INA240A1DR
Product Specifications
INA240A1DR
Texas Instruments — Power Management & Signal Chain
Product Overview
The INA240A1DR 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
INA240A1DR: 80 V bidirectional current sense amplifier with 132 dB rejection and a gain of 20
INA240A1DR is a bidirectional current sense amplifier in the eight pin SOIC body that measures across a shunt anywhere between 4 V below ground and 80 V above it, with 25 uV of maximum offset and a fixed gain of 20 in a SOIC body of 8 pins. The INA240A1 is the gain of 20 member of a four part family, with gains of 50, 100 and 200 in the A2, A3 and A4 codes. Common mode input spans minus 4 V to 80 V, the supply is 2.7 V to 5.5 V and the quiescent current is 2.4 mA maximum. A differential output stage and an enhanced rejection scheme let it measure a shunt in a switching motor leg, and two reference pins set the zero point for bidirectional current.
Order code and package
| Order code | INA240A1DR |
| Gain | 20 V per V |
| Common mode input range | minus 4 V to 80 V |
| Supply voltage | 2.7 V to 5.5 V |
| Supply current | 1.8 mA typical, 2.4 mA maximum |
| Package | eight pin SOIC, D body |
| Reference inputs | VREF1 and VREF2, 0 V to VS |
| Family gains | 20, 50, 100 and 200 V per V in A1 to A4 |
| Document | SBOS662C, July 2016, revised December 2021 |
| Temperature range | minus 40 C to 125 C |
Electrical parameters that decide the application
| Common mode rejection, DC | 120 dB minimum, 132 dB typical |
| Common mode rejection at 50 kHz | 93 dB typical |
| Offset voltage, input referred | plus or minus 5 uV typical, plus or minus 25 uV maximum |
| Offset drift | plus or minus 50 nV per C typical, plus or minus 250 nV per C maximum |
| Gain error | plus or minus 0.05 percent typical, plus or minus 0.20 percent maximum |
| Gain drift | plus or minus 0.5 ppm per C typical, plus or minus 2.5 ppm per C maximum |
| Non linearity | plus or minus 0.01 percent |
| Reference divider accuracy | 0.02 percent typical, 0.1 percent maximum |
| Reference rejection, gain of 20 | 20 uV per V |
| Bandwidth at minus 3 dB | 400 kHz at every gain |
| Bandwidth at 2 percent distortion | 100 kHz |
| Settling to 0.5 percent, gain of 20 | 9.6 us |
| Slew rate | 2 V per us |
| Input referred noise density | 40 nV per root hertz |
| Output swing to the positive rail | VS minus 50 mV typical, VS minus 200 mV maximum |
| Output swing to ground | 1 mV typical, 10 mV maximum |
| Maximum capacitive load | 1 nF |
| Power supply rejection | plus or minus 1 uV per V typical, plus or minus 10 uV per V maximum |
Points to check before this part goes on a BOM
Offset sets the floor of the measurement and the shunt sets where that floor lands. Twenty five microvolts of maximum offset across a 1 milliohm shunt is 25 mA of apparent current, and across a 10 milliohm shunt it is 2.5 mA. With a 1 milliohm shunt and a 20 A full scale, the sense voltage is 20 mV, so the offset is 0.125 percent of full scale and the gain error of 0.2 percent is the larger term. Below about 10 percent of full scale the picture reverses and the offset dominates, which is why a drive that has to measure small currents accurately needs a larger shunt and a thermal budget for it.
The common mode range reaches 4 V below ground, which is what makes the part usable in a motor phase. During commutation the phase node is pulled below the negative rail by the freewheeling path, and a device that stops at ground would clip its input and produce a false zero. Four volts is a working figure rather than an absolute maximum, so a phase that rings to minus 8 V needs a series resistor and a capacitor at the input pins, and the resistor then adds its own error with the 90 uA of input bias current.
The 93 dB rejection at 50 kHz is the number that separates this part from a general purpose amplifier. A 48 V common mode step attenuated by 93 dB is 1.07 mV referred to the input, and at a gain of 20 that is 21 mV at the output, which is about 0.1 percent of a 4.8 V full scale output swing. Without the rejection scheme the same step would arrive nearly whole and the current reading on a PWM edge would be dominated by the switching rather than by the load.
The two reference pins set the zero current point and their divider accuracy of 0.1 percent is a direct offset error. With both pins driven from a 3.3 V rail through a divider to mid supply, 0.1 percent of 1.65 V is 1.65 mV at the output, which is 82 uV referred to the input and therefore 82 mA across a 1 milliohm shunt. That is three times the amplifier offset, so a board that stops at the amplifier specification and ignores the reference network leaves most of the achievable accuracy on the table.
Bandwidth is 400 kHz at every gain, which is unusual and useful: the internal stages are arranged so the gain does not trade bandwidth. The distortion figure, however, is quoted at 100 kHz, and the settling time to 0.5 percent is 9.6 us. In a 20 kHz PWM drive the period is 50 us, so a 9.6 us settle is about a fifth of the period and the current sample has to be taken after that window, not at the start of the on time.
The output can swing to 50 mV below the positive rail and to within 1 mV of ground, so a bidirectional measurement at mid supply uses most of the 3.3 V range. The 200 mV maximum figure at the top rail is the one to use when the output feeds an analogue to digital converter whose input range ends at the rail, because a 200 mV shortfall at the top of a 3.3 V range is 6 percent of the span and looks like a gain error unless it is trimmed.
Quiescent current of 2.4 mA maximum over temperature is modest for a current sense amplifier and it is not zero. On a battery pack monitor that samples once a second, the amplifier can be powered down between samples if the reference network is also switched, but a continuous monitor draws 2.6 mA over temperature, which is 23 mAh a day and rules the part out for a coin cell product.
Where this order code fits
This order code belongs on phase current sensing in three phase motor drives, solenoid and valve control, battery current on 48 V and 60 V packs, and current limited outputs on telecom and industrial supplies. It is the part chosen when the shunt sits in a node that swings tens of volts at tens of kilohertz, because the rejection figure rather than the gain error is what decides whether the current reading survives the switching edge. 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 Pin to Pin Cross Reference page.
Questions buyers ask
What shunt should go with the gain of 20 version? The gain of 20 exists for shunts with a large voltage across them, typically 20 mV to 100 mV at full scale. With a 5 milliohm shunt and 20 A the sense voltage is 100 mV and the output at a gain of 20 is 2 V, which fits a 3.3 V supply with room for the bidirectional offset. A 1 milliohm shunt and the same 20 A gives 20 mV, which uses only a fifth of the output range and lets the 25 uV offset become 0.125 percent of full scale.
Can it measure a phase node that goes below ground? The common mode range starts at minus 4 V, so a phase node that rings a few volts below the negative rail is inside the part. Anything beyond that needs input resistors and a differential capacitor to keep the pins inside the range, and those components add error through the 90 uA of bias current. The right answer for a high side drive with fast commutation is to check the ringing amplitude at the pins with a scope before the first board is committed.
How accurate is the current reading at 10 percent load? At 10 percent of a 20 A full scale with a 1 milliohm shunt the sense voltage is 2 mV against an offset limit of 25 uV, which is 1.25 percent, plus 0.2 percent of gain error and the reference divider term. The honest figure is about 1.5 percent of reading at that point, and it improves as the load rises. A drive that needs 1 percent at 10 percent load needs a larger shunt, not a better amplifier.
Does the 400 kHz bandwidth mean the part can follow a PWM current ripple? It can follow the fundamental of a 20 kHz ripple with phase lag but not the edges. The 9.6 us settling time to 0.5 percent is the useful number: at 20 kHz the current has to be sampled at least 10 us after the switching edge, which leaves roughly 15 us of a 50 us period for a clean sample. Averaging the sample over several periods, or synchronising the conversion to the PWM, is the usual approach.
SUPERB supplies this family against the exact order code, including the gain variant, and states the date code and packing on the offer. When a drive is moving from one current sense part to another, the How to Verify a Drop in Replacement page lists the checks that decide whether the swap holds across the full temperature range.