LM317AEMP
Product Specifications
LM317AEMP
Texas Instruments — Power Management & Signal Chain
Product Overview
The LM317AEMP 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
LM317AEMP: the A grade of the 1.5 A floating regulator
LM317AEMP is an A grade string of the three terminal adjustable positive regulator, the 1.5 A member of the LM317 family. The circuit is floating: there is no ground pin, so what the device regulates is the difference between its input and its output, and any output voltage from 1.25 V to 37 V can be set with two resistors. A build that needs a 48 V rail or a 200 V rail can still use this part, as long as the differential across the device never exceeds the 37 V operating window.
The A letter is the reason buyers ask about this string at all. The device comparison table in SLVS044Z prints the plain device and the A grade side by side, and the row that separates them is load regulation accuracy, printed as 1 percent on the A grade against 1.5 percent on the plain device. Everything else in the same table, the 4.25 V to 40 V input window, the 1.5 A current class and the 80 dB ripple rejection figure, is shared.
What follows is what the data sheet fixes for the family, the rows that move with temperature, what the packaging addendum actually prints for carriers, and how a buyer should read an EMP suffix that the current revision does not itemise.
What the order code sits on
SLVS044Z, revised April 2025, is the document behind this family. It publishes the regulator as a three terminal floating device with internal current limiting, thermal overload protection and safe area compensation, and it states that overload protection stays active even if the adjustment pin is left open. That last property is why the part survives a broken solder joint on the adjustment node in the field instead of running the load at full input voltage.
| Function | adjustable positive voltage regulator, floating |
| Output current class | 1.5 A |
| Output voltage window | 1.25 V to 37 V, set by two external resistors |
| Input to output differential | 4.25 V to 40 V operating window |
| Reference voltage at 25 degrees | 1.2 V minimum, 1.25 V typical, 1.3 V maximum |
| Line regulation at 25 degrees | 0.01 percent per volt typical, 0.04 maximum |
| Line regulation over temperature | 0.02 percent per volt typical, 0.07 maximum |
| Load regulation at 25 degrees | 0.1 percent typical, 0.5 maximum |
| Load regulation over temperature | 0.3 percent typical, 1.5 maximum |
| Thermal regulation, 20 ms pulse | 0.04 percent per watt typical, 0.07 maximum |
| Adjustment pin current | 50 microamps typical, 100 maximum |
| Adjustment pin current change | 0.2 microamps typical, 5 maximum |
| Minimum load current | 3.5 milliamps typical, 10 maximum |
| Current limit | 1.5 A minimum, 2.2 A typical at 25 degrees |
| Ripple rejection at 120 Hz | 64 dB legacy chip, 80 dB new chip with 10 microfarads on the adjustment pin |
| Output noise, 10 Hz to 10 kHz | 0.003 percent of output voltage |
| Temperature stability | 0.7 percent on the legacy chip, 1 percent on the new chip |
The rows that move with temperature
The reference sits at 1.2 V to 1.3 V at room, and that 80 millivolt spread is the whole error budget of a resistor divider before anything else is counted. A 240 ohm resistor from output to adjustment pin plus a 5 k ohm resistor to ground gives roughly 27 V, and a 1 percent spread on the reference alone moves that output by 270 millivolts. On a 3.3 V rail built from the same part the spread is 33 millivolts, which is why the part is usually used for the higher rails and left out of tight digital supplies.
Load regulation widens from 0.5 percent at 25 degrees to 1.5 percent over the full window on the new chip, and line regulation moves from 0.04 to 0.07 percent per volt. For a 12 V rail feeding a 500 mA load, 1.5 percent of load regulation is 180 millivolts of sag between no load and full load, so a system that samples the rail with a 12 bit converter across a 4.096 V reference has to be checked at both ends of the load range and not at the bench condition alone.
The adjustment pin current matters more than its 50 microamps suggest. It flows through the upper resistor of the divider, so a 5 k ohm upper resistor turns 100 microamps into 500 millivolts of offset that appears as an output error and drifts with temperature. A 240 ohm upper resistor keeps that term under 24 millivolts, which is the reason the data sheet example uses 240 ohm rather than a larger value, and the reason a low current version of the family exists for standby rails.
Minimum load current is the row that catches battery backed circuits. Regulation is only guaranteed with at least 3.5 milliamps, and up to 10 milliamps in the worst case, flowing out of the device. A rail that is allowed to sit at no load for long periods will drift high, so a bleeder resistor sized for 10 milliamps is part of the circuit and not an optional extra.
Carriers and what the addendum prints
The packaging addendum in this revision prints four bodies: the four pin SOT-223 in a 6.5 mm by 7 mm outline, the three pin TO-263 in a 10.16 mm by 15.24 mm outline, and two TO-220 rows at 10.16 mm by 4.55 mm, one of them marked as the legacy chip. Reel quantities follow the body: 2500 pieces on a 330 mm reel with 12.4 mm tape for the SOT-223, and 500 pieces on a large reel for the TO-263.
| LM317DCY | SOT-223, 4 pins, 80 pieces per tube, moisture level 2, 0 to 125 degrees |
| LM317DCYR | SOT-223, 4 pins, 2500 pieces on a large reel, moisture level 2 |
| LM317KTTR | TO-263, 3 pins, 500 pieces on a large reel, moisture level 3, 245 degrees for 168 hours |
| LM317KCS | TO-220, 3 pins, 50 pieces per tube, printed not recommended for new designs |
| Part marking on SOT-223 | L3 |
| Reel diameter, SOT-223 | 330 mm with 12.4 mm tape, box 340 mm by 340 mm by 38 mm |
| Lead finish | tin over the standard finish, RoHS compliant on the reel rows |
| Order code in this page | LM317AEMP, A grade with the EMP carrier suffix |
| What the addendum prints for A grade | the A grade column in the comparison table, no EMP row in this revision |
| Practical consequence | carrier and reel quantity for the EMP suffix have to be confirmed before a production order |
The EMP suffix is where a buyer has to slow down. The current revision prints DCY, KTT, KCS and KCT rows, and the A grade appears in the device comparison table with its tighter load regulation figure, but no orderable row in this revision carries an EMP suffix. That does not mean the string is unavailable, it means the suffix belongs to an older ordering form and the carrier has to be confirmed against the vendor's live orderable list rather than inferred from this page.
Thermal budget in SOT-223 and TO-263
The SOT-223 body dissipates through the tab and the copper under it, and the data sheet section on heatsinking is explicit that the junction to ambient figure is a function of the copper plane rather than of the package alone. With a 12 V input and a 5 V output at 700 mA, the device drops 4.9 W, and the SOT-223 will not hold that without a plane of several square centimetres on the tab.
The TO-263 handles the same dissipation with a far smaller board area because the tab is the heatsink interface, and the 500 piece reel quantity is the practical difference between the two for a production line: the SOT-223 runs on a 2500 piece reel that suits a pick and place feeder, while the TO-263 reel holds 500 pieces and is often bought as a partial reel for a build that runs once a quarter.
Thermal regulation of 0.04 percent per watt typical means a 5 W dissipation shifts the output by 0.2 percent, about 33 millivolts on a 16 V rail, purely from the junction heating up. That figure is a settled state shift, not a transient, so a test that runs the board for two seconds and reads the rail will not see it.
Where a 1.5 A floating regulator still earns its place
Three cases keep this device in production. The first is a high voltage rail where a switching regulator is not wanted, for example a 90 V bias supply in an instrument, because the floating topology lets the device sit on top of the output and see only the differential. The second is a current source, where a single resistor between output and adjustment turns the part into a constant current sink for LED strings or battery charging. The third is a post regulator after a switching stage, where 80 dB of ripple rejection at 120 Hz cleans up the switching residue without a second converter.
In each case the part is chosen for what it does at the extremes rather than at 25 degrees. A rail that runs at 90 V output and 15 mA is a 1.35 W dissipation, which the TO-263 absorbs on a modest plane and which the TO-220 absorbs only with a clip on heatsink. The power management cross reference page covers the pin compatible alternatives for this body, and the drop in replacement check is the procedure we use before a second source is approved on a board that already has a qualification file.
Sourcing notes and questions that come with this code
Is the A grade a different silicon? The comparison table prints the same 1.5 A class, the same 4.25 V to 40 V input window and the same ripple rejection, and the visible difference is the load regulation accuracy row. Treat the A grade as a tighter tested grade rather than as a different circuit.
Can the TO-263 row be used where the board has the SOT-223 footprint? No. The two are different footprints and different pin counts, three pins against four, with the tab on a different pitch. Substitution is a board change and not a carrier change.
What should a buyer confirm before ordering the EMP string? Two things: the carrier, because this revision does not itemise it, and the date code, because a part that sits at a distributor for three years is still within the moisture level 2 window as long as the bag is sealed, but the solderability of the tab is what fails first on a shelf. The date code management page sets out the checks we run on incoming reels for exactly this reason.
Is a 2500 piece reel the right buy for a build that uses 400 pieces a year? That is six years of parts on one reel, so the answer depends on the programme length rather than on price. Where the programme is short, a partial reel from a broker with a documented date code is the cheaper route, and where the programme is long, the full reel is the safer one because the parts stay in one lot.