respectively.