Contact Us

UCC25630-3DDBR

Brand: Texas Instruments
Texas Instruments UCC25630-3DDBR is a power management component designed for AI Servers & Low-Altitude Economy (Drone and UAV) and related applications. In stock at Superb Automation — request a quote for competitive pricing and fast delivery.
quote now

Product Specifications

UCC25630-3DDBR — Power Management & Signal Chain | Superb Automation

UCC25630-3DDBR

Texas Instruments — Power Management & Signal Chain

Product Overview

The UCC25630-3DDBR 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 VoltageV
Output VoltageV
Output CurrentA
Switching FrequencykHz
Efficiency%
Quiescent CurrentµA
Topology—
Package—
Operating Temp°C
Protection—

Applications

Industrial Automation Automotive Electronics Telecom Infrastructure Server & Data Center Power Medical Devices Consumer Electronics

Request Quote — In Stock, 100% New & Original, Full Warranty

UCC25630-3DDBR: the LLC controller that holds 75 mW of standby with the PFC running

UCC25630-3DDBR is the LLC resonant controller from the UCC25630x family that carries an integrated high side gate driver, hybrid hysteretic control and a burst mode built for standby power. It is meant to sit behind a power factor correction stage and to keep the whole supply under 75 mW at no load with the PFC still switching, which is the requirement that decides whether an adapter meets the CoC Tier II standby limit without a separate standby converter.

The string on our page is the DDB carrier, which the vendor addendum prints as a 14-pin SOIC body on a 2500 piece large reel. The same device is published on a 250 piece small reel as the DDBT tail. Both rows carry the same marking, UCC256303, so the reel label is the only thing that tells them apart once the parts are out of the box.

One line on the addendum changes how this part is bought rather than how it works: the status column prints NRND for both reel rows, the vendor term for a device that is not recommended for new designs. A build that already has this controller on a board keeps buying it, and a build that is starting now should be asking whether the same function is available on a device the vendor still recommends. The numbers below come from SLUSD49A, revised January 2019, and from the addendum of that document.

What the order code sits on

FunctionLLC resonant half bridge controller with integrated high side driver
Control methodhybrid hysteretic control, first order open loop transfer function
Operating frequency35 kHz to 1 MHz
Gate drive0.6 A source and 1.2 A sink capability on the high and low side drivers
Standby75 mW system standby with the PFC stage running
Burst modeprogrammable burst power level that adapts to input voltage
Dead timeadaptive, set by the resonant tank behaviour rather than by a fixed resistor
Soft startinternal, written to avoid hard switching during the ramp
Protectionover temperature, output over voltage, input over and under voltage, three levels of over current
Biasexternal bias required, no internal high voltage regulator
PackageSOIC, 14 pins, 9.9 mm by 3.9 mm, with high voltage pin spacing
Reel2500 pieces on a large reel, 250 on the small reel
Moisture levelLevel-2-260C-1 YEAR
Operating temperatureminus 40 to plus 125 degrees
Vendor statusNRND on both reel rows of the addendum

Why hybrid hysteretic control changes the compensation work

An LLC converter has a resonant tank whose gain falls as frequency rises above resonance and rises again below it, and the usual control problem is that the loop behaves differently on either side of the peak. Hybrid hysteretic control keeps the control effort close to linearly proportional to the average input current within one cycle, which makes the open loop transfer function a first order system. A first order system is compensated with one pole and one zero, and the data sheet says as much: the loop is stable with proper compensation rather than needing a model of the tank.

That property saves bench time on a new build, but it does not remove the tank from the schedule. The resonant inductor and capacitor still set the gain curve, the turns ratio still sets the operating point, and the dead time still has to be long enough for the primary current to commutate at the lightest load the supply will run. Adaptive dead time handles the last item by watching the switch node rather than by trusting a fixed delay, and the practical effect is that a supply built with a 20 percent tolerance on the resonant inductor still commutates cleanly at no load.

The burst mode is where the standby number comes from. In burst the controller stops switching for a period, lets the output fall by a set amount, and then restarts with a burst whose power level is programmed and which follows the input voltage. Because the burst power is consistent from one burst to the next, the output ripple during standby is predictable, and because the controller can leave burst quickly, a load step out of standby does not produce the long recovery that a simple hiccup mode gives. Fast exit from burst is the feature that keeps a set top box or a monitor from dropping its rail when a remote control wakes it.

Driving the half bridge from a 14-pin body

The high side driver is inside the device, which is why the pin count is 14 rather than the 16 that the feature bullet prints. The high side gate and the switch node come out on adjacent pins, the bootstrap capacitor sits between them, and the spacing between those pins and the low side pins is what the body size buys. A 9.9 mm by 3.9 mm SOIC with high voltage spacing is a part that can be used on a supply whose input rail sits at 400 V after the PFC stage, provided the creepage on the board matches the spacing on the package.

The 0.6 A source and 1.2 A sink figures are modest by the standards of a discrete driver, and they set the gate charge the part can drive at a given frequency. At 200 kHz a 20 nanocoulomb gate is comfortable; at 800 kHz the same gate needs more average drive current than the sink rating allows and the switching loss climbs. A build that wants to run at the top of the 1 MHz range is looking at a small die or at an external driver, and the data sheet curves show where the internal driver stops being the right answer.

Bias is external, and that is a real constraint rather than a footnote. The controller needs a low voltage rail of its own before it can start, so the supply needs a startup circuit that charges the bias capacitor from the high voltage rail. Once the converter is running the bias can come from an auxiliary winding, but the startup path is what decides how fast the output comes up and whether the supply can restart after a fault without a hiccup on the input.

Three levels of over current and what each one protects

The over current protection is split into three thresholds that act on different time scales. The fastest one compares the switch current against a limit within a cycle and terminates the pulse; the middle one counts repeated fast events and increases the frequency to reduce the gain; the slowest one integrates over many cycles and shuts the supply down. The reason for three levels is that a hard short, an overload and a slowly rising load are three different events, and a single threshold either trips on inrush or fails to protect the switches.

For a customer building a charger or an adapter, the practical question is what the supply does after a short is removed. A latched shutdown needs a power cycle, a hiccup restart tries again and again, and the choice between them is set by how the protection pins are wired. The data sheet gives the thresholds and the restart behaviour; what it cannot give is the answer for a given product, because that depends on whether the load is allowed to keep the supply off until a user acts.

Reel, moisture and the NRND status

UCC25630-3DDBRSOIC DDB, 14 pins, 2500 pieces, large reel, NRND
UCC25630-3DDBTsame body, 250 pieces, small reel, NRND
Lead finishNIPDAU, nickel palladium gold
Moisture levelLevel-2-260C-1 YEAR
Operating temperatureminus 40 to plus 125 degrees
Part markingUCC256303
DocumentSLUSD49A, revised January 2019

The NRND status does not change the part that arrives on the reel, and it does not change the electrical behaviour. What it changes is the buying conversation. A customer with a released board asks for the part by number and the answer is a reel from stock or from a distributor's shelf; a customer starting a build asks whether the controller has a successor, and the honest answer is that the vendor still produces this device while marking it not recommended for new designs, so the schedule risk belongs to the buyer rather than to the silicon.

For a production line the moisture level 2 with a one year floor life is the line that decides the storage rule. A reel opened and left on a shelf for more than a year needs a bake before reflow, and a baked reel needs a new seal date on the label or the next shift will bake it again. The small reel tail exists for exactly the case where a line needs a few hundred parts for a pilot run and does not want to open a 2500 piece reel that will then age on the shelf.

Where this controller still fits

The applications the vendor lists are digital television supplies, AC to DC adapters, gaming consoles and desktop computers, and the common requirement is a supply that has to hold a very low standby power while the main converter stays ready. An adapter that must meet a 75 mW standby limit with the PFC running is the case this device was written for, because the alternative is a second converter that runs only in standby and that costs more board area than the controller saves.

When a customer brings a supply with this controller on it and asks for a replacement, the comparison starts from the resonant tank rather than from the part number. The operating frequency range, the gate drive current, the bias arrangement and the protection behaviour all have to line up before a sample is worth ordering, and the workflow on our power management cross reference page lists the fields to compare. If the original string is already end of life at the vendor, the component obsolescence management page covers how a last time buy is sized and what a bridge buy looks like.

Questions buyers ask before a first order

Is this the same device as the UCC25630-1 and UCC25630-2? They are the same family with different protection and burst settings, and the suffix matters. The three digit tail is not a packaging code, so a bill of materials that lists one of the siblings cannot be filled with this string without checking the settings.

Does it need a separate bias supply? Yes. There is no internal high voltage regulator, so the board needs a startup path to charge the bias rail before the converter runs.

Can it run at 1 MHz? The oscillator covers up to 1 MHz, but the internal gate driver and the switching loss set a practical ceiling well below that for a large die, and the curves in the data sheet are the place to check a specific gate charge.

What does the reel hold? 2500 pieces on the large reel and 250 on the small reel, both at moisture level 2 with a one year floor life.