INA296A1IDDFR
Product Specifications
INA296A1IDDFR
Texas Instruments — Power Management & Signal Chain
Product Overview
The INA296A1IDDFR 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
INA296A1IDDFR: 10 V/V bidirectional current sense amplifier with a 110 V common mode window in the eight pin SOT-23-THIN
INA296A1IDDFR is an ultra precise bidirectional current shunt monitor for high voltage rails, the 10 V/V member of the A grade in a SOT-23-THIN body of 8 pins. It measures across a shunt at common mode from minus 5 V to plus 110 V with a survival rating of minus 20 V to plus 120 V, runs from a 2.7 V to 20 V supply at 2.5 mA, and holds a plus or minus 10 uV offset with a plus or minus 0.01 per cent gain error in the A grade. The bandwidth is 1.1 MHz at every gain, which is the figure that makes it usable for transient detection and fast overcurrent protection rather than only for DC metering.
Order code and package
| INA296A1IDDFR | SOT-23-THIN DDF body, 8 pins, 3000 pieces on a large reel, Active in the addendum dated 8 November 2025, NIPDAU finish, moisture level 1 with no floor life limit, marking 2SA3, minus 40 to plus 125 degrees |
| INA296A1IDGKR | VSSOP DGK body, 8 pins, 2500 pieces on a large reel at moisture level 2 with a one year floor life, marking I296A1 |
| INA296A1IDR | SOIC D body, 8 pins, 2500 pieces on a large reel at moisture level 2, marking I296A1, which is the body to use where the board has the room and reflow profile of a leaded package |
| INA296B1IDDFR | the B grade in the same body, with a plus or minus 150 uV offset and a plus or minus 0.1 per cent gain error, which is ten times the offset and ten times the gain error of the A grade at a lower price |
| INA296A2IDDFR | the 20 V/V member of the A grade in the same DDF body and the same 3000 piece reel, marking 2SB3, which is the sibling to compare when the shunt value is fixed by thermal limits |
Electrical parameters that decide the application
| Common mode range | minus 5 V to plus 110 V operational, with a survival rating of minus 20 V to plus 120 V, independent of the supply |
| Supply | 2.7 V to 20 V single rail, 2.5 mA of supply current, which is an order of magnitude above the low power monitors and the price of the bandwidth |
| Offset voltage | plus or minus 10 uV maximum in the A grade, plus or minus 150 uV in the B grade, both referred to the input |
| Offset drift | 0.1 uV per degree in the A grade and 0.5 uV per degree in the B grade, over minus 40 to plus 125 degrees |
| Gain | 10 V/V on the A1 member, against 20, 50, 100 and 200 V/V on A2 to A5, each also available in the B grade |
| Gain error | plus or minus 0.01 per cent maximum in the A grade with 1 ppm per degree of drift, plus or minus 0.1 per cent and 5 ppm per degree in the B grade |
| Common mode rejection | 166 dB typical, which is the figure the vendor prints in the features list rather than a guaranteed minimum |
| Bandwidth | 1.1 MHz at every gain, so the small signal bandwidth does not fall as the gain rises |
| Slew rate and settling | 8 V per microsecond, with 1 us settling to within 1 per cent of final value, which is what a fast overcurrent loop needs |
| Reference pins | two reference pins rather than one, which allows the zero current level to be placed anywhere inside the output range for bidirectional sensing |
| Noise | quoted per gain in the electrical table, with the 10 V/V member carrying the lowest output referred noise of the family |
| Temperature | minus 40 to plus 125 degrees on both grades and all three bodies |
Family variants and how they differ
| A grade against B grade | the A grade carries 10 uV of offset and 0.01 per cent of gain error, the B grade 150 uV and 0.1 per cent. The two are the same silicon and the same pinout, so the grade is chosen by the error budget and never by the footprint |
| INA240 as the motor drive alternative | a 80 V common mode window with enhanced common mode rejection for pulse width modulated buses, where this part's 110 V window is not needed and the PWM rejection is |
| INA296 against INA290 | the INA290 is the 110 V member of the lower bandwidth family, with a 1 MHz figure but a much lower supply current, for DC metering where 8 V per microsecond is not needed |
| INA186 against this part | 40 V common mode and 48 uA of supply current, which is a fifth of the current and a third of the window, for battery equipment rather than for a 48 V rack |
| Body options | DDF, DGK and D are the same silicon, so the choice is footprint, reel size and moisture level rather than electrical grade |
Points to check before this part goes on a BOM
The 166 dB common mode rejection is a typical figure and the guaranteed minimum is not printed in the features list. That matters in a 48 V system because the common mode is not static: a hot plug event moves the bus by tens of volts in a microsecond and any common mode to differential conversion appears as a current step. The switched supply rejection of the stage is what actually protects the reading, and a circuit that needs a guaranteed number has to read the electrical table rather than the front page.
The A grade offset of 10 uV is what makes a low value shunt viable. On a 1 milliohm shunt, 10 uV is 10 mA of apparent current, and at 50 A full scale that is 0.02 per cent of range. The same shunt with the B grade is 150 mA of apparent current, which at 50 A is 0.3 per cent, so the grade choice is worth fifteen times the offset and has to be made before the shunt is chosen, not after.
The 1.1 MHz bandwidth at every gain is unusual for a current sense amplifier and it comes with a cost: 2.5 mA of supply current. On a 48 V server supply that current is irrelevant. On a battery powered instrument it is the entire budget. The family therefore splits cleanly into metering parts at tens of microamps and protection parts at milliamps, and this is the protection part.
Two reference pins allow the zero current output to sit anywhere in the output range, which is what bidirectional sensing needs. A single reference pin tied to mid supply would do the same thing on a fixed 3.3 V rail, but on a 5 V rail with a 20 V supply the divider has to be matched to the output swing, and the second pin allows the positive and negative current ranges to be trimmed separately. The datasheet asks for both pins to be driven from low impedance sources.
The survival rating of minus 20 V to plus 120 V is not an operating range and the difference matters. A 48 V bus that rings to 90 V on a hot plug event is inside the survival rating and outside the operating range, so the reading during the ring is not specified but the part survives. A bus that rings to 130 V destroys the input stage, and no amount of common mode rejection helps, so a transient voltage suppressor on the bus is the correct answer rather than a higher rated amplifier.
The 8 V per microsecond slew rate and 1 us settling to 1 per cent are what separate this part from the metering monitors. A 50 A overcurrent event on a 1 milliohm shunt produces a 50 mV step and the output reaches within a per cent of its final value in a microsecond, so a comparator on the output can act inside 2 us. The low power parts at 0.3 V per microsecond take twenty times as long, which is fine for a gauge and useless for protection.
Using the part on a 48 V bus
The operational window covers 48 V with 62 V of headroom and the supply can be the 3.3 V logic rail, which is the whole point of a high voltage current sense amplifier: no divider, no isolated supply, and the shunt sits in the bus rather than in the return. The one condition is that the supply has to come up before or with the bus, because the input stage is powered from the supply and an unpowered amplifier presents its inputs as a diode drop to ground.
The alternative that people reach for first is a shunt in the return with a low voltage amplifier. That works, but it puts the measurement on the ground side where a short circuit takes out both the shunt and the amplifier, and it makes the ground of the load and the ground of the measurement the same node, which is a noise problem in a rack. The high side part costs more and removes both issues.
The B grade as a cost route
Where the reading is used for a status flag rather than a control loop, the 150 uV offset of the B grade is immaterial and the saving is real. The trap is a later change of use: a board that started as a status indicator and became a revenue meter needs the A grade, and the two are pin compatible so the change is a part number edit that nobody schedules a validation for.
Where this order code fits
48 V DC to DC converters, 48 V battery management, test and measurement, remote radio units and 48 V rack servers are the applications the vendor names, and every one of them puts the shunt on a bus that is far above the logic supply. Teams comparing it with a neighbouring part start from our Amplifier Cross Reference page, and teams replacing a part that has gone end of life start from the Power Management Cross Reference page.
Questions buyers ask
Does it need a separate high voltage supply? No. The supply is 2.7 V to 20 V and the common mode window is independent of it, so a 3.3 V rail powers a measurement on a 48 V bus.
Is 166 dB common mode rejection guaranteed? No. That figure is typical, and the guaranteed minimum has to be read from the electrical table row that carries the test conditions.
What is the difference between the A and B grades? Offset 10 uV against 150 uV and gain error 0.01 per cent against 0.1 per cent, in the same body and the same pinout.
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.