LTC3633A-2/LTC3633A-3
19
3633a23fb
For more information www.linear.com/LTC3633A-2
APPLICATIONS INFORMATION
Efficiency Considerations
The percent efficiency of a switching regulator is equal to
the output power divided by the input power times 100%.
It is often useful to analyze individual losses to determine
what is limiting the efficiency and which change would
produce the most improvement. Percent efficiency can
be expressed as:
% Efficiency = 100% – (L1 + L2 + L3 +…)
where L1, L2, etc. are the individual losses as a percent
-
age of input power.
Although all dissipative elements in the circuit produce
losses, three main sources usually account for most of the
losses in LTC3633A-2 circuits: 1) I
2
R losses, 2) switching
losses and quiescent power loss 3) transition losses and
other losses.
1. I
2
R losses are calculated from the DC resistances of the
internal switches, R
SW
, and external inductor, R
L
. In con-
tinuous mode, the average output current flows through
inductor L but is “chopped” between the internal top and
bottom power MOSFET
s. Thus, the series resistance look
-
ing into the SW pin is a function of both top and bottom
MOSFET R
DS(ON)
and the duty cycle (DC) as follows:
R
SW
= (R
DS(ON)TOP
)(DC) + (R
DS(ON)BOT
)(1 – DC)
The R
DS(ON)
for both the top and bottom MOSFETs can be
obtained from the Typical Performance Characteristics
curves. Thus to obtain I
2
R losses:
I
2
R losses = I
OUT
2
(R
SW
+ R
L
)
2. The internal LDO draws power from the SV
IN
input to
regulate the INTV
CC
rail. The total power loss here is
the sum of the switching losses and quiescent current
losses from the control circuitry.
Each time a power MOSFET gate is switched from low
to high to low again, a packet of charge dQ moves from
V
IN
to ground. The resulting dQ/dt is a current out of
INTV
CC
that is typically much larger than the DC control
bias current. In continuous mode, I
GATECHG
= f(Q
T
+ Q
B
),
where Q
T
and Q
B
are the gate charges of the internal
top and bottom power MOSFETs and f is the switching
frequency. For estimation purposes, (Q
T
+ Q
B
) on each
LTC3633A-2 regulator channel is approximately 2.3nC.
To calculate the total power loss from the LDO load,
simply add the gate charge current and quiescent cur
-
rent and multiply by the voltage applied to SV
IN
:
P
LDO
= (I
GATECHG
+ I
Q
) • SV
IN
3. Other “hidden” losses such as transition loss, cop-
per trace resistances, and internal load currents can
account for additional efficiency degradations in the
overall power system. Transition loss arises from the
brief amount of time the top power MOSFET spends
in the saturated region during switch node transitions.
The L
TC3633A-2 internal power devices switch quickly
enough that these losses are not significant compared
to other sources.
Other losses, including diode conduction losses during
dead-time and inductor core losses, generally account
for less than 2% total additional loss.
Thermal Considerations
The LTC3633A-2 requires the exposed package backplane
metal (PGND) to be well soldered to the PC board to
provide good thermal contact. This gives the QFN and
TSSOP packages exceptional thermal properties, which
are necessary to prevent excessive self-heating of the part
in normal operation.
In a majority of applications, the LTC3633A-2 does not
dissipate much heat due to its high efficiency and low
thermal resistance of its exposed-back QFN package.
However, in applications where the LTC3633A-2 is running
at high ambient temperature, high input supply voltage,
high switching frequency, and maximum output current
load, the heat dissipated may exceed the maximum junc
-
tion temperature of the part. If the junction temperature
reaches approximately 150°C, both power switches will
be turned off until temperature returns to 140°C.
To prevent the L
TC3633A-2 from exceeding the maximum
junction temperature of 125°C, the user will need to do
some thermal analysis. The goal of the thermal analysis
is to determine whether the power dissipated exceeds the
maximum junction temperature of the part. The tempera
-
ture rise is given by:
T
RISE
= P
D
θ
JA
LTC3633A-2/LTC3633A-3
20
3633a23fb
For more information www.linear.com/LTC3633A-2
APPLICATIONS INFORMATION
As an example, consider the case when one of the regu-
lators is used in an application where V
IN
= SV
IN
= 12V,
I
OUT
= 2A, frequency = 2MHz, V
OUT
= 1.8V. From the R
DS(ON)
graphs in the Typical Performance Characteristics section,
the top switch on-resistance is nominally 145mΩ and the
bottom switch on-resistance is nominally 70mΩ at 70°C
ambient. The equivalent power MOSFET resistance R
SW
is:
R
DS(ON)
TOP
1.8V
12V
+R
DS(ON)
BOT
10.2V
12V
= 81.3m
From the previous section’s discussion on gate drive, we
estimate the total gate drive current through the LDO to be
2MHz • 2.3nC = 4.6mA, and I
Q
of one channel is 0.65mA
(see Electrical Characteristics). Therefore, the total power
dissipated by a single regulator is:
P
D
= I
OUT
2
• R
SW
+ SV
IN
• (I
GATECHG
+ I
Q
)
P
D
= (2A)
2
• (0.0813Ω) + (12V) • (4.6mA + 0.65mA)
= 0.388W
Running two regulators under the same conditions would
result in a power dissipation of 0.776W. The QFN 5mm
× 4mm package junction-to-ambient thermal resistance,
θ
JA
, is around 43°C/W. Therefore, the junction temperature
of the regulator operating in a 70°C ambient temperature
is approximately:
T
J
= 0.776W • 43°C/W + 70°C = 103°C
which is below the maximum junction temperature of
125°C. With higher ambient temperatures, a heat sink
or cooling fan should be considered to drop the junc
-
tion-to-ambient thermal resistance. Alternatively, the
TSSOP package may be a better choice for high power
applications, since it has better thermal properties than
the QFN package.
Remembering that the above junction temperature is
obtained from an R
DS(ON)
at 70°C, we might recalculate
the junction temperature based on a higher R
DS(ON)
since
it increases with temperature. Redoing the calculation
assuming that R
SW
increased 12% at 103°C yields a new
junction temperature of 107°C. If the application calls for
a higher ambient temperature and/or higher load currents,
care should be taken to reduce the temperature rise of the
part by using a heat sink or air flow.
Figure 8. Temperature Derating Curve for DC1347 Demo Circuit
Figure 8 is a temperature derating curve based on the
DC1347 demo board (QFN package). It can be used to
estimate the maximum allowable ambient temperature
for given DC load currents in order to avoid exceeding
the maximum operating junction temperature of 125°C.
0
CHANNEL 1 LOAD CURRENT (A)
0.5
1.0
1.5
2.0
3.0
2.5
50
125
3633a23 F08
0
3.5
25 75 100
MAXIMUM ALLOWABLE AMBIENT
TEMPERATURE (°C)
CH2 LOAD = 0A
CH2 LOAD = 1A
CH2 LOAD = 2A
CH2 LOAD = 3A
Junction Temperature Measurement
The junction-to-ambient thermal resistance will vary de-
pending on the size and amount of heat sinking copper
on the PCB board where the part is mounted, as well as
the amount of air flow on the device. In order to properly
evaluate this thermal resistance, the junction temperature
needs to be measured. A clever way to measure the junction
temperature directly is to use the internal junction diode
on one of the pins (PGOOD) to measure its diode voltage
change based on ambient temperature change.
First
remove any external passive component on the PGOOD
pin, then pull out 100μA from the PGOOD pin to turn on its
internal junction diode and bias the PGOOD pin to a negative
voltage. With no output current load, measure the PGOOD
voltage at an ambient temperature of 25°C, 75°C and 125°C
to establish a slope relationship between the delta voltage on
PGOOD and delta ambient temperature. Once this slope is es
-
tablished, then the junction temperature rise can be measured
as a function of power loss in the package with corresponding
output load current. Although making this measurement with
this method does violate absolute maximum voltage ratings
on the PGOOD pin, the applied power is so low that there
should be no significant risk of damaging the device.
LTC3633A-2/LTC3633A-3
21
3633a23fb
For more information www.linear.com/LTC3633A-2
Board Layout Considerations
When laying out the printed circuit board, the following
checklist should be used to ensure proper operation of the
LTC3633A-2. Check the following in your layout:
1) Do the input capacitors connect to the PV
IN
and PGND
pins as close as possible? These capacitors provide
the AC current to the internal power MOSFETs and their
drivers.
2) The output capacitor, C
OUT
, and inductor L should be
closely connected to minimize loss. The (–) plate of
C
OUT
should be closely connected to both PGND and
the (–) plate of C
IN
.
3) The resistive divider, (e.g. R1 to R4 in Figure 9) must be
connected between the (+) plate of C
OUT
and a ground
line terminated near SGND. The feedback signal V
FB
should be routed away from noisy components and
traces, such as the SW line, and its trace length should
be minimized. In addition, the R
T
resistor and loop com-
pensation components should be terminated to SGND.
4)
Keep sensitive components away from the SW pin.
The R
T
resistor, the compensation components, the
feedback resistors, and the INTV
CC
bypass capacitor
should all be routed away from the SW trace and the
inductor L.
5) A ground plane is preferred, but if not available, the
signal and power grounds should be segregated with
both connecting to a common, low noise reference point.
The connection to the PGND pin should be made with
a minimal resistance trace from the reference point.
6) Flood all unused areas on all layers with copper in order
to reduce the temperature rise of power components.
These copper areas should be connected to the exposed
backside of the package (PGND).
Refer to Figures 10 and 11 for board layout examples.
Design Example
As a design example, consider using the LTC3633A-2 in
an application with the following specifications: V
IN(MAX)
=
13.2V, V
OUT1
= 1.8V, V
OUT2
= 3.3V, I
OUT(MAX)
= 3A, I
OUT(MIN)
= 10mA, f = 2MHz, V
DROOP
~ (5% • V
OUT
). The following
discussion will use equations from the previous sections.
APPLICATIONS INFORMATION
Because efficiency is important at both high and low load
current, Burst Mode operation will be utilized.
First, the correct R
T
resistor value for 2MHz switching fre-
quency must be chosen. Based on the equation discussed
earlier, R
T
should be 160k; the closest standard value is
162k. RT can be tied to INTV
CC
if switching frequency
accuracy is not critical.
Next, determine the channel 1 inductor value for about
40% ripple current at maximum V
IN
:
L1=
1.8V
2MHz 1.2A
1
1.8V
13.2V
= 0.64µH
A standard value of 0.68µH should work well here. Solving
the same equation for channel 2 results in a 1µH inductor.
C
OUT
will be selected based on the charge storage require-
ment. For a V
DROOP
of 90mV for a 3A load step:
C
OUT1
3 I
OUT
f V
DROOP
=
3 (3A)
(2MHz)(90mV)
= 50µF
A 47µF ceramic capacitor should be sufficient for channel 1.
Solving the same equation for channel 2 (using 5% of
V
OUT
for V
DROOP
) results in 27µF of capacitance (22µF is
the closest standard value).
C
IN
should be sized for a maximum current rating of:
I
RMS
= 3A
1.8V 13.2V 1.8V
( )
13.2V
= 1A
Solving this equation for channel 2 results in an RMS
input current of 1.3A. Decoupling each PV
IN
input with
a 47µF ceramic capacitor should be adequate for most
applications.
Lastly, the feedback resistors must be chosen. Picking
R1 and R3 to be 12.1k, R2 and R4 are calculated to be:
R2 = (12.1k) •
1.8V
0.6V
1
= 24.2k
R4 = (12.1k) •
3.3V
0.6V
1
= 54.5k
The final circuit is shown in Figure 9.

LTC3633AEFE-3#PBF

Mfr. #:
Manufacturer:
Analog Devices / Linear Technology
Description:
Switching Voltage Regulators Dual 3A, 20Vin, 4MHz, Monolithic Synchronous Step-Down Regulator
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union