TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
3-A DUAL NON-SYNCHRONOUS CONVERTER WITH INTEGRATED HIGH-SIDE MOSFET
Check for Samples: TPS54386-Q1
FEATURES
APPLICATIONS
•
•
•
•
1
23
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Qualified for Automotive Applications
AEC-Q100 Qualified With the Following
Results:
– Device Temperature Grade 2: –40°C to
+105°C Ambient Operating Temperature
– Device HBM ESD Classification Level H2
– Device CDM ESD Classification Level C3B
4.5-V to 28-V Input Range
Output Voltage Range 0.8 V to 90% of Input
Voltage
Output Current Up to 3 A
Fixed Switching Frequency: 600 kHz
Three Selectable Levels of Overcurrent
Protection (Output 2)
0.8-V 1.5% Voltage Reference
2.1-ms Internal Soft Start
Dual PWM Outputs 180° Out-of-Phase
Ratiometric or Sequential Startup Modes
Selectable by a Single Pin
85-mΩ Internal High-Side MOSFETs
Current Mode Control
Internal Compensation
Pulse-by-Pulse Overcurrent Protection
Thermal Shutdown Protection at 148°C
14-Pin PowerPAD™ HTSSOP Package
Power for DSP
Consumer Electronics
CONTENTS
Device Ratings
2
Electrical Characteristics
3
Device Information
9
Application Information
12
Design Examples
32
Additional References
44
DESCRIPTION
The TPS54386-Q1 are dual-output, non-synchronous
buck converters capable of supporting 3-A output
applications that operate from a 4.5-V to 28-V input
supply voltage, and require output voltages between
0.8 V and 90% of the input voltage.
With an internally-determined operating frequency,
soft-start time, and control-loop compensation, these
converters provide many features with a minimum of
external
components.
Channel-1
overcurrent
protection is set at 4.5 A, whereas the channel-2
overcurrent protection level is selected by connecting
a pin to ground, to BP, or left floating. The setting
levels are used to allow for scaling of external
components for applications that do not need the full
load capability of both outputs.
The outputs may be enabled independently, or may
be configured to allow either ratiometric or sequential
start-up sequencing. Additionally, the two outputs
may be powered from different sources.
VIN
TPS54386
1
PVDD1
PVDD2 14
2
BOOT1
BOOT2 13
3
SW1
SW2 12
4
GND
BP 11
5
EN1
6
EN2
7
FB1
OUTPUT1
SEQ 10
ILIM2
9
FB2
8
GND
1
2
3
Please be aware that an important notice concerning availability, standard warranty, and use in critical applications of
Texas Instruments semiconductor products and disclaimers thereto appears at the end of this data sheet.
PowerPAD is a trademark of Texas Instruments.
All other trademarks are the property of their respective owners.
PRODUCTION DATA information is current as of publication date.
Products conform to specifications per the terms of the Texas
Instruments standard warranty. Production processing does not
necessarily include testing of all parameters.
Copyright © 2012, Texas Instruments Incorporated
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
These devices have limited built-in ESD protection. The leads should be shorted together or the device placed in conductive foam
during storage or handling to prevent electrostatic damage to the MOS gates.
Table 1. ORDERING INFORMATION (1)
PART NUMBER
OPERATING FREQUENCY (kHz)
PACKAGE
MEDIA
UNITS
TOP-SIDE MARKING
TPS54386TPWPRQ1
600
14-HTSSOP package
Tape and reel
2000
54386T
(1)
For the most current package and ordering information see the Package Option Addendum at the end of this document, or see the TI
web site at www.ti.com.
DEVICE RATINGS
ABSOLUTE MAXIMUM RATINGS (1)
VALUE
PVDD1, PVDD2, EN1, EN2
Input voltage range
UNIT
30
BOOT1, BOOT2
VSW+ 7
SW1, SW2
–2 to 30
SW1, SW2 transient (< 50 ns)
–3 to 31
BP
V
6.5
SEQ, ILIM2
–0.3 to 6.5
FB1, FB2
–0.3 to 3
SW1, SW2 output current
7
A
BP load current
35
mA
Tstg
Storage temperature
–55 to 165
°C
TA
Operating temperature
–40 to 105
°C
2
kv
750
V
Human Body Model (HBM) AEC-Q100 Classification Level H2
ESD ratings
(1)
Charged Device Model (CDM) AEC-Q100 Classification Level C3B
Permanent device damage may occur if Absolute Maximum Ratings are exceeded. Functional operation should be limited to the
Recommended DC Operating Conditions detailed in this data sheet. Exposure to conditions beyond the operational limits for extended
periods of time may affect device reliability.
RECOMMENDED OPERATING CONDITIONS
MIN
MAX
UNIT
VPVDD2
Input voltage
4.5
28
V
TA
Operating junction
temperature
–40
125
°C
PACKAGE DISSIPATION RATINGS (1)
(1)
(2)
(3)
(4)
2
(2) (3)
PACKAGE
THERMAL IMPEDANCE
JUNTION-TO-THERMAL PAD
(°C/W)
TA = 25°C
POWER RATING (W)
TA = 105°C
POWER RATING (W)
Plastic 14-Pin HTSSOP (PWP)
2.07 (4)
1.6
0.8
For more information on the PWP package, see TI Technical Brief (SLMA002A).
TI device packages are modeled and tested for thermal performance using printed circuit board designs outlined in JEDEC standards
JESD 51-3 and JESD 51-7.
For application information, see the Power Derating section.
TJ-A = 40°C/W.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
ELECTRICAL CHARACTERISTICS
–40°C ≤ TA ≤ 105°C, VPVDD1 = VPVDD2 = 12 V, unless otherwise noted.
PARAMETER
TEST CONDITIONS
MIN
TYP
MAX
UNIT
INPUT SUPPLY (PVDD)
VPVDD1
VPVDD2
Input voltage range
4.5
28
4.5
28
V
μA
IDDSDN
Shutdown
V EN1 = V EN2 = VPVDD2
70
150
IDDQ
Quiescent, non-switching
VFB = 0.9 V, outputs off
1.8
3
IDDSW
Quiescent, while-switching
SW node unloaded; Measured as BP sink
current
VUVLO
Minimum turnon voltage
PVDD2 only
VUVLO(hys)
Hysteresis
tSTART
(1) (2)
CBP = 10 μF, EN1 and EN2 go low
simultaneously
Time from start-up to soft-start begin
mA
5
3.8
4.1
V
4.4
V
400
mV
2
ms
ENABLE (EN)
VEN1
Enable threshold
VEN2
0.9
1.2
1.5
0.9
1.2
1.5
Hysteresis
IEN1
IEN2
tEN (1)
50
Enable pullup current
V EN1 = V EN2 = 0 V
Time from enable to soft-start begin
Other EN pin = GND
V
V
mV
6
12
μA
6
12
μA
μs
10
BP REGULATOR (BP)
BP
BPLDO
IBP
(1)
IBPS
Regulator voltage
8 V < PVDD2 < 28 V
Dropout voltage
PVDD2 = 4.5 V; switching, no external load on
BP
5
5.25
400
Regulator external load
Regulator short circuit
4.5 V < PVDD2 < 28 V
5.6
V
mV
2
mA
10
20
30
mA
510
630
750
kHz
OSCILLATOR
fSW
Switching frequency
tDEAD (1)
Clock dead time
140
ns
ERROR AMPLIFIER (EA) and VOLTAGE REFERENCE (REF)
VFB1
Feedback input voltage
VFB2
IFB1
0°C < TA < 85°C
788
–40°C < TA < 125°C
786
Feedback input bias current
IFB2
gM1 (1)
gM2 (1)
Transconductance
800
812
mV
812
mV
3
50
nA
3
50
nA
30
μS
30
μS
SOFT START (SS)
TSS1
Soft-start time
TSS2
1.5
2.1
2.7
ms
1.5
2.1
2.7
ms
3.6
4.5
5.6
A
3.6
4.5
5.6
OVERCURRENT PROTECTION
ICL1
Current limit channel 1
VILIM2 = VBP
ICL2
Current limit channel 2
VILIM2 = (floating)
VILIM2 = GND
VUV1
Low-level output threshold to declare a fault
VUV2
tHICCUP (1)
tON1(oc) (1)
tON2(oc) (1)
(1)
(2)
Measured at feedback pin
Hiccup timeout
2.4
3
3.6
1.15
1.5
1.75
670
mV
670
mV
10
Minimum overcurrent pulse duration
A
ms
90
150
ns
90
150
ns
Ensured by design. Not production tested.
When both outputs are started simultaneously, a 20-mA current source charges the BP capacitor. Faster times are possible with a lower
BP capacitor value. More information can be found in the Input UVLO and Startup section.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
3
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
ELECTRICAL CHARACTERISTICS (continued)
–40°C ≤ TA ≤ 105°C, VPVDD1 = VPVDD2 = 12 V, unless otherwise noted.
PARAMETER
TEST CONDITIONS
MIN
TYP
MAX
UNIT
BOOTSTRAP
RBOOT1
Bootstrap switch resistance
RBOOT2
From BP to BOOT1 or BP to BOOT2,
IEXT = 50 mA
18
TA = 25°C, VPVDD2 = 8 V
85
–40°C < TA < 125°C, VPVDD2 = 8 V
85
165
100
200
Ω
OUTPUT STAGE (Channel 1 and Channel 2)
rDS(on)
(3)
MOSFET on-resistance plus bond-wire resistance
(3)
Minimum controllable pulse duration
ISWx peak current > 1 A (4)
DMIN
Minimum duty cycle
VFB = 0.9 V
DMAX
Maximum duty cycle
fSW = 600 kHz
ISW
Switching-node leakage current (sourcing)
Outputs OFF
tON(min)
mΩ
ns
0%
85%
90%
2
12
μA
THERMAL SHUTDOWN
TSD
(3)
TSD(hys)
(3)
(4)
4
Shutdown temperature
(3)
Hysteresis
148
°C
20
°C
Ensured by design. Not production tested.
See Figure 14 for ISWx peak current
enable threshold voltage
Active
Tie EN1 to > enable threshold voltage
for low quiescent current (BP inactive)
when VEN2 > enable threshold voltage
Ignored by the device.when V EN1 <
enable threshold voltage
GND
Sequential, output 1 then output 2
Tie EN2 to < enable threshold voltage
for BP to be active when VEN1 >
enable threshold voltage
Active
Tie EN2 to > enable threshold voltage
for low quiescent current (BP inactive)
when V EN1 > enable threshold voltage
(floating)
Independent or ratiometric, output 1
and output 2
Active. EN1 and EN2 must be tied
together for Ratio-metric startup.
Active. EN1 and EN2 must be tied
together for ratiometric start-up.
If the SEQ pin is connected to BP, then when output 2 is enabled, output 1 is allowed to start approximately 400
μs after output 2 has reached regulation; that is, sequential start-up where output 1 is slave to output 2. If EN2 is
allowed to go high after the outputs have been operating, then both outputs are disabled immediately, and the
output voltages decay according to the load that is present.
If the SEQ pin is connected to GND, then when output 1 is enabled, output 2 is allowed to start approximately
400 μs after output 1 has reached regulation; that is, sequential start-up where output 2 is slave to output 1. If
EN1 is allowed to go high after the outputs have been operating, then both outputs are disabled immediately,
and the output voltages decay according to the load that is present.
SEQ = BP
Sequential
CH2 then CH1
SEQ = GND
Sequential
CH1 then CH2
5-V VOUT1
(2 V/div)
5-V VOUT1
(2 V/div)
3.3-V VOUT2
(2 V/div)
3.3-V VOUT2
(2 V/div)
T - Time - 1 ms/div
T - Time - 1 ms/div
Figure 18. SEQ Pin Tied to BP
Figure 19. SEQ Pin Tied to GND
NOTE
An R-C network connected to the ENx pin may be used in addition to the SEQ pin in
sequential mode to delay the start-up of the first output voltage. This approach may be
necessary in systems with a large number of output voltages and elaborate voltagesequencing requirements. See Enable and Timed Turn On of the Outputs.
14
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
If the SEQ pin is left floating, output 1 and output 2 each start ratiometrically when both outputs are enabled at
the same time. Output 1 and output 2 soft-start at a rate that is determined by the respective final output voltages
and enter regulation at the same time. If the EN1 and EN2 pins are allowed to operate independently, then the
two outputs also operate independently.
5-V VOUT1
(2 V/div)
3.3-V VOUT2
(2 V/div)
T - Time - 1 ms/div
Figure 20. SEQ Pin Floating
Soft Start
Each output has a dedicated soft-start circuit. The soft-start voltage is an internal digital reference ramp to one of
two noninverting inputs of the error amplifier. The other input is the (internal) precision 0.8-V reference. The total
ramp time for the FB voltage to charge from 0 V to 0.8 V is about 2.1 ms. During a soft-start interval, the
TPS54386-Q1 output slowly increases the voltage to the noninverting input of the error amplifier. In this way, the
output voltage ramps up slowly until the voltage on the noninverting input to the error amplifier reaches the
internal 0.8-V reference voltage. At that time, the voltage at the noninverting input to the error amplifier remains
at the reference voltage.
NOTE
To avoid a disturbance in the output voltage during the stepping of the digital soft-start, a
minimum output capacitance of 50 μF is recommended. See Feedback Loop and InductorCapacitor (L-C) Filter Selection. Once the filter and compensation components have been
established, laboratory measurements of the physical design should be performed to
confirm converter stability.
During the soft-start interval, pulse-by-pulse current limiting is in effect. If an overcurrent pulse is detected, six
PWM pulses are skipped to allow the inductor current to decay before another PWM pulse is applied. (See the
Output Overload Protection section.) There is no pulse-skipping if a current-limit pulse is not detected.
DESIGN HINT
If the rate of rise of the input voltage (PVDDx) is such that the input voltage is too low to
support the desired regulation voltage by the time soft-start has completed, then the
output UV circuit may trip and cause a hiccup in the output voltage. In this case, use a
timed-delay start-up from the ENx pin to delay the start-up of the output until the PVDDx
voltage has the capability of supporting the desired regulation voltage. See Operating
Near Maximum Duty Cycle and Maximum Output Capacitance for related information.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
15
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
Output Voltage Regulation
Each output has a dedicated feedback loop comprising a voltage-setting divider, an error amplifier, a pulse-width
modulator, and a switching MOSFET. The regulation output voltage is determined by a resistor divider
connecting the output node, the FBx pin, and GND (see Figure 21). Assuming the value of the upper resistor of
the voltage-setting divider is known, the value of the lower divider resistor for a desired output voltage is
calculated by Equation 2.
VREF
R2 = R1´
VOUT - VREF
where
•
VREF is the internal 0.8-V reference voltage.
(2)
TPS5438x
1
PVDD1
PVDD2 14
2
BOOT1
BOOT2 13
3
SW1
SW2 12
4
GND
BP 11
5
EN1
SEQ 10
6
EN2
ILIM2
9
7
FB1
FB2
8
OUTPUT1
R1
R2
UDG-07011
Figure 21. Feedback Network for Channel 1
DESIGN HINT
There is a leakage current of up to 12 μA out of the SW pin when a single output of the
TPS54386-Q1 is disabled. Keeping the series impedance of R1 + R2 less than 50 kΩ
prevents the output from floating above the reference voltage while the controller output is
in the OFF state.
16
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
Feedback Loop and Inductor-Capacitor (L-C) Filter Selection
In the feedback signal path, the output voltage-setting divider is followed by an internal gM-type error amplifier
with a typical transconductance of 30 μs. An internal series-connected R-C circuit from the gM amplifier output to
ground serves as the compensation network for the converter. The signal from the error amplifier output is then
buffered and combined with a slope compensation signal before it is mirrored to be referenced to the SW node.
Here, it is compared with the current feedback signal to create a pulse-width-modulated (PWM) signal to drive
the upper MOSFET switch. A simplified equivalent circuit of the signal control path is depicted in Figure 22.
NOTE
Noise coupling from the SWx node to internal circuitry of BOOTx may impact narrow
pulse-width operation, especially at load currents less than 1 A. See SW Node Ringing for
further information on reducing noise on the SWx node.
TPS5438x
BOOT
ICOMP - ISLOPE
Error Amplifier
0.8 VREF
+
+
FB
PWM to
Switch
x2
ISLOPE
ICOMP
Offset
f(IDRAIN)
RCOMP
CCOMP
SW
11.5 kW
RCOMP
(kW)
CCOMP
(pF)
TPS54383
700
40
TPS54386
700
20
UDG-07012
Figure 22. Feedback-Loop Equivalent Circuit
A more conventional small-signal equivalent block diagram is shown in Figure 23. Here, the full closed-loop
signal path is shown. Because the TPS54386-Q1 contains internal slope-compensation and loop-compensation
components, the external L-C filter must be selected appropriately so that the resulting control loop meets criteria
for stability. This approach differs from an externally-compensated controller, where the L-C filter is generally
selected first, and the compensation network is found afterwards. To find the appropriate L and C filter
combination, the output-to-Vc signal path plots (see the next section) of gain and phase are used along with
other design criteria to aid in finding the combination that best results in a stable feedback loop.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
17
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
VIN
VREF
VC
+
VOUT
+
Modulator
_
_
Filter
Current
Feedback
Network
Compensation
Network
Figure 23. Small-Signal Equivalent Block Diagram
Inductor-Capacitor (L-C) Selection
The following figures plot the TPS54386-Q1 output-to-Vc gain and phase versus frequency for various duty
cycles (10%, 30%, 50%, 70%, 90%) at three (200 mA, 400 mA, 600 mA) peak-to-peak ripple-current levels. The
loop response curve selected to compensate the loop is based on the duty cycle of the application and the ripple
current in the inductor. Once the curve has been selected and the inductor value has been calculated, the output
capacitor is found by calculating the L-C resonant frequency required to compensate the feedback loop. A brief
example follows the curves.
Note that the internal error-amplifier compensation is optimized for output capacitors with an ESR zero frequency
between 20 kHz and 60 kHz. See the following sections for further details.
GAIN AND PHASE
vs
FREQUENCY
GAIN AND PHASE
vs
FREQUENCY
100
270
85
225
100
270
225
55
135
40
90
25
45
Duty Cycle %
Gain Phase
10
30
50
70
90
10
-5
-20
100
1k
60
-45
-90
1M
90
45
20
0
10 k
100 k
f - Frequency - Hz
135
40
0
Figure 24. TPS54386-Q1 at 200-mApp Ripple
Current
18
180
Phase - °
180
Gain - dB
70
Phase - °
Gain - dB
80
Duty Cycle %
Gain Phase
10
30
50
70
90
-20
100
1k
0
-45
10 k
100 k
f - Frequency -Hz
-90
1M
Figure 25. TPS54386-Q1 at 400-mApp Ripple
Current
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
GAIN AND PHASE
vs
FREQUENCY
85
225
70
180
55
135
40
90
45
25
10
-5
-20
100
Duty Cycle %
Gain Phase
10
30
50
70
90
1k
Phase - °
270
Gain - dB
100
0
-45
10 k
100 k
f - Frequency - Hz
-90
1M
Figure 26. TPS54386-Q1 at 600-mApp Ripple Current
Maximum Output Capacitance
With internal pulse-by-pulse current limiting and a fixed soft-start time, there is a maximum output capacitance
which may be used before start-up problems begin to occur. If the output capacitance is large enough so that the
device enters a current-limit protection mode during startup, then there is a possibility that the output will never
reach regulation. Instead, the TPS54386-Q1 simply shuts down and attempts a restart as if the output were
short-circuited to ground. The maximum output capacitance (including bypass capacitance distributed at the
load) is given by Equation 3:
R1
VREF (1 + R2 ) ´ TS
tSS
R1
1
+
)
COUTmax =
I
- VREF (1 + R2 ) (1 R
VREF CLx
2 ´ VIN ´ L
LOAD
(3)
Minimum Output Capacitance
Ensure the value of capacitance selected for closed-loop stability is compatible with the requirements of Soft
Start.
Modifying The Feedback Loop
Within the limits of the internal compensation, there is flexibility in the selection of the inductor and outputcapacitor values. A smaller inductor increases ripple current, and raises the resonant frequency, thereby
incerasing the required amount of output capacitance. A smaller capacitor could also be used, increasing the
resonant frequency, and increasing the overall loop bandwidth—perhaps at the expense of adequate phase
margin.
The internal compensation of the TPS54x8x is designed for capacitors with an ESR zero frequency between 20
kHz and 60 kHz. It is possible, with additional feedback compensation components, to use capacitors with higher
or lower ESR zero frequencies. For either case, the components C1 and R3 (see Figure 30) are added to recompensate the feedback loop for stability. In this configuration, a low frequency pole is followed by a higherfrequency zero. The placement of this pole-zero pair is dependent on the type of output capacitor used and the
desired closed-loop frequency response.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
19
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
TPS5438x
1
PVDD1
PVDD2 14
2
BOOT1
BOOT2 13
3
SW1
SW2 12
4
GND
BP 11
5
EN1
SEQ 10
6
EN2
ILIM2
9
7
FB1
FB2
8
OUTPUT1
C2
R1
C1
R2
R3
UDG-07013
Figure 27. Optional Loop Compensation Components
NOTE
Once the filter and compensation components have been established, laboratory
measurements of the physical design should be performed to confirm converter stability.
Using High-ESR Output Capacitors
If a high-ESR capacitor is used in the output filter, a zero appears in the loop response that could lead to
instability. To compensate, a small R-C series connected network is placed in parallel with the lower voltagesetting divider resistor (see Figure 27). The values of the components are determined such that a pole is placed
at the same frequency as the ESR zero and a new zero is placed at a frequency location conducive to good loop
stability.
The value of the resistor is calculated using a ratio of impedances to match the ratio of ESR zero frequency to
the desired zero frequency.
R3 =
R2
æ æ fZERO(desired)
çç
ç ç fESR(zero)
èè
ö ö
÷ - 1÷
÷ ÷
ø ø
where:
•
•
20
f ESR(zero) is the ESR zero frequency of the output capacitor.
f ZERO(desired) is the desired frequency of the zero added to the feedback. This frequency should be placed
between 20 kHz and 60 kHz to ensure good loop stability.
Submit Documentation Feedback
(4)
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
The value of the capacitor is calculated in Equation 5.
C1 =
1
2p ´ REQ ´ fESR(zero)
where:
•
REQ is an equivalent impedance created by the parallel combination of the voltage-setting divider resistors (R1
and R2) in series with R3.
(5)
REQ = R3 +
1
ææ 1 ö æ 1 öö
çç ÷ + ç
÷÷
è è R1 ø è R2 ø ø
(6)
Using All Ceramic Output Capacitors
With low-ESR ceramic capacitors, there may not be enough phase margin at the crossover frequency. In this
case (see Figure 27), resistor R3 is set equal to 1/2 R2. This lowers the gain by 6 dB, reduces the crossover
frequency, and improves phase margin.
The value of C1 is found by determining the frequency at which to place the low-frequency pole. The minimum
frequency at which to place the pole is 1 kHz. Any lower, and the time constant will be too slow and interfere with
the internal soft-start (see Soft Start). The upper bound for the pole frequency is determined by the operating
frequency of the converter. It is 3 kHz for the TPS54x83, and 6 kHz for the TPS54x86. C1 is then found from
Equation 7. Keep component tolerances in mind when selecting the desired pole frequency.
C1 =
1
2p ´ REQ ´ fPOLE(desired)
where:
•
•
f POLE(desired) is the desired pole frequency between 1 kHz and 3 kHz (TPS54x83) or 1 kHz and 6 kHz
(TPS54x86).
REQ is an equivalent impedance created by the parallel combination of the voltage-setting divider resistors (R1
and R2) in series with R3.
(7)
REQ = R3 +
1
ææ 1 ö æ 1 öö
çç ÷ + ç
÷÷
è è R1 ø è R2 ø ø
(8)
If it is necessary to increase phase margin, place a capacitor in parallel with the upper voltage-setting divider
resistor (C2 in Equation 9).
C2 =
1
R1
´ 1+
2p ´ fC ´ R1
æ (R2 ´ R3 ) ö
çç
÷÷
è (R2 + R3 ) ø
where
•
f C is the unity-gain crossover frequency, (approximately 50 kHz for most designs following these guidelines).
(9)
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
21
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
Example: TPS54386-Q1 Buck Converter Operating at 12-V Input, 3.3-V Output and 400-mA(P-P) Ripple
Current
First, the steady-state duty cycle is calculated. Assuming the rectifier diode has a voltage drop of 0.5 V, the duty
cycle is approximated using Equation 10.
VOUT + VDIODE
3.3 + 0.5
= 30%
d=
=
VIN + VDIODE
12 + 0.5
(10)
The filter inductor is then calculated; see Equation 11.
V - VOUT
12 - 3.3
1
L = IN
´ d ´ TS =
´ 0.3 ´
= 10.9 mH
0.4
600000
DIL
(11)
A custom-designed inductor may be used for the application, or a standard value close to the calculated value
may be used. For this example, a standard 10-μH inductor is used. Using Figure 25, find the 30% duty cycle
curve. The 30% duty cycle curve has a down slope from low frequency and rises at approximately 6 kHz. This
curve is the resonant frequency that must be compensated. Any frequency within an octave of the peak may be
used in calculating the capacitor value. In this example, 6 kHz is used.
1
C=
2
L ´ (2 ´ p ´ fRES )
1
=
10 ´ 10
-6
2
= 70 mF
´ (2 ´ 3.14 ´ 6000 )
(12)
A 68-μF capacitor should be used as a bulk capacitor, with up to 10 μF of ceramic bypass capacitance. To
ensure the ESR zero does not significantly impact the loop response, the ESR of the bulk capacitor should be
placed a decade above the resonant frequency.
RESR <
1
1
=
» 40 mW
2 ´ p ´ 10 ´ fRES ´ C 2 ´ 3.14 ´ 10 ´ 6000 ´ 68 ´ (10 )-6
(13)
The resulting loop gain and phase are shown in Figure 28. Based on measurement, loop crossover is 45 kHz
with a phase margin of 60 degrees.
GAIN AND PHASE
vs
FREQUENCY
80
180
70
Phase
135
60
90
45
40
0
30
20
Phase - °
Gain - dB
50
-45
10
-90
0
Gain
-135
-10
-20
100
1k
10 k
100 k
f - Frequency - Hz
-180
1M
Figure 28. Example Loop Result
22
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
Bootstrap for the N-Channel MOSFET
A bootstrap circuit provides a voltage source higher than the input voltage and of sufficient energy to fully
enhance the switching MOSFET each switching cycle. The PWM duty cycle is limited to a maximum of 90%,
allowing an external bootstrap capacitor to charge through an internal synchronous switch (between BP and
BOOTx) during every cycle. When the PWM switch is commanded to turn ON, the energy used to drive the
MOSFET gate is derived from the voltage on this capacitor.
To allow the bootstrap capacitor to charge each switching cycle, an internal pulldown MOSFET (from SW to
GND) is turned ON for approximately 140 ns at the beginning of each switching cycle. In this way, if, during light
load operation, there is insufficient energy for the SW node to drive to ground naturally, this MOSFET forces the
SW node toward ground and allows the bootstrap capacitor to charge.
Because this is a charge transfer circuit, care must be taken in selecting the value of the bootstrap capacitor. It
must be sized such that the energy stored in the capacitor on a per-cycle basis is greater than the gate charge
requirement of the MOSFET being used.
DESIGN HINT
For the bootstrap capacitor, use a ceramic capacitor with a value between 22 nF and 82
nF.
NOTE
For 5-V input applications, connect PVDDx to BP directly. This connection bypasses the
internal control-circuit regulator and provides maximum voltage to the gate-drive circuitry.
In this configuration, shutdown mode IDDSDN is the same as quiescent IDDQ.
Light Load Operation
There is no special circuitry for pulse skipping at light loads. The normal characteristic of a nonsynchronous
converter is to operate in the discontinuous-conduction mode (DCM) at an average load current less than onehalf of the inductor peak-to-peak ripple current. Note that the amplitude of the ripple current is a function of input
voltage, output voltage, inductor value, and operating frequency, as shown in Equation 14.
1 VIN - VOUT
IDCM = ´
´ d ´ TS
2
L
(14)
Further, during discontinuous-mode operation the commanded pulse duration may become narrower than the
capability of the converter to resolve. To maintain the output voltage within regulation, skipping switching pulses
at light load conditions is a natural byproduct of that mode. This condition may occur if the output capacitor is
charged to a value greater than the output regulation voltage and there is insufficient load to discharge the
capacitor. A byproduct of pulse skipping is an increase in the peak-to-peak output ripple voltage.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
23
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
SW Waveform
SW Waveform
VOUT
Ripple
Inductor
Current
VOUT
Ripple
Skipping
VIN = 12 V
VOUT = 5 V
Inductor
Current
Steady State
VIN = 12 V
VOUT = 5 V
Figure 29. Steady State
Figure 30. Skipping
DESIGN HINT
If additional output capacitance is required to reduce the output-voltage ripple during DCM
operation, be sure to recheck the Feedback Loop and Inductor-Capacitor (L-C) Filter
Selection and Maximum Output Capacitance sections.
SW Node Ringing
A portion of the control circuitry is referenced to the SW node. To ensure jitter-free operation, it is necessary to
decrease the voltage waveform ringing at the SW node to less than 5 volts peak and of a duration of less than
30-ns. In addition to following good printed-circuit board (PCB) layout practices, there are a couple of design
techniques for reducing ringing and noise.
SW Node Snubber
Voltage ringing observable at the SW node is caused by fast switching edges and parasitic inductance and
capacitance. If the ringing results in excessive voltage on the SW node, or erratic operation of the converter, an
R-C snubber may be used to dampen the ringing and ensure proper operation over the full load range.
DESIGN HINT
A series-connected R-C snubber (C = between 330 pF and 1 nF, R = 10 Ω) connected
from SW to GND reduces the ringing on the SW node.
Bootstrap Resistor
A small resistor in series with the bootstrap capacitor reduces the turnon time of the internal MOSFET, thereby
reducing the rising-edge ringing of the SW node.
DESIGN HINT
A resistor with a value between 1 Ω and 3 Ω may be placed in series with the bootstrap
capacitor to reduce ringing on the SW node.
DESIGN HINT
Placeholders for these components should be placed on the initial prototype PCBs in case
they are needed.
24
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
Output Overload Protection
In the event of an overcurrent during soft-start on either output (such as starting into an output short), pulse-bypulse current limiting and PWM frequency division are in effect for that output until the internal soft-start timer
ends. At the end of the soft-start time, a UV condition is declared and a fault is declared. During this fault
condition, both PWM outputs are disabled and the small pulldown MOSFETs (from SWx to GND) are turned ON.
This process ensures that both outputs discharge to GND in the event that overcurrent is on one output while the
other is not loaded. The converter then enters a hiccup-mode time-out before attempting to restart. Frequency
division means if an overcurrent pulse is detected, six clock cycles are skipped before the next PWM pulse is
initiated, effectively dividing the operating frequency by six and preventing excessive current buildup in the
inductor.
In the event of an overcurrent on either output after the output reaches regulation, pulse-by-pulse current limit is
in effect for that output. In addition, an output undervoltage (UV) comparator monitors the FBx voltage (that
follows the output voltage) to declare a fault if the output drops below 85% of regulation. During this fault
condition, both PWM outputs are disabled and the small pulldown MOSFETs (from SWx to GND) are turned ON.
This design ensures that both outputs discharge to GND, in the event that overcurrent is on one output while the
other is not loaded. The converter then enters a hiccup-mode timeout before attempting to restart.
The overcurrent threshold for output 1 is set nominally at 4.5 A. The overcurrent level of output 2 is determined
by the state of the ILIM2 pin. The ILIM setting of output 2 is not latched in place and may be changed during
operation of the converter.
Table 3. Current Limit Threshold Adjustment for
Output 2
ILIM2 Connection
OCP Threshold for Output 2
BP
4.5-A nominal setting
(floating)
3-A nominal setting
GND
1.5-A nominal setting
DESIGN HINT
The OCP threshold refers to the peak current in the internal switch. Be sure to add onehalf of the peak inductor ripple current to the dc load current in determining how close the
actual operating point is to the OCP threshold.
Operating Near Maximum Duty Cycle
If the TPS54386-Q1 operates at maximum duty cycle, and if the input voltage is insufficient to support the output
voltage (at full load or during a load-current transient), then there is a possibility that the output voltage will fall
from regulation and trip the output UV comparator. If this should occur, the TPS54386-Q1 protection circuitry
declares a fault and enters a shut-down-and-restart cycle.
DESIGN HINT
Ensure that under ALL conditions of line and load regulation, there is sufficient duty cycle
to maintain output-voltage regulation.
To calculate the operating duty cycle, use Equation 15.
d=
VOUT + VDIODE
VIN + VDIODE
where
•
VDIODE is the voltage drop of the rectifier diode.
(15)
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
25
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
Dual-Supply Operation
It is possible to operate a TPS54386-Q1 from two supply voltages. If this application is desired, then the
sequencing of the supplies must be such that PVDD2 is above the UVLO voltage before PVDD1 begins to rise.
This level requirement ensures that the internal regulator and the control circuitry are in operation before PVDD1
supplies energy to the output. In addition, output 1 must be held in the disabled state (EN1 high) until there is
sufficient voltage on PVDD1 to support output 1 in regulation. (See the Operating Near Maximum Duty Cycle
section.)
The preferred sequence of events is:
1. PVDD2 rises above the input UVLO voltage.
2. PVDD1 rises with output 1 disabled until PVDD1 rises above the level to support output 1 regulation.
With these two conditions satisfied, there is no restriction on PVDD2 to be greater than or less than PVDD1.
DESIGN HINT
An R-C delay on EN1 may be used to delay the start-up of output 1 for a long-enough
period of time to ensure that PVDD1 can support the output 1 load.
Cascading Supply Operation
It is possible to source PVDD1 from output 2 as depicted in Figure 31 and Figure 32. This configuration may be
preferred if the input voltage is high, relative to the voltage on output 1.
VIN
TPS54383
1
PVDD1
PVDD2 14
2
BOOT1
BOOT2 13
3
SW1
SW2 12
4
GND
BP 11
5
EN1
SEQ 10
6
EN2
ILIM2
9
7
FB1
FB2
8
OUTPUT2
OUTPUT1
UDG-07015
Figure 31. Schematic Showing Cascading PVDD1 From Output 2
26
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
PVDD2
Output2
PVDD1
Output1
T - Time
Figure 32. Waveforms Resulting From Cascading PVDD1 From Output 2
In this configuration, the following conditions must be maintained:
1. Output 2 must be of a voltage high enough to maintain regulation of output 1 under all load conditions.
2. The sum of the current drawn by output 2 load plus the current into PVDD1 must be less than the overload
protection current level of output 2.
3. The method of output sequencing must be such that the voltage on output 2 is sufficient to support output 1
before output 1 is enabled. This requrement may be accomplished by:
(a) a delay of the enable function
(b) selecting sequential sequencing of output 1 starting after output 2 is in regulation
Multiphase Operation
The TPS54386-Q1 is not designed to operate
http://www.power.ti.com for appropriate device selection.
as
a
two-channel
multiphase
converter.
See
Bypass and FIltering
As with any integrated circuit, supply bypassing is important for jitter-free operation. To improve the noise
immunity of the converter, ceramic bypass capacitors must be placed as close to the package as possible.
1. PVDD1 to GND: Use a 10-μF ceramic capacitor.
2. PVDD2 to GND: Use a 10-μF ceramic capacitor.
3. BP to GND: Use a 4.7-μF to 10-μF ceramic capacitor.
Overtemperature Protection and Junction Temperature Rise
The overtemperature thermal protection limits the maximum power to be dissipated at a given operating ambient
temperature. In other words, at a given device power dissipation, the maximum ambient operating temperature is
limited by the maximum allowable junction operating temperature. The device junction temperature is a function
of power dissipation and the thermal impedance from the junction to ambient. If the internal die temperature
should reach the thermal shutdown level, the TPS54386-Q1 shuts off both PWMs and remains in this state until
the die temperature drops below the hysteresis value, at which time the device restarts.
The first step to determine the device junction temperature is to calculate the power dissipation. The power
dissipation is dominated by the two switching MOSFETs and the BP internal regulator. The power dissipated by
each MOSFET is composed of conduction losses and output (switching) losses incurred while driving the
external rectifier diode. To find the conduction loss, first find the rms current through the upper switch MOSFET.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
27
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
2
æ
æ
2 ç (D IOUTPUTx )
ç
IRMS(outputx) = D ´ (IOUTPUTx ) +
çç
ç
12
è
è
öö
÷÷
÷ ÷÷
øø
where
•
•
•
D is the duty cycle.
IOUTPUTx is the dc output current.
ΔIOUTPUTx is the peak ripple current in the inductor for output x.
(16)
Notice the impact of the operating duty cycle on the result.
Multiplying the result by the RDS(on) of the MOSFET gives the conduction loss.
PD(cond) = IRMS(outputx)2 ´ RDS(on)
(17)
The switching loss is approximated by:
2
(VIN) ´ CJ ´ fS
PD(SW) =
2
where
•
•
where CJ is the parallel capacitance of the rectifier diode and snubber (if any).
fS is the switching frequency.
(18)
The total power dissipation is found by summing the power loss for both MOSFETs plus the loss in the internal
regulator.
PD = PD(cond)output1 + PD(SW )output1 + PD(cond)output2 + PD(SW )output2 + VIN ´ Iq
(19)
The temperature rise of the device junction depends on the thermal impedance from the junction to the mounting
pad (see the Package Dissipation Ratings table), plus the thermal impedance from the thermal pad to ambient.
The thermal impedance from the thermal pad to ambient depends on the PCB layout (thermal-pad interface to
the PCB, the exposed pad area) and airflow (if any). See the PCB Layout Guidelines, Additional References
section.
The operating junction temperature is shown in Equation 20.
(
TJ = TA + PD ´ qTH(pkg) + qTH(pad-amb)
)
(20)
Power Derating
The TPS54386-Q1 delivers full current at ambient temperatures up to 85°C if the thermal impedance from the
thermal pad maintains the junction temperature below the thermal shutdown level. At higher ambient
temperatures, the device power dissipation must be reduced to maintain the junction temperature at or below the
thermal shutdown level. Figure 33 illustrates the power derating for elevated ambient temperature under various
airflow conditions. Note that these curves assume that the thermal pad is properly soldered to the recommended
board. (See the References section for further information.)
28
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
POWER DISSIPATION
vs
AMBIENT TEMPERATURE
1.8
LFM = 250
1.6
PD - Power Dissipation - W
LFM = 500
1.4
LFM = 0
1.2
LFM = 150
1.0
0.8
0.6
LFM
0
150
250
500
0.4
0.2
0
0
20
40
60
80
100
120
TA - Ambient Temperature - °C
140
Figure 33. Power-Derating Curves
PowerPAD Package
The PowerPAD package provides low thermal impedance for heat removal from the device. The thermal pad
derives its name and low thermal impedance from the large bonding pad on the bottom of the device. The circuit
board must have an area of solder-tinned-copper underneath the package. The dimensions of this area depend
on the size of the PowerPAD package. Thermal vias connect this area to internal or external copper planes and
should have a drill diameter sufficiently small so that the via hole is effectively plugged when the barrel of the via
is plated with copper. This plug is needed to prevent wicking the solder away from the interface between the
package body and the solder-tinned area under the device during solder reflow. Drill diameters of 0.33 mm (13
mils) work well when 1-oz. copper is plated at the surface of the board while simultaneously plating the barrel of
the via. If the thermal vias are not plugged when the copper plating is performed, then a solder mask material
should be used to cap the vias with a diameter equal to the via diameter of 0.1 mm minimum. This capping
prevents the solder from being wicked through the thermal vias and potentially creating a solder void under the
package. (See the Additional References section.)
PCB Layout Guidelines
The layout guidelines presented here are illustrated in the PCB layout examples given in Figure 34 and
Figure 35.
• The thermal pad must be connected to a low-current (signal) ground plane having a large copper surface
area to dissipate heat. Extend the copper surface well beyond the IC package area to maximize thermal
transfer of heat away from the IC.
• Connect the GND pin to the thermal pad through a 10-mil (0.010-in, or 0.254-mm) wide trace.
• Place the ceramic input capacitors close to PVDD1 and PVDD2; connect using short, wide traces.
• Maintain a tight loop of wide traces from SW1 or SW2 through the switch node, inductor, output capacitor,
and rectifier diode. Avoid using vias in this loop.
• Use a wide ground connection from the input capacitor to the rectifier diode, placed as close to the power
path as possible. Placement directly under the diode and the switch node is recommended.
• Locate the bootstrap capacitor close to the BOOT pin to minimize the gate-drive loop.
• Locate voltage-setting resistors and any feedback components over the ground plane and away from the
switch node and the rectifier diode to the input-capacitor ground connection.
• Locate snubber components (if used) close to the rectifier diode with minimal loop area.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
29
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
•
•
www.ti.com
Locate the BP bypass capacitor very close to the IC; a minimal loop area is recommended.
Locate the output ceramic capacitor close to the inductor output terminal between the inductor and any
electrolytic capacitors, if used.
L2
C18
R8
C14
C17
R6
C13
D2
C19
VOUT2
GND
C15
R7
R4
R9
C16
C8
C12
C11
C6
U1
1
VIN
R2
C10
R5
C1
D1
GND
C7
C3
C4
C9
GND
C5
R3
VOUT1
L1
Figure 34. Top Layer Copper Layout and Component Placement
30
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
Figure 35. Bottom Layer Copper Layout
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
31
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
DESIGN EXAMPLES
Example 1: Detailed Design of a 12-V to 5-V and 3.3-V Converter
The following example illustrates a design process and component selection for a 12-V to 5-V and 3.3-V dual
non-synchronous buck regulator using the TPS54386-Q1 converter. Design Example List of Materials and
Table 5, Definition of Symbols is found at the end of this section.
PARAMETER
NOTES AND CONDITIONS
MIN
NOM
MAX
UNIT
V
INPUT CHARACTERISTICS
VIN
Input voltage
12
13.2
IIN
Input current
VIN = nom, IOUT = max
6.9
1.6
2
A
No load input current
VIN = nom, IOUT = 0 A
12
20
mA
OUTPUT CHARACTERISTICS
VOUT1
Output voltage 1
VIN = nom, IOUT = nom
4.8
5
5.2
VOUT2
Output voltage 2
VIN = nom, IOUT = nom
3.2
3.3
3.4
Line regulation
VIN = min to max
1%
Load regulation
IOUT = min to max
1%
Output voltage ripple
VIN = nom, IOUT = max
50
IOUT1
Output current 1
VIN = min to max
0
2
IOUT2
Output current 2
VIN = min to max
0
2
IOCP1
Output overcurrent channel
1
VIN = nom, VOUT = VOUT1 = 5%
2.4
3
3.5
IOCP2
Output overcurrent channel
2
VIN = nom, VOUT = VOUT2 = 5%
2.4
3
3.5
Transient response ΔVOUT
from load transient
ΔIOUT = 1 A at 3 A/μs
VOUT(ripple
)
Transient response settling
time
V
mVPP
A
200
mV
1
ms
SYSTEM CHARACTERISTICS
fSW
Switching frequency
η
Full-load efficiency
TJ
Operating temperature
range
250
310
370
kHz
60
°C
85%
0
25
+
+
+
Figure 36. Design Example Schematic
32
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
Design Procedure
Duty Cycle Estimation
The first step is to estimate the duty cycle of each switching FET.
VOUT + VFD
VIN(min) + VFD
Dmax »
Dmin »
(21)
VOUT + VFD
VIN(max) + VFD
(22)
Using an assumed forward drop of 0.5 V for a Schottky rectifier diode, the channel 1 duty cycle is approximately
40.1% (minimum) to 48.7% (maximum), while the channel 2 duty cycle is approximately 27.7% (minimum) to
32.2% (maximum).
Inductor Selection
The peak-to-peak ripple is limited to 30% of the maximum output current. This places the peak current far
enough from the minimum overcurrent trip level to ensure reliable operation.
For both channel 1 and channel 2, the maximum inductor ripple current is 600 mA. The inductor size is estimated
in Equation 23.
L min »
VIN(max) - VOUT
ILRIP(max)
´ D min ´
1
fSW
(23)
The inductor values are
• L1 = 18.3 μH
• L2 = 15.3 μH
The next-higher standard inductor value of 22 μH is used for both inductors.
The resulting ripple currents are :
IRIPPLE »
VIN(max) - VOUT
1
´ Dmin ´
L
fSW
(24)
Peak-to-peak ripple currents of 0.498 A and 0.416 A are estimated for channel 1 and channel 2, respectively.
The rms current through an inductor is approximated by Equation 25.
(IL(avg) ) + 121 (IRIPPLE )2
2
IL(rms ) =
(25)
and is approximately 2 A for both channels.
The peak inductor current is found using:
IL(peak ) » IOUT(max) +
1
IRIPPLE
2
(26)
An inductor with a minimum rms current rating of 2 A and minimum saturation current rating of 2.25 A is required.
A Coilcraft MSS1278-223ML 22-μH, 6.8-A inductor is selected.
Rectifier Diode Selection
A Schottky diode is selected as a rectifier diode for its low forward-voltage drop. Allowing 20% over VIN for
ringing on the switch node, the required minimum reverse-breakdown voltage of the rectifier diode is:
V(BR )R(min ) ³ 1.2 ´ VIN
(27)
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
33
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
The diode must have reverse breakdown voltage greater than 15.8 V, therefore a 20-V device is used.
The average current in the rectifier diode is estimated by Equation 28.
ID(avg) » IOUT(max ) ´ (1 - D )
(28)
For this design, 1.2-A (average) and 2.25 A (peak) is estimated for channel 1 and 1.5-A (average) and 2.21-A
(peak) for channel 2.
An MBRS320, 20-V, 3-A diode in an SMC package is selected for both channels. This diode has a forward
voltage drop of 0.4 V at 2 A.
The power dissipation in the diode is estimated by Equation 29.
PD (m ax ) » VFM ´ ID (avg )
(29)
For this design, the full-load power dissipation is estimated to be 480 mW in D1, and 580 mW in D2.
Output Capacitor Selection
The TPS54386-Q1 internal compensation limits the selection of the output capacitors. From , the internal
compensation has a double zero resonance at about 3 kHz. The output capacitor is selected by Equation 30.
COUT =
1
2
2
4 ´ p ´ (fRES ) ´ L
(30)
Solving for COUT using
• fRES = 3 kHz
• L = 22 μH
The resulting is COUT = 128 μF. The output ripple voltage of the converter is composed of the ripple voltage
across the output capacitance and the ripple voltage across the ESR of the output capacitor. To find the
maximum ESR allowable to meet the output ripple requirements, the total ripple is partitioned and the equation
solved to find the ESR.
ESR(max) =
VRIPPLE(tot) - VRIPPLE(cap)
IRIPPLE
=
VRIPPLE(tot)
IRIPPLE
-
D
fS ´ C OUT
(31)
Based on 128 μF of capacitance, 300-kHz switching frequency, and 50-mV ripple voltage, plus rounding up the
ripple current to 0.5 A and the duty cycle to 50%, the capacitive portion of the ripple voltage is 6.5 mV, leaving a
maximum allowable ESR of 87 mΩ.
To meet the ripple-voltage requirements, a low-cost 100-μF electrolytic capacitor with 400 mΩ ESR (C5, C17)
and two 10-μF ceramic capacitors (C3 and C4; and C18 and C19) with 2.5-mΩ ESR are selected. From the data
sheets for the ceramic capacitors, the parallel combination provides an impedance of 28 mΩ at 300 kHz for 14
mV of ripple.
Voltage Setting
The primary feedback divider resistors (R2, R9) from VOUT to FB should be between 10 kΩ and 50 kΩ to
maintain a balance between power dissipation and noise sensitivity. For this design, 20 kΩ is selected.
The lower resistors, R4 and R7 are found using the following equations.
R4 =
R7 =
•
•
34
VFB ´ R2
VOUT1 - VFB
(32)
VFB ´ R9
VOUT2 - VFB
(33)
R2 = R9 = 20 kΩ
VFB = 0.8 V
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
•
•
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
R4 = 3.8 kΩ (3.83 kΩ standard value is used)
R7 = 6.4 kΩ (6.34 kΩ standard value is used)
Compensation Capacitors
Checking the ESR zero of the output capacitors:
fESR(zero) =
•
•
•
1
2 ´ p ´ C ´ ESR
C = 100 μF
ESR = 400 mΩ
ESR(zero) = 3980 Hz
(34)
Because the ESR zero of the main output capacitor is less than 20 kHz, an R-C filter is added in parallel with R4
and R7 to compensate for the ESR of the electrolytic capacitor and add a zero of approximately 40 kHz.
R5 =
•
•
•
•
•
•
R4
æ æ fZERO(desired)
çç
ç ç fESR(zero)
èè
fESR(zero) = 4 kHz
fESR(desired) = 40 kHz
R4 = 3.83 kΩ
R5 = 424 Ω (422 Ω selected)
R7 = 6.34 kΩ
R8 = 702 Ω (698 Ω selected)
REQ = R5 +
•
•
•
C8 =
•
•
ö ö
÷ - 1÷
÷ ÷
ø ø
(35)
1
ææ 1 ö æ 1 öö
çç
÷+ç
÷÷
è è R2 ø è R4 ø ø
R2 = R9 = 20 kΩ
REQ1 = 3.63 kΩ
REQ2 = 5.51 kΩ
(36)
1
2 ´ p ´ REQ ´ fESR(zero)
C8 = 10.9 nF (10 nF selected)
C15 = 7.22 nF (6800 pF selected)
(37)
Input Capacitor Selection
The TPS54386-Q1 data sheet recommends a minimum 10-μF ceramic input capacitor on each PVDD pin. These
capacitors must be capable of handling the rms ripple current of the converter. The rms current in the input
capacitors is estimated by Equation 38.
2
æ
æ
2 ç (D IOUTPUTx )
ç
IRMS(outputx) = D ´ (IOUTPUTx ) +
çç
ç
12
è
è
•
öö
÷÷
÷ ÷÷
øø
(38)
IRMS(CIN) = 0.43 A
One 1210 10-μF, 25-V, X5R ceramic capacitor with 2-mΩ ESR and a 2-A rms current rating is selected for each
PVDD input. Higher-voltage capacitors are selected to minimize capacitance loss at the dc bias voltage to ensure
the capacitors maintain sufficient capacitance at the working voltage.
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
35
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
Bootstrap Capacitor
To ensure proper charging of the high-side FET gate and limit the ripple voltage on the boost capacitor, a 33-nF
bootstrap capacitor is used.
ILIM
Current limit must be set above the peak inductor current IL(peak). Comparing IL(peak) to the available minimum
current limits, ILIM is connected to BP for the highest current-limit level.
SEQ
The SEQ pin is left floating, leaving the enable pins to function independently. If the enable pins are tied
together, the two supplies start up ratiometrically. Alternatively, SEQ could be connected to BP or GND to
provide sequential start-up.
Power Dissipation
The power dissipation in the TPS54386-Q1 is composed of FET conduction losses, switching losses, and
internal regulator losses. The rms FET current is found using Equation 39.
2
æ
æ (D I
2
OUTPUTx )
IRMS(outputx) = D ´ ç (IOUTPUTx ) + ç
çç
ç
12
è
è
öö
÷÷
÷ ÷÷
øø
(39)
This results in 1.05 -A rms for channel 1 and 0.87 A rms for channel 2.
Conduction losses are estimated by:
2
(
PCON = RDS(on ) ´ IQSW (rms )
)
(40)
Conduction losses of 198 mW and 136 mW are estimated for channel 1 and channel 2 respectively.
The switching losses are estimated in Equation 41.
PSW
(V
»
IN(max )
2
) ´ (C
DJ
+ COSS )´ fSW
2
(41)
From the data sheet of the MBRS320, the junction capacitance is 658 pF. Because this is large compared to the
output capacitance of the TPS54x8x, the FET capacitance is neglected, leaving switching losses of 17 mW for
each channel.
The regulator losses are estimated in Equation 42.
(
PREG » IDD ´ VIN(max ) + IBP ´ VIN(max ) - VBP
)
(42)
With no external load on BP (IBP = 0), the power dissipation of the regulator is 66 mW.
Total power dissipation in the device is the sum of conduction and switching for both channels, plus regulator
losses.
The total power dissipation is PDISS = 0.198 + 0.136 + 0.017 + 0.017 + 0.066 = 434 mW.
Design Example Test Results
The following results are from the TPS54386-Q1-001 EVM.
36
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
VIN = 12 V
SW 3.3 V
SW 5 V
t − Time − 40 ns/div
Figure 37. Switching-Node Waveforms
100
100
VIN = 9.6 V
VIN = 9.6 V
90
90
80
VIN = 13.2 V
70
h - Efficiency - %
70
h - Efficiency - %
80
VIN = 12.0 V
60
50
40
30
VOUT = 5.0 V
20
10
VIN = 13.2 V
60
VIN = 12.0 V
50
40
30
VOUT = 3.3 V
VIN (V)
20
VIN (V)
9.6
12.0
13.2
10
9.6
12.0
13.2
0
0
0
0.5
1.0
1.5
2.0
2.5
3.0
0
ILOAD - Load Current - A
Figure 38. 5-V Output Efficiency vs Load Current
0.5
1.0
1.5
2.0
2.5
3.0
ILOAD - Load Current - A
Figure 39. 3.3-V Output Efficiency vs Load Current
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
37
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
1.004
1.004
1.003
VOUT - Output Voltage (Normalized) - V
1.005
VOUT - Output Voltage (Normalized) - V
1.005
1.002
VIN = 9.6 V
1.003
VIN = 9.6 V
1.002
VIN = 12.0 V
1.001
VIN = 13.2 V
1.001
1.000
1.000
0.999
0.999
VOUT = 5.0 V
0.998
VIN = 13.2 V
0.997
0.998
VIN (V)
9.6
12.0
13.2
0.996
0.995
VOUT = 3.3 V
VIN = 12.0 V
0.997
VIN (V)
0.996
9.6
12.0
13.2
0.995
0.5
1.0
1.5
2.0
2.5
0
3.0
0.5
1.0
Gain - dB
Figure 40. 5-V Output Voltage vs Load Current
180
60
135
40
90
20
45
0
0
-20
2.5
3.0
-45
-40
-90
Gain
-80
1k
2.0
Figure 41. 3.3-V Output Voltage vs Load Current
80
-60
1.5
IOUT - Load Current - A
IOUT - Load Current - A
Phase - °
0
Phase
5.0 V
3.3 V
-135
10 k
f - Frequency -Hz
100 k
-180
300 k
Figure 42. Example 1 Loop Response
38
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
Table 4. Design Example List of Materials
QTY
REFERENCE
DESIGNATOR
VALUE
DESCRIPTION
SIZE
PART NUMBER
MANUFACTURER
1
C1
100 μF
Capacitor, Aluminum, 25V, 20%
E-can
EEEFC1E101P
Panasonic
2
C10, C11
10 μF
Capacitor, Ceramic, 25V, X5R 20%
1210
C3216X5R1E106M
TDK
1
C12
4.7 μF
Capacitor, Ceramic, 10V, X5R 20%
0805
Std
Std
2
C14, C16
470 pF
Capacitor, Ceramic, 25V, X7R, 20%
0603
Std
Std
1
C15
6.8 nF
Capacitor, Ceramic, 25V, X7R, 20%
0603
Std
Std
100 μF
Capacitor, Aluminum, 10V, 20%, FC
Series
F-can
EEEFC1A101P
Panasonic
1
C17, C5
4
C3, C4, C18, C19 10 μF
Capacitor, Ceramic, 6.3V, X5R 20%
0805
C2012X5R0J106M
TDK
1
C8
10 nF
Capacitor, Ceramic, 25V, X7R, 20%
0603
Std
Std
2
C9, C13
0.033 μF
Capacitor, Ceramic, 25V, X7R, 20%
0603
Std
Std
2
D1, D2
MBRS320
Diode, Schottky, 3-A, 30-V
SMC
MBRS330T3
On Semi
2
L1, L2
22 μH
Inductor, Power, 6.8A, 0.038 Ω
0.484 x
0.484
MSS1278-153ML
Coilcraft
2
R2, R9
20 kΩ
Resistor, Chip, 1/16W, 1%
0603
Std
Std
1
R5
422 Ω
Resistor, Chip, 1/16W, 1%
0603
Std
Std
2
R6, R10
10 Ω
Resistor, Chip, 1/16W, 5%
0603
Std
Std
1
R8
698 Ω
Resistor, Chip, 1/16W, 1%
0603
Std
Std
1
R4
3.83 kΩ
Resistor, Chip, 1/16W, 1%
0603
Std
Std
1
R7
6.34 kΩ
Resistor, Chip, 1/16W, 1%
0603
Std
Std
TPS54386-Q1 DC-DC Switching
Converter w/ FET
HTSSOP
TPS54386-Q1PWP
-14
1
U1
TI
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
39
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
Table 5. Definition of Symbols
CDJ
Average junction capacitance of the rectifier diode from 0 V to VIN(max)
COSS
Average output capacitance of the switching MOSFET from 0 V to VIN(max)
COUT
Output capacitor
D(max)
Maximum steady-state operating duty cycle
D(min)
Minimum steady-state operating duty cycle
ESR(max)
Maximum allowable output-capacitor ESR
fSW
Switching frequency
IBP
Output current of BP regulator due to external loads
IDD
Switching quiescent current with no load on BP
ID(avg)
Average diode conduction current
ID(peak)
Peak diode conduction current
IIN(avg)
Average input current
IIN(rms)
Root mean squared (RMS) input current
IL(avg)
Average inductor current
IL(rms)
Root mean squared (RMS) inductor current
IL(peak)
Peak current in inductor
ILRIP(max)
Maximum allowable inductor ripple current
L(min)
Minimum inductor value to maintain desired ripple current
IOUT(max)
Maximum designed output current
IRMS(cin)
Root mean squared (RMS) current through the input capacitor
IRIPPLE
Inductor peak-to-peak ripple current
IQSW(rms)
Root mean squared current through the switching MOSFET
PCON
Power loss due to conduction through switching MOSFET
PD(max)
Maximum power dissipation in diode
RDS(on)
Drain-to-source resistance of the switching MOSFET when ON
PSW
Power loss due to switching
PREG
Power loss due to the internal regulator
VBP
Output voltage of BP regulator
V(BR)R(min)
Minimum reverse-breakdown voltage rating for rectifier diode
VFB
Regulated feedback voltage
VFD
Forward voltage drop across rectifier diode
VIN
Power-stage input voltage
VOUT
Regulated output voltage
VRIPPLE(cap)
Peak-to-peak ripple voltage due to ideal capacitor (ESR = 0 Ω)
VRIPPLE(tot)
Maximum allowable peak-to-peak output ripple voltage
40
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
Example 2: 24 V to 12 V and 24 V to 5 V
For a higher input voltage, both a snubber and bootstrap resistors are added to reduce ringing on the switch
node and a 30-V Schottky diode is selected. A higher-resistance feedback network is chosen for the 12-V output
to reduce the feedback current.
+
+
Figure 43. 24 V to 12 V and 24 V to 5 V Using the TPS54386-Q1
VIN = 24 V
IOUT = 2 A
VIN = 24 V
IOUT = 2 A
VOUT
(5 V/div)
VOUT
(5 V/div)
T − Time − 10 ns / div
T − Time − 10 ns / div
Figure 44. Switch Node Ringing Without Snubber
and Boost Resistor
Figure 45. Switch Node Ringing With Snubber and
Boost Resistor
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
41
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
90
80
VOUT = 5 V
h - Efficiency - %
70
60
VOUT = 12 V
50
40
30
VIN = 24 V
20
VOUT (V)
5
12
10
0
0
0.5
1.0
1.5
2.0
2.5
3.0
IOUT - Load Current - A
Figure 46. Efficiency vs Load Current
42
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
TPS54386-Q1
www.ti.com
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
Example 3: 5 V to 3.3 V and 5 V to 1.2 V
For a low-input-voltage application, the TPS54386-Q1 is selected for reduced size, and all ceramic output
capacitors are used. 22-μF input capacitors are selected to reduce input ripple and lead capacitors are placed in
the feedback to boost phase margin.
Figure 47. 5 V to 3.3 V and 5 V to 1.2 V
80
100
80
70
Gain - dB
VOUT = 3.3 V
60
VOUT = 1.2 V
50
40
60
135
40
90
20
45
0
0
-20
-45
-90
30
VIN = 5 V
-40
20
VOUT (V)
-60
1.2
3.3
10
Gain
Phase
WIth Lead
Without Lead
-80
1k
0
0
0.5
1.0
1.5
2.0
2.5
3.0
10 k
f - Frequency -Hz
100 k
Phase - °
90
h - Efficiency - %
180
VOUT = 1.2 V
-135
-180
300 k
IOUT - Load Current - A
Figure 48. Efficiency vs Load Current
Figure 49. Example 3 Loop Response
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
43
TPS54386-Q1
SLUSAZ9A – MARCH 2012 – REVISED MARCH 2012
www.ti.com
ADDITIONAL REFERENCES
Related Devices
The following parts have characteristics similar to the TPS54386-Q1 and may be of interest.
Table 6. Devices Related to the TPS54386-Q1
TI LITERATURE
NUMBER
DEVICE
DESCRIPTION
SLUS642
TPS40222
5-V input, 1.6-A non-synchronous buck converter
SLUS749
TPS54283 /
TPS54286
2-A dual non-synchronous converter with integrated high-side MOSFET
References
These references, design tools, and links to additional references, including design software, may be found at
http:www.power.ti.com
Table 7. References
TI LITERATURE
NUMBER
DESCRIPTION
SLMA002
PowerPAD Thermally Enhanced Package Application Report
SLMA004
PowerPAD™ Made Easy
SLUP206
Under the Hood Of Low Voltage DC/DC Converters. SEM1500 Topic 5, 2002 Seminar Series
SLVA057
Understanding Buck Power Stages in Switchmode Power Supplies
SLUP173
Designing Stable Control Loops. SEM 1400, 2001 Seminar Series
Package Outline and Recommended PCB Footprint
The following pages outline the mechanical dimensions of the 14-Pin PWP package and provide
recommendations for PCB layout.
44
Submit Documentation Feedback
Copyright © 2012, Texas Instruments Incorporated
Product Folder Link(s): TPS54386-Q1
PACKAGE OPTION ADDENDUM
www.ti.com
10-Dec-2020
PACKAGING INFORMATION
Orderable Device
Status
(1)
Package Type Package Pins Package
Drawing
Qty
Eco Plan
(2)
Lead finish/
Ball material
MSL Peak Temp
Op Temp (°C)
Device Marking
(3)
(4/5)
(6)
TPS54386TPWPRQ1
ACTIVE
HTSSOP
PWP
14
2000
RoHS & Green
NIPDAU
Level-3-260C-168 HR
-40 to 105
54386T
(1)
The marketing status values are defined as follows:
ACTIVE: Product device recommended for new designs.
LIFEBUY: TI has announced that the device will be discontinued, and a lifetime-buy period is in effect.
NRND: Not recommended for new designs. Device is in production to support existing customers, but TI does not recommend using this part in a new design.
PREVIEW: Device has been announced but is not in production. Samples may or may not be available.
OBSOLETE: TI has discontinued the production of the device.
(2)
RoHS: TI defines "RoHS" to mean semiconductor products that are compliant with the current EU RoHS requirements for all 10 RoHS substances, including the requirement that RoHS substance
do not exceed 0.1% by weight in homogeneous materials. Where designed to be soldered at high temperatures, "RoHS" products are suitable for use in specified lead-free processes. TI may
reference these types of products as "Pb-Free".
RoHS Exempt: TI defines "RoHS Exempt" to mean products that contain lead but are compliant with EU RoHS pursuant to a specific EU RoHS exemption.
Green: TI defines "Green" to mean the content of Chlorine (Cl) and Bromine (Br) based flame retardants meet JS709B low halogen requirements of