TPS548A28RWWR
Product Specifications
TPS548A28RWWR
Texas Instruments — Power Management & Signal Chain
Product Overview
The TPS548A28RWWR from Texas Instruments is a power management / signal chain IC delivering high efficiency, precision regulation, and robust protection. Texas Instruments (TI) is a global leader in analog ICs, power management solutions, and embedded processors.
Superb Automation partners with leading brands including Qorvo, TI, ADI, XILINX, NVIDIA, Intel, ST, MACOM, Renesas, and Coilcraft to provide complete BOM and supply-chain solutions.
Key Features
- High-efficiency power conversion with low quiescent current
- Wide input voltage range with precision output regulation
- Integrated over-current, over-temperature, and short-circuit protection
- Small footprint packaging for space-constrained designs
- AEC-Q100 qualified options available for automotive applications
- Flexible sequencing and soft-start control
Key Specifications
| Input Voltage | V |
| Output Voltage | V |
| Output Current | A |
| Switching Frequency | kHz |
| Efficiency | % |
| Quiescent Current | µA |
| Topology | — |
| Package | — |
| Operating Temp | °C |
| Protection | — |
Applications
Request Quote — In Stock, 100% New & Original, Full Warranty
TPS548A28RWWR 15 A D-CAP3 buck with a 3 V LDO and remote sense
The TPS548A28RWWR is a 2.7 V to 16 V, 15 A synchronous buck in a 4 mm by 3 mm, 21 pin QFN. The control loop is D-CAP3, an adaptive on time scheme that needs no external compensation at all, so the parts around it are the inductor, the output bank, a boot capacitor, a soft start capacitor and one resistor that sets the current limit. On a board where the power stage is the easy part and the schedule is not, that count is the argument.
The internal LDO is 3 V, which is what lets the part run from a 3 V rail without an external bias. Below that, from 2.7 V to 3 V, an external bias of 3.13 V to 5.3 V on VCC keeps the gate drive alive, and the full 15 A is available across the whole 2.7 V to 16 V window with that bias present. Without any bias the part is good for 15 A from 4 V and 12 A from 3 V.
The reference is 0.6 V at 1 per cent over 40 to 125 degrees junction, and the sense is a true differential pair, so the voltage at the load is the voltage the loop regulates. Current limit is set by a resistor on TRIP and the switching frequency is pin selectable at 600 kHz, 800 kHz or 1 MHz. Fault handling is hiccup on overcurrent and undervoltage, latch off on overvoltage.
Key parameters from the SNVSBB1D document
| Input range | 4 V to 16 V for 15 A without bias, 3 V to 16 V for 12 A without bias, 2.7 V to 16 V for 15 A with 3.13 V to 5.3 V external bias |
| Output range | 0.6 V to 5.5 V, set by the feedback divider against a 0.6 V reference |
| Output current | 15 A continuous, 12 A in the 3 V to 4 V input window without external bias |
| Power FETs | 10.2 milliohm high side and 3.1 milliohm low side |
| Switching frequency | Pin selectable 600 kHz, 800 kHz or 1 MHz, no external timing component |
| Reference accuracy | 0.6 V at 1 per cent over 40 to 125 degrees junction, with differential remote sense |
| Current limit | Programmable with a resistor on TRIP, temperature compensated, hiccup on a sustained fault |
| Package | 21 pin QFN, 4 mm by 3 mm, pin compatible with the 12 A TPS54JA20 |
Why the 3 V LDO matters in a 5 V system
A 5 V bus with plus or minus 10 per cent tolerance dips to 4.5 V, and a converter that needs a 4.75 V bias for its gate drive is already outside its window at that point. With a 3 V LDO the part keeps switching down to 3 V input, so a 5 V rail brownout does not take the core rail with it. In a telecom card that holds up on a 48 V to 12 V brick and then a 12 V to 5 V stage, that margin is the difference between a clean restart and a reset loop.
The trade is dissipation. A 3 V LDO dropping from 12 V at the gate drive current is a small number, but the LDO is inside the package and it counts against the thermal budget at 15 A. Boards that run 12 V input at full load should consider the external bias option, which is the same pin used at low input.
Fault behaviour and the TRIP resistor
The TRIP resistor sets a current limit that is temperature compensated, so the limit does not drift with the die temperature the way a plain RDS(on) sense does. A sustained overcurrent puts the part into hiccup: it stops, waits, retries, and keeps the average dissipation low while the fault persists. An overvoltage latches off, which is the right choice for a rail feeding a processor that must not see a sustained high voltage. Undervoltage is hiccup, so a shorted rail retries instead of staying dead.
Auto skip Eco mode is enabled by the MODE pin for light load efficiency, and forced continuous conduction is the alternative when output ripple at no load has to stay small. The two settings change the no load consumption by an order of magnitude, which is the number a standby budget is built on.
Where this part fits
Space constrained data centre boards, point of load rails for processors and FPGAs, and any 12 V or 5 V intermediate bus that needs 15 A in a 4 by 3 mm footprint. Because it is pin compatible with the 12 A TPS54JA20, a board can move between the two ratings without a revision, which is a useful hedge while a program is still deciding its load. Our desk keeps the cross reference for this class of part, see power management cross reference and one stop BOM sourcing.
What to check before a purchase order
The RWW suffix is the 4 by 3 mm 21 pin QFN; the family also ships in other bodies that do not share the pad. Confirm which LDO version the board was routed for, because the A28 carries 3 V and the A29 carries 4.5 V, and the two differ in the external bias window even though the pinout matches. Then confirm the TRIP resistor value against the intended limit, since a board built with the 12 A part's resistor will trip early at 15 A.
See pin to pin cross reference for the 12 A to 15 A move, and incoming inspection plan for how we check the marking on a 21 pin QFN that carries little more than a code.
Questions that come up on this part
Is external compensation ever needed, is the usual question. With D-CAP3 and all ceramic output capacitors, no. The loop is stable across the recommended output capacitance range, and the document gives the range. The second question is whether the part can start into a prebiased output. It can, and it holds the low side off until the internal ramp passes the sensed voltage, which matters on rails that come up after their load is already partly powered.