Contact Us

UCC27712DR

Brand: Texas Instruments
Texas Instruments UCC27712DR 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

UCC27712DR — Power Management & Signal Chain | Superb Automation

UCC27712DR

Texas Instruments — Power Management & Signal Chain

Product Overview

The UCC27712DR 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

UCC27712DR: a 620 V high-side and low-side gate driver with 1.8 A of source current, 2.8 A of sink current, 100 ns of typical propagation delay and 150 ns of internal dead time, in the eight pin SOIC on a 2500 piece reel

UCC27712DR is a half bridge driver for two N-channel FETs or IGBTs, with separate high-side and low-side inputs, an interlock that stops both outputs being on together and a floating channel fed from a bootstrap capacitor in a SOIC body of 8 pins. The D body measures 4.90 mm by 3.91 mm and is the outline the industry already routes, with 75 pieces in a tube for a first article and 2500 on a reel for production.

Order code and package

Catalogue stringUCC27712DR, where D is the eight pin SOIC and R is the 2500 piece large tape and reel
BodySOIC, D, eight pins, 4.90 mm by 3.91 mm of moulding; the body carries both channels and the level shift, so no second part is needed for the high side
TubeUCC27712D ships 75 pieces in a tube, which is the route for a prototype or a first article build
Revision rowsthe plain row and the .B row are both Active at the same reel size, so the vendor has an active revision running alongside the original
Finish and moistureNIPDAU plating, moisture level 2 at 260 C with one year of floor life, so the bag has to be resealed after the first use
Temperatureminus 40 to 125 C on the published rows, with the electrical table specified to 150 C junction
MarkingU27712 on the body
DocumentSLUSCE9B, June 2017, revised March 2020
StatusActive and in production on the addendum dated 2026

Electrical parameters that decide the application

Switch node620 V fully operational and 700 V absolute maximum on the HB pin, which is the rating that decides whether the part fits a 400 V bus with an 80 V margin of overshoot
Bias rangeVDD recommended from 10 V to 20 V for an IGBT and 10 V to 17 V for a power MOSFET, with 20 V as the top of the useful range
Peak current1.8 A source and 2.8 A sink, asymmetric on purpose, because turning a FET off quickly is what limits shoot through while turning it on slowly is what limits ringing
VDD lockoutrising at 8.9 V typical, 8.0 V minimum and 9.8 V maximum, falling at 8.4 V typical, so a 12 V rail starts the part and a 10 V rail is inside the window at the bottom of its tolerance
Bootstrap lockoutrising at 8.2 V typical, 7.2 V minimum and 9.2 V maximum measured from HB to HS, falling at 7.3 V typical
Quiescent current255 uA typical and 420 uA maximum total, with 65 to 100 uA drawn from the bootstrap rail, all measured with the inputs held static
Dynamic current3.8 mA at 100 kHz and 50 per cent duty into a 1 nF load, which is what the bias supply has to deliver when the bridge is switching
Input thresholds2.0 V typical high and 1.2 V typical low, with 0.8 V of hysteresis and an absolute maximum that is independent of the bias rail
Propagation delay100 ns typical and 160 ns maximum in both directions, with 12 ns of typical matching between the two channels
Dead time150 ns typical, 100 ns minimum and 200 ns maximum, generated internally by the interlock and therefore not adjustable
Minimum pulse25 ns on and 35 ns off, so a controller that produces narrow pulses at light load does not lose them in the driver
Output stage30 to 80 Ohm of pull-up and pull-down resistance, with 60 to 136 mV of drop at 20 mA, and the stages survive a 500 mA reverse current without a clamp diode
Negative ratingsthe switch node is specified for logic operation to minus 11 V and the inputs to minus 5 V, which is the margin that keeps a bridge alive on the turn-off edge

What the eight pins do

HI, pin 1the high-side input, referenced to the logic return rather than to the switch node, with an internal pull-down that holds the output off when the pin floats
LI, pin 2the low-side input, with the same logic threshold and the same pull-down, and the interlock between HI and LI stops both outputs being high at once
VDD, pin 3the low-side bias input, monitored by a lockout that holds both outputs low until the rail is above 8.9 V
HB, pin 6the bootstrap rail for the floating high-side driver, charged through an external diode from VDD while the low side is on
HO, pin 7the high-side gate output, referenced to HS, with 1.8 A of source current and 2.8 A of sink current
HS, pin 8the return for the high-side driver and the connection to the switch node, specified to 620 V and tolerant of minus 11 V of logic excursion
LO, pin 5the low-side gate output, referenced to COM and driven from the same bias rail as the logic

Points to check before this part goes on a BOM

The interlock is the reason this part exists in a bridge rather than two single drivers. Both outputs cannot be high at the same time, and the internal dead time of 150 ns is inserted between the turn-off of one channel and the turn-on of the other. That dead time is fixed, which is both the safety and the limit: a build with a slow IGBT that needs 800 ns of dead time cannot get it from this part, and a build with a fast silicon carbide FET that wants 50 ns has to accept the 150 ns and the body diode conduction that comes with it. At 100 kHz the dead time is 1.5 per cent of the period, and in a 400 V bridge that is a real loss rather than a rounding error.

The 620 V switch node rating and the minus 11 V logic tolerance are two different numbers for two different events. The 620 V is the continuous and switching level the high side can sit at, with 700 V as the absolute maximum for a transient that lasts less than the recovery time of the bootstrap diode. The minus 11 V is the excursion the high-side logic survives when the low-side FET turns off and the source inductance drives the switch node below ground. A 20 nH source connection carrying 20 A switched in 20 ns produces about 20 V of kick, and a driver rated to minus 5 V on that pin will latch on the first hard turn-off.

The bootstrap path is external on this device, which means a diode and a capacitor per bridge leg and a choice about the diode. A slow recovery diode in that position dissipates the charge that the switch node pumps back through it, and a diode with too much forward drop leaves the high-side gate under-driven at low duty. The data sheet's 65 to 100 uA of bootstrap current plus the gate charge of the FET sets the capacitor: a 100 nF capacitor with 8 nC of gate charge and 2 uC of leakage loses about 0.1 V per switching cycle at 100 kHz, so 1 uF is the usual answer and 100 nF is the usual mistake.

The 8.9 V VDD lockout is a narrow window for a 12 V rail that has to start in a cold room. A regulated 12 V rail with 5 per cent tolerance sits between 11.4 V and 12.6 V, which is comfortably above the 9.8 V maximum turn-on threshold, so the part starts. The problem appears when the gate drive is fed from a bootstrap winding that only produces 11 V at low line, or from a resistor dropper that sags under load: the part then works on the bench and sits in lockout in the field, and the symptom is a bridge that stops switching at low mains rather than a part that failed. The lockout on the high side at 8.2 V typical is the second half of the same problem, because the bootstrap rail sits one diode drop below VDD.

Power dissipation is small compared with the FETs and still worth the arithmetic. Driving a 60 nC gate at 15 V and 100 kHz is 90 mW per channel, and both channels running through the 3.8 mA of dynamic current add about 57 mW at 15 V, so the driver dissipates roughly 240 mW in the eight pin body. The thermal resistance of that body is not quoted in the table, but the same outline in this family sits near 120 C per W, which puts the rise at about 29 C and makes the part safe at 85 C ambient and marginal inside a sealed enclosure at 105 C. The fix is a smaller gate charge or a driver in a body with an exposed pad.

The input logic is CMOS and TTL compatible and independent of the bias rail, so a 3.3 V controller drives a driver that runs from 15 V with no level shifter. What the input is not is tolerant of a slow edge: a signal that spends more than a few microseconds between 1.2 V and 2.0 V can produce more than one transition at the output, and the interlock then inserts dead time where none was wanted. A series resistor and a capacitor at the controller output, which is a common way to slow a gate drive, is therefore the wrong tool at this pin; the slow edge belongs at the gate, not at the input.

Dead time, body diode loss and the 150 ns fixed number

Every nanosecond of dead time is a nanosecond during which the inductor current flows in a body diode with a forward drop of about 1 V, and in a 400 V to 12 V converter at 10 A that is 10 W of loss during the interval. At 100 kHz and 150 ns the diode conducts 1.5 per cent of the time, which is 150 mW, and the same bridge at 300 kHz would lose 450 mW. The number is not a reason to reject the part, but it is the reason a build with a high switching frequency looks at a driver with an adjustable dead time instead.

The other side of the same arithmetic is shoot through. A dead time that is too short puts both FETs on for the duration of the turn-off tail, and the current that flows is limited only by the loop inductance. The 100 ns minimum of this part exists to make that impossible with a FET whose turn-off is faster than 100 ns, and a board that parallels devices with a slow turn-off should measure the actual switch node edge rather than trust the number on the page.

Choosing between this driver and one with an integrated bootstrap diode

A driver with the bootstrap diode on the die removes a component and a placement decision, and it caps the bootstrap voltage at VDD minus the internal diode drop, which rises with current. This part leaves the choice of diode to the board, which costs one part and allows a Schottky with 0.3 V of drop where the high-side gate needs every volt of drive, or a fast recovery diode where the switch node is at 400 V and the reverse recovery of a Schottky would be a problem.

The trade is worth making when the bridge runs at a low duty cycle, because the bootstrap capacitor then has very little time to recharge and the diode drop matters. It is not worth making when the bill of materials is the constraint and the duty cycle is near 50 per cent.

Where this order code fits

Half bridges and full bridges in motor drives, offline converters and inverters, synchronous bucks that run from a 48 V or 400 V rail, and Class D output stages where the two channels have to be interlocked 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 both outputs be on at once? No. The interlock prevents it, and the internal dead time of 150 ns sits between one channel turning off and the other turning on.

How much gate charge can it drive? 1.8 A source and 2.8 A sink, which is a gate charge of about 100 nC at 20 kHz and needs a buffer stage above that.

Does it include the bootstrap diode? No. The bootstrap diode and capacitor are external, which lets the diode be chosen for the switch node voltage and the duty cycle.

What is the switch node rating? 620 V operational and 700 V absolute maximum, with logic operation to minus 11 V on the switch node and minus 5 V on the inputs.

Is the dead time adjustable? No. It is fixed at 150 ns typical by the interlock, so a board that needs a longer dead time needs a different driver or an external delay.

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.