Contact Us

INA110SG

Brand: Texas Instruments
Texas Instruments INA110SG is a power management component designed for Industrial Automation and related applications. In stock at Superb Automation — request a quote for competitive pricing and fast delivery.
quote now

Product Specifications

INA110SG — Power Management & Signal Chain | Superb Automation

INA110SG

Texas Instruments — Power Management & Signal Chain

Product Overview

The INA110SG 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

INA110SG: the military grade FET input amplifier in a ceramic DIP

The INA110SG is a monolithic FET input instrumentation amplifier in the side braze ceramic DIP, and the S in the suffix is the military temperature grade that the vendor publishes from minus 55 to plus 125 degrees. It carries internal gain setting resistors for 1, 10, 100, 200 and 500 V/V, so most boards never fit an external gain resistor at all, and it settles to 0.01 per cent in 4 microseconds, which is the number that put the part into multiplexed data acquisition systems in the first place. Input bias current is 20 pA typical and 100 pA maximum, and common mode rejection is 106 dB minimum at a gain of 100, falling to 90 dB at 10 kHz.

The circuit is a current feedback topology with a laser trimmed input stage, and both inputs are protected against differential and common mode voltages up to the positive supply rail. That protection is not a small detail on a military platform: a transducer cable that is plugged in while the rack is powered will put the full rail on one input, and a part without the protection dies at the connector rather than at the sensor. The ceramic body is hermetic, which is the reason this grade still exists for programmes that cannot use a plastic package in a sealed enclosure with a desiccant pack.

The INA110SG rows as the addendum publishes them

Part classFET input instrumentation amplifier, current feedback
Vendor statusACTIVE, production data
Data sheetSBOS147A
Internal gains1, 10, 100, 200, 500 V/V by pin strapping
Gain equation with external resistorG = 1 + 40 kΩ divided by the sum of RG and 50 Ω
Gain range1 to 800 V/V
Gain temperature coefficient3 to 25 ppm per degree, worst row 100 ppm per degree
Input offset voltageLaser trimmed, expressed as a fixed term plus a term divided by gain
Offset drift2 µV per degree plus 20 µV per degree divided by gain
Input bias current20 pA typical, 100 pA maximum, each input
Input offset current2 pA typical, 50 pA maximum
Differential input impedance5 x 10^12 Ω in parallel with 6 pF
Common mode input impedance2 x 10^12 Ω in parallel with 1 pF
Common mode rejection at G = 100100 dB minimum, 110 dB typical, 90 dB at 10 kHz
Input noise at 10 kHz10 nV per root Hz, 1.8 fA per root Hz current noise
Bandwidth at G = 12.5 MHz small signal
Bandwidth at G = 100470 kHz small signal
Full power bandwidth190 kHz typical, 270 kHz maximum, 10 V output, G = 2 to 100
Slew rate12 V per microsecond typical, 17 V per microsecond maximum
Settling to 0.01 per cent4 µs at G = 1, 3 µs at G = 100, 11 µs at G = 500
Output swing10 V minimum, 12.7 V typical into 2 kΩ
Output current5 mA minimum, 25 mA short circuit
Capacitive load stability5000 pF
SupplyPlus or minus 15 V nominal, 18 V absolute maximum
BodyCDIP side braze, JD outline, 16 pins
Carrier24 pieces in a tube, the only carrier the addendum lists
Lead finishGold over nickel
Moisture levelNot applicable to this package type
Temperature rangeMinus 55 to plus 125 degrees
Part markingINA110SG

Two of those rows decide how the part is bought rather than how it is used. The 24 piece tube is a small quantity for a production line, so a programme that needs a few hundred pieces is buying ten tubes and the carrier count becomes part of the receiving inspection rather than an afterthought. The gold finish is the second: it solders without the wetting problem that a nickel palladium gold finish can show on a ceramic body, but it also means a tube that has been opened and closed in a humid room is the failure mode to watch, not the solder joint itself.

Internal gain taps and what they cost in bandwidth

The five internal taps are the reason this part is still specified in systems that were drawn decades ago. A gain of 100 by pin strapping removes the external resistor and its tolerance from the error budget, and the residual gain error is the internal ratio, which the vendor trims. The price is paid in bandwidth: the same amplifier that passes 2.5 MHz at unity gain is down to 470 kHz at a gain of 100 and 100 kHz at a gain of 500. On a multiplexed front end running at 100 thousand samples a second that is irrelevant, and on a pulse system that needs a 2 microsecond rise time it is the whole story.

The full power bandwidth of 190 kHz at 10 V output is the figure that is usually missed. Small signal bandwidth says nothing about how fast the output can move a 10 V step, and a channel that looks fine in a sine sweep will show a triangle on a square wave once the amplitude goes above a couple of volts. If the signal of interest is a pulse rather than a tone, the number to quote in a specification is the settling time at the gain actually used, not the unity gain bandwidth.

Bias current, guard rings and a 5 x 10^12 ohm input

A 20 pA bias current against a 5 x 10^12 Ω differential input is a combination that behaves well only if the board is clean. At that impedance a fingerprint across two traces is a resistor of a few hundred megohms, and the resulting offset is larger than the amplifier's own. The practical guard is a ring around both inputs driven from the reference pin, kept on the same layer as the input traces, and a cable whose shield is driven rather than grounded. The part has no dedicated guard pin, so the reference node is the only low impedance point available to drive the ring, which is acceptable because the ring only has to sit within a few volts of the inputs.

Current noise of 1.8 fA per root Hz is low enough that a 10 MΩ source resistance produces less noise than the 10 nV per root Hz voltage noise of the amplifier itself. That is the crossover that decides whether a FET input part is needed at all: below about 10 kΩ of source resistance a bipolar input amplifier has the better noise, and above a megohm the FET input wins by an order of magnitude. A piezoelectric transducer or a pH electrode sits firmly in the second group.

The two offset trim pins and the drift they do not fix

The package brings out an input offset adjustment and an output offset adjustment, and the difference matters. Trimming at the input removes the offset before the gain stage, so the correction scales with gain and stays correct when the gain tap is changed. Trimming at the output removes a fixed voltage regardless of gain, which is convenient for a fixed gain channel and wrong for a channel that switches taps. On a part with 2 µV per degree of drift, a trim at 25 degrees leaves a residual that grows with every degree away from the bench, and a 40 degree excursion adds 80 µV before the gain term.

The gain temperature coefficient of 10 to 50 ppm per degree is the other half of the budget. At a gain of 500 that is 0.5 to 2.5 per cent over a 100 degree military excursion, which is larger than the trimmed offset error on a 10 mV signal. A channel that is specified to 1 per cent over temperature on a 10 mV input has to be checked at both ends of the range, not just at room temperature.

Driving a capacitive load in a sealed chassis

The 5000 pF capacitive load stability is generous for an amplifier of this age, and it is the reason the part survives a long shielded cable without an isolation resistor. Above that value the output stage starts to ring, and the usual fix is a 50 to 100 Ω resistor in series with the output, placed close to the pin so that the cable capacitance is isolated from the feedback loop. A 100 Ω series resistor into a 10 kΩ load costs 1 per cent of gain, which is cheaper than a channel that oscillates at 400 kHz in a thermal vacuum test.

Output current of 5 mA minimum into 2 kΩ is enough for a 10 V swing and no more. A rack that fans the signal out to four destinations needs a buffer per destination rather than four loads on one amplifier, because 25 mA of short circuit current is a protection limit rather than something to operate at. Supply at plus or minus 15 V with an 18 V absolute maximum also means the part does not tolerate a 24 V rail, which is a common mistake when a card is moved from a 15 V chassis into a 24 V one.

Buying INA110SG

The part is active, the carrier is a 24 piece tube, and the finish is gold, so a purchase order is a simple string with a small quantity. The complication is the grade. The K and P suffixes in the same family are the plastic and SOIC versions in the industrial range, and the S is the ceramic military range; they share the data sheet and the internal gain taps but they are not interchangeable in a qualification, because the temperature range and the package are the qualification. A customer who quotes the SG for a commercial build is paying for a hermetic body they do not need, and a customer who substitutes a KU into a minus 55 degree slot has a part that is only specified to minus 40.

Stock for a ceramic military grade is thin by nature. The addendum lists one tube row, the vendor does not publish a reel, and the parts are usually bought against a programme schedule rather than held on a shelf. That makes the date code on the tube label the first thing to check on receipt, and a photograph of the label next to the tube the second. For a programme that has been running for twenty years, the bigger risk is not that the vendor stops building the part, it is that the part in the stores has been there longer than the solderability window.

Questions that arrive with this part number

Whether the SG can be replaced by the KU is the first question and the answer is no for a military slot, because the plastic body is specified to minus 40 and plus 85. Whether it can be replaced by a modern FET input amplifier is the second, and the honest answer is that a modern part will be faster and cheaper but will not be pin compatible, because the internal gain taps and the trim pins do not exist on the newer families.

The third question is about the trim pins: whether they can be left floating. They can, and the part meets its untrimmed offset specification with them open. What they cannot do is fix a drift problem, and a customer who trims at 25 degrees and expects the drift to fall with the offset will be disappointed at the cold end of the range.

For a defence programme that needs the grade documented, our defence grade sourcing page sets out what we ask the vendor for, and the amplifier cross reference page lists the families we can compare against when a tube cannot be found.