TPS54626PWPR
Product Specifications
TPS54626PWPR
Texas Instruments — Power Management & Signal Chain
Product Overview
The TPS54626PWPR 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
TPS54626PWPR: a 4.5 V to 18 V, 6.5 A synchronous buck with Eco-mode in the 14 pin HTSSOP
TPS54626PWPR is a 6.5 A synchronous step down converter with D-CAP2 control in a HTSSOP (PWP) body of 14 pins. The PWP body is the 14 pin HTSSOP with an exposed pad, 2000 pieces on the large reel. The 6.5 A rating and the 36 mOhm high side and 28 mOhm low side FETs are the reason the package has a pad at all: at full load the conduction loss alone is about 0.6 W, and the pad is the only path out of the body.
Order code and package
| TPS54626PWPR | HTSSOP (PWP), 14 pins with exposed pad, 2000 piece large reel, NIPDAU finish, MSL level 2 at 260 C with a one year floor life, -40 to 85 C |
| TPS54626PWPT | the small reel string in the same body |
| Reference | 0.765 V internal reference, which is the value the soft start equation uses and the one the feedback divider is calculated against |
| Control | D-CAP2 adaptive on time with no external compensation and an internal ramp that stands in for output ripple |
| Eco-mode | auto-skip at light load for efficiency, with the transition point set by the inductor and the load |
| Document | SLVSC34, August 2013, single revision |
Electrical parameters that decide the application
| Input range | 4.5 V to 18 V operating, which covers a 12 V bus with a 20 percent transient |
| Output range | 0.76 V to 5.5 V set by the divider, with the reference at 0.765 V |
| Output current | 6.5 A continuous, with the 36 mOhm high side and 28 mOhm low side FETs optimised for low duty cycle conversions |
| Current limit | 7.2 A minimum, 8.2 A typical and 9.5 A maximum measured with a 1.5 uH inductor, which is a valley limit and sits above the 6.5 A rating |
| Switching frequency | 650 kHz fixed, so the inductor and the ripple follow from the part |
| Quiescent current | 3.6 uA typical and 10 uA maximum in shutdown, with the Eco-mode skip behaviour giving a low no load input current |
| Lockout | 3.45 V minimum, 3.75 V typical and 4.05 V maximum on the internal regulator, which is 450 mV under the operating floor |
| Soft start | an external capacitor on SS sets the ramp, with the time following the 0.765 V reference and the 1.1 factor in the document equation |
| Power good | open drain with a 25 kOhm to 150 kOhm pull up window, deglitched and referenced to the internal soft start ramp by a factor of 1.7 |
| Protection | cycle by cycle current limit, output overvoltage and undervoltage comparators, and thermal shutdown |
| Thermal | the exposed pad is the only thermal path, and the document's numbers assume it is soldered to copper with vias |
Family variants and how they differ
| TPS54620 and TPS54622 | the 6 A members of the same converter family with a different reference and a different limit window, in the same pin out |
| Against TPS54821 | that part is 8 A with a split rail and external compensation; this one is 6.5 A with D-CAP2 and no compensation network |
| Eco-mode against forced PWM | Eco-mode skips cycles at light load for efficiency and lets the ripple rise; a forced PWM part holds the frequency and pays the idle loss |
| Fixed frequency | there is no synchronisation pin, so several of these on one board will beat against each other |
| Output range | 0.76 V to 5.5 V covers core rails and 5 V rails from the same part, which is why it appears on both sides of a set top box |
Points to check before this part goes on a BOM
The current limit is a valley limit at 8.2 A typical on a 6.5 A part, and the gap between the two figures is the ripple. With a 1.5 uH inductor at 650 kHz and a 12 V input, the peak to peak ripple at 5 V out is roughly 1.4 A, so the DC current at which the valley reaches the limit is about 7.5 A. A 6.5 A load with that ripple runs its valley near 5.8 A, which is the margin the part is meant to have rather than an accident.
D-CAP2 with no compensation network is the attraction and the constraint. The internal ramp is generated from the input and the switch node rather than from the output ripple, so a low ESR ceramic output capacitor works without a compensation network. The constraint is that the output capacitor value and ESR window is what sets the loop, so a board that changes the output capacitance by a large factor is outside the document's recommended range even though nothing in the parts list says so.
Eco-mode is what keeps the no load input current low, and the transition point is set by the inductor value. A larger inductor moves the transition to a higher load, which improves light load efficiency but slows the transient. A board that runs a 6.5 A part at 100 mA most of the time should choose the inductor for the skip threshold, not for the full load ripple.
The 3.6 uA shutdown current is small enough that the part can be left connected to a battery, and the 650 kHz fixed frequency means the input filter can be sized once. The price is that two of these on one board cannot be phase shifted, and their beat products will land wherever the two 650 kHz clocks happen to sit. On a board with a radio, that is the difference between a clean spectrum and a spurious response.
Why the FET resistances are uneven
36 mOhm on the high side and 28 mOhm on the low side is a deliberate ratio for a converter that spends most of its life at a low duty cycle, converting 12 V to something near 1 V. The high side conducts for about 8 percent of the period and the low side for the rest, so the loss budget is weighted toward the low side switch.
The same part converting 12 V to 5 V runs at a 42 percent duty and the balance shifts, which is why the document's efficiency curves are drawn separately for 1.2 V, 1.8 V, 3.3 V and 5 V outputs rather than as one figure.
What the 6.5 A rating assumes
The rating is at 85 C ambient with the pad soldered to the document's copper area. At 0.6 W of conduction loss plus switching, the junction rise is around 30 C on that board, which leaves margin at 85 C and none at 100 C. In an enclosure without airflow, the derating starts well before 6.5 A.
Where this order code fits
It runs the 1 V to 5 V rails on set top boxes, digital televisions, Blu-ray players and network equipment, where 6.5 A comes from a 12 V bus with no compensation network. Teams comparing it with a neighbouring part start from our Power Management Cross Reference page, and teams replacing a part that has gone end of life start from the Pin to Pin Cross Reference page.
Questions buyers ask
Does this part need a compensation network? No. D-CAP2 closes the loop internally, so the output capacitor value and ESR set the loop behaviour and no COMP components are fitted.
What is the current limit on a 6.5 A part? 8.2 A typical, with a 7.2 A minimum and a 9.5 A maximum. It is a valley limit, so it sits above the DC load current by half the ripple.
Can the output be set to 5 V? Yes. The range is 0.76 V to 5.5 V, set by the feedback divider against a 0.765 V reference.
Is there a synchronisation pin? No. The switching frequency is fixed at 650 kHz, so multiple rails on one board cannot be phase shifted.
SUPERB holds and sources this line for production and repair demand, quoted against the exact ordering string with date codes, packing and country of origin stated on the offer. When a part is moving, the How to Verify a Drop in Replacement page sets out the checks that keep a substitute from failing on the bench.