UCC27517ADBVR
Product Specifications
UCC27517ADBVR
Texas Instruments — Power Management & Signal Chain
Product Overview
The UCC27517ADBVR 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
UCC27517ADBVR: a single channel low side gate driver with 4 A of symmetrical source and sink current, 13 ns of typical propagation delay, a 4.5 V to 18 V bias range and minus 5 V of input tolerance, in the five pin SOT-23 on a 3000 piece reel
UCC27517ADBVR is a buffer between a controller output and a power FET gate, with two inputs so the same part can invert or not invert, and an output stage that holds the gate low whenever the input floats in a SOT-23 body of 5 pins. The DBV body is 2.90 mm by 1.60 mm with five pins, and the part replaces a totem pole pair of bipolar transistors without their base current or their saturation drop.
Order code and package
| Catalogue string | UCC27517ADBVR, where A is the revision of the UCC27517, DBV is the five pin SOT-23 and R is the 3000 piece large tape and reel |
| Body | SOT-23, DBV, five pins, 2.90 mm by 1.60 mm, with the gate loop on one side and the bias on the other |
| Reel | 3000 pieces on a large tape and reel, and 250 pieces on a small reel as UCC27517ADBVT |
| Material rows | the plain row, the .A row and the two G4 rows are all Active at 3000 pieces, so the packing does not depend on the material declaration |
| Finish and moisture | NIPDAU or SN plating depending on the row, moisture level 1 at 260 C with no floor life limit |
| Temperature | the addendum rows print minus 40 to 125 C while the electrical table is specified to 140 C junction, and that gap has to be read before a high ambient application is signed off |
| Marking | 517A on the body |
| Document | SLUSBQ0C, August 2013, revised August 2015 |
| Status | Active and in production on the addendum dated 2026 |
Electrical parameters that decide the application
| Bias range | 4.5 V to 18 V recommended, with 20 V as the absolute maximum on VDD and no separate logic supply needed |
| Lockout | rising at 4.20 V typical, 3.91 V minimum and 4.5 V maximum, falling at 3.90 V typical with 0.3 V of hysteresis, which is what lets a 5 V rail drive a GaN device directly |
| Startup current | 25 uA typical and 75 uA maximum before the lockout releases, rising to 40 uA and 100 uA in the other input states |
| Peak current | plus or minus 4 A into a 0.22 uF load at 1 kHz, quoted as a bench measurement rather than as a guaranteed production limit |
| Input thresholds | 2.2 V typical high and 1.0 V typical low, with 1.0 V of hysteresis and thresholds that do not follow the bias rail |
| Propagation delay | 13 ns typical, 4 ns minimum and 23 ns maximum at 12 V, and 13 ns typical with 26 ns maximum at 4.5 V |
| Rise and fall | 8 ns typical and 12 ns maximum rise at 12 V into 1.8 nF, 7 ns typical and 11 ns maximum fall, with 16 ns and 22 ns rise at 4.5 V |
| Output resistance | 5.0 to 7.5 Ohm pull-up and 0.5 to 1.0 Ohm pull-down at 12 V, and the hybrid pull-up behaves like about 1.4 times the pull-down resistance in a transient |
| Output swing | 50 to 90 mV of drop at 10 mA on a 12 V rail and 60 to 130 mV at 4.5 V, which is what rail to rail means in practice |
| Reverse current | the output withstands 500 mA of reverse current without damage or logic upset, so an external Schottky clamp is usually not needed |
| Negative input | minus 5 V on the inputs, which covers the gate spike that couples back through the Miller capacitance when the drain swings |
| Quiescent | the device draws microamps when the output is low and the input is static, so a controller that parks the gate does not pay a bias penalty |
What the five pins do and how the inputs combine
| VDD, pin 1 | bias input, 4.5 V to 18 V, decoupled with 1 uF close to the pin because the peak current is drawn in a few nanoseconds |
| GND, pin 2 | the return for both the bias and the gate loop, and the reference for the input thresholds |
| IN+, pin 3 | the non-inverting input; high on this pin with IN- low turns the output on |
| IN-, pin 4 | the inverting input; high on this pin turns the output off regardless of IN+ |
| OUT, pin 5 | the gate output, with 4 A of source and sink and a body diode that clamps the gate to the rails |
| Logic table | IN+ low and IN- low gives low, IN+ low and IN- high gives low, IN+ high and IN- low gives high, both high gives low, and either pin floating gives low |
Points to check before this part goes on a BOM
Two inputs rather than one is what makes this part usable in both polarities. Driving IN+ with the control signal and tying IN- to ground gives a non-inverting driver; tying IN+ to VDD and driving IN- gives an inverting one. The fourth combination, both high, forces the output low, which is a useful way to hold a gate off from a fault signal without an extra gate. What it is not is a logic gate: the two inputs are not a Schmitt trigger with a defined precedence beyond the table, and a board that relies on a race between them is relying on a behaviour the vendor did not specify.
The minus 5 V rating on the inputs is a consequence of the gate loop rather than of the input signal. When the FET turns off and the drain rises, the Miller capacitance couples a spike back into the gate node, and a driver whose input sits on the same node sees it. The input ESD structure on this part is two diodes back to back, which is why an input can go below ground without conducting into the bias rail. An engineer who adds an external clamp diode from the input to ground removes that tolerance and should stop at 0.3 V instead.
The 4.2 V lockout is what makes the part interesting for a GaN device. A GaN FET wants 5 V of gate drive and cannot take more than 6 V, so a driver that runs from 5 V and holds its output low until the rail is at 4.2 V is exactly the protection the device needs. The same part on a 12 V rail drives a silicon MOSFET, and the 20 V absolute maximum leaves room for a 15 V rail with a bootstrap on top. What the part lacks is a separate logic rail, so the input thresholds are fixed at 2.2 V and a controller that only swings to 1.8 V cannot drive it directly.
The 4 A peak current is a pulse rating and the average current is set by the gate charge. Driving 40 nC at 12 V and 200 kHz is 96 mW of gate charge, which is the entire dissipation the five pin body can take without a copper pour; the peak current only decides how fast the transition is. An engineer who picks the driver on peak current alone and runs it at 500 kHz with a 200 nC gate is dissipating 1.2 W in a body with no exposed pad, and the part fails at the junction rather than at the current.
The propagation delay of 13 ns with a 4 ns minimum and 23 ns maximum is what makes a synchronous converter work at high frequency. The dead time has to cover the delay plus the FET turn-off, and 23 ns of part to part variation is a real fraction of a 300 ns period at 3 MHz. The delay is also nearly flat over temperature and bias, which the vendor shows in the typical curves, and that flatness is what allows a fixed dead time instead of an adaptive one.
The temperature rows disagree and the disagreement matters. The electrical table is specified over minus 40 to 140 C junction, and the addendum on the same document prints minus 40 to 125 C ambient for every ordering row. In a 105 C ambient cabinet the part is inside the addendum limit and outside nothing, but a board that assumes 140 C of ambient margin is reading the junction figure as an ambient one. The check is to convert dissipation to junction temperature with the package thermal resistance and compare that with 140 C, not to compare the ambient with the table header.
What a 4 A driver does to the gate resistor
A 4 A driver into a 10 Ohm gate resistor produces 40 V of drop, which is four times the gate voltage, so the resistor is no longer a resistor in the first 20 ns and the current is set by the driver output resistance and the gate capacitance instead. The usual value for a part like this is 2.2 Ohm to 4.7 Ohm, which limits the peak to about 1 A and slows the edge to 30 ns, and that is a deliberate trade of switching loss for ringing.
A board that wants the full 4 A puts no resistor in the gate at all and relies on the loop inductance to limit the di/dt, which works when the driver sits within a few millimetres of the gate. Adding a resistor to a board that already has a short loop is the usual way to cure an oscillation that the loop length did not cause, and the cure then costs 5 ns of dead time margin.
Where the A suffix comes from
The UCC27517A sits in the UCC2751x family next to the UCC27516 and the UCC27517 without the A, and the family table shows them differing in input logic rather than in drive strength. The A part has a TTL and CMOS threshold that does not follow the bias rail; the non-A part has a CMOS threshold that scales with VDD. On a 12 V rail that difference is a threshold of 2.2 V against 6 V, and a 3.3 V controller drives one of them and not the other.
The suffix is therefore a logic fact and not a packing fact, and substituting one for the other on a bill of materials can leave a board that works at 12 V and fails at 5 V, or one that works with a 3.3 V controller and not with a 5 V one.
Where this order code fits
Gate drive for silicon MOSFETs and GaN devices in switched mode supplies, DC to DC converters, motor drives, solar inverters, UPS stages and anywhere a controller output has to become a 4 A gate pulse 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 Component Lead Time Solutions page.
Questions buyers ask
Can it drive a GaN FET? Yes. The 4.2 V lockout and the 18 V ceiling suit a 5 V gate, and the separate input thresholds do not depend on the bias rail.
Does it invert? Yes. Drive IN- and tie IN+ to VDD for an inverting driver, or drive IN+ and tie IN- to ground for a non-inverting one.
What happens if the input floats? The output is held low by internal pull-up and pull-down resistors, so a disconnected controller cannot leave a gate floating.
Is an external clamp diode needed? Usually not. The output withstands 500 mA of reverse current and the body diodes clamp overshoot.
What temperature is it rated to? The electrical table is specified to 140 C junction while the ordering rows print minus 40 to 125 C ambient, so the junction has to be calculated.
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.