INA148UA
Product Specifications
INA148UA
Texas Instruments — Power Management & Signal Chain
Product Overview
The INA148UA 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
INA148UA: unity gain, plus or minus 200 volts, 260 microamps
INA148UA is the industrial INA148: unity gain, a plus or minus 200 volt common mode window at plus or minus 15 V, 0.075 per cent gain error, 86 dB of rejection typical, and 260 microamps of supply current. The UA suffix is the only grade in the family, and at 1 millivolt typical offset on a 200 volt range that is not a limitation.
The window is the whole point. On a single 5 V rail it becomes minus 47.5 to plus 32.5 volts with the reference at mid supply, or minus 4 to plus 75 volts with the reference at 0.25 volts, and that asymmetry is what a battery stack measurement has to work around. Fix the reference first, then check that the stack sits inside the window.
INA148UA as the vendor document publishes it
| Gain | 1 V/V fixed |
| Gain error | 0.01 per cent typical, 0.075 maximum |
| Gain drift | 3 ppm per degree typical, 10 maximum |
| Nonlinearity | 0.002 per cent maximum of full scale range |
| Offset | 1 millivolt typical, 5 maximum |
| Offset drift | 10 microvolts per degree maximum |
| Common mode range at plus or minus 15 volts | minus 200 to plus 200 volts |
| Common mode range at plus or minus 5 volts | minus 100 to plus 100 volts |
| Common mode range on 5 volts, reference at mid | minus 47.5 to plus 32.5 volts |
| Common mode range on 5 volts, reference at 0.25 volts | minus 4 to plus 75 volts |
| Common mode rejection | 70 dB minimum, 86 typical |
| Input impedance | 2 megohms differential, 1 megohm common mode |
| Voltage noise | 17 microvolts peak to peak, 880 nanovolts per root hertz at 1 kHz |
| Bandwidth | 100 kHz on SHE, 155 kHz on TID |
| Slew rate | 1 volt per microsecond |
| Settling to 0.01 per cent | 25 microseconds on a 10 V step |
| Overload recovery | 4 microseconds on SHE, 1.2 on TID |
| Quiescent current | 260 microamps |
| Supply range | 2.7 to 36 volts single, plus or minus 1.35 to plus or minus 18 dual |
| Stable load capacitance | 10 nanofarads |
| Body | eight pin SOIC |
| Temperature range | minus 40 to plus 85 degrees |
| Catalogue code | INA148UA |
| Grade | UA, the only grade published for this family |
| Carrier | tube and reel, confirm the row on the order |
Plus or minus 200 volts of common mode at unity gain
INA148UA is a unity gain difference amplifier with a thin film resistor network and a common mode window that reaches plus or minus 200 volts on a plus or minus 15 volt supply, plus or minus 100 volts on plus or minus 5, minus 47.5 to plus 32.5 volts on a single 5 V supply with the reference at mid rail, and minus 4 to plus 75 volts on the same 5 V supply with the reference at 0.25 volts.
The last two rows are the ones that catch people out. On a single supply the window is not symmetric and it moves bodily with the reference, so a channel that measures a 48 V battery stack on a 5 V rail has to set the reference near 0.25 volts to reach the top of the stack, and then loses the negative side. That is a property of the internal divider, not a limitation that can be trimmed away, and it decides the reference circuit before the converter is chosen.
2 megohms of differential input and 880 nanovolts of noise
The input impedance of INA148UA is 2 megohms differential and 1 megohm common mode, high enough that a 10 kohm source loses half a per cent and a 1 kohm source loses 0.05 per cent, both of which are inside the gain error. Voltage noise is 17 microvolts peak to peak from 0.1 to 10 Hz and 880 nanovolts per root hertz at 1 kHz, referred to the input, which on a 200 volt range is a negligible contribution to a 16 bit measurement.
The high input impedance is what makes the part usable across a high voltage divider without loading it, and it is also why the part does not need the source balancing that a 20 kohm difference amplifier demands. A protection resistor of 1 kohm in series with one input shifts the ratio by 0.05 per cent, so a matched pair is still the right practice, but the penalty for not matching is a hundred times smaller than on a low impedance network.
100 to 155 kHz, 1 volt per microsecond, 21 microseconds to settle
The bandwidth of INA148UA is 100 kHz on the SHE chip and 155 on TID, with a slew rate of 1 volt per microsecond, a settling time of 21 microseconds to 0.1 per cent and 25 to 0.01 on a 10 V step at plus or minus 15 volts, and an overload recovery of 4 microseconds on SHE and 1.2 on TID after a 50 per cent overdrive. The output reaches to within 0.25 volts of the negative rail and 1 volt of the positive into 100 kohm.
The scan arithmetic is straightforward: sixteen channels at 25 microseconds each is 400 microseconds, 2500 scans per second, which is faster than most high voltage multiplexers can switch and far slower than an audio channel needs. The overload recovery is the row that decides behaviour at power up, when the source is live and the supplies are still ramping, and 1.2 microseconds on the TID chip is the reason a board that hot plugs can behave differently from one that does not.
0.075 per cent gain error and 86 dB of rejection
Gain error on INA148UA is 0.01 per cent typical and 0.075 per cent maximum, with 3 ppm per degree typical and 10 maximum of drift, and nonlinearity at 0.002 per cent maximum of full scale range. Common mode rejection is 70 dB minimum and 86 typical on all three supply conditions the vendor specifies, which is the figure that survives a real source, because it is quoted at zero ohm source impedance with the window swept end to end.
On a 200 volt common mode, 70 dB converts to 63 millivolts of input referred error and 86 dB to 10 millivolts. That sounds poor until it is compared with the alternative, a resistive divider into an ordinary amplifier, where the ratio tolerance alone is worth more error and the divider dissipates continuously. The integrated network holds its ratio over temperature at 0.001 per cent of full scale, which is the part a discrete divider cannot match without a matched network of its own.
260 microamps for a plus or minus 200 volt front end
The quiescent current of INA148UA is 260 microamps, from a single supply between 2.7 and 36 volts or a dual supply between plus or minus 1.35 and plus or minus 18 volts. That is low enough for a loop powered transmitter and low enough that a battery instrument can leave the front end powered between conversions, which is unusual for a part that has to stand off 200 volts.
The supply window is wide in both directions, and the low end matters as much as the high end: at plus or minus 1.35 volts the part still works, with the common mode window scaled down to twice the supply, so the same circuit on a 3 V rail measures plus or minus 2.7 volts of common mode. Short circuit current is 13 milliamps on the SHE chip and 20 on TID, continuous into the common point, and the output is stable into 10 nanofarads, which is more than most parts in this class.
The SOIC-8 body and the automotive suffix question
INA148UA is an eight pin SOIC and the vendor specifies the industrial device over minus 40 to plus 85 degrees. The catalogue entry this page describes carries a Q1 suffix, which is the automotive qualification marker on the same body, and the vendor document that was read for these figures does not carry a separate automotive electrical table or an orderable addendum row for that string.
That is not a reason to avoid the part, but it is a reason to put the grade on the purchase order rather than in a note. The industrial and automotive rows differ in test flow and in documentation, not in the silicon, and a build that needs the automotive paperwork needs the specific orderable code on the line. Until that code is confirmed against a customer purchase order, the honest position is that the page describes the device and the suffix has to be verified with the supplier.
This is a one grade part with a wide supply window, so the buying question is carrier and lot, not bin. Our amplifier cross reference page covers the alternative high voltage parts, including the pin compatible INA149, and the pin to pin page covers what a swap has to preserve.
Can the part measure a 400 volt stack on a plus or minus 15 V supply? The vendor specifies plus or minus 200 volts, so a 400 V total that swings symmetrically sits exactly at the limit with nothing left for tolerance. A stack that floats or that can ring needs either a divider ahead of the part or a part with a wider window.
Is the 2 megohm input impedance real at dc? It is the differential figure the vendor publishes, with 1 megohm common mode, and it is high enough that a 10 kohm source loses half a per cent, which is inside the gain error window but not inside a 16 bit budget without calibration.