Skip to content

Linear Voltage Regulator Selection Guide

Linear voltage regulator selection guide: how to choose between linear, LDO, and switching regulators by matching voltage, current, dropout, and thermal.

PinCompatible Engineering Team • • Updated: 10/7/2026
Linear voltage regulator selection guide comparing linear, LDO, and switching regulators
Linear voltage regulator selection guide comparing linear, LDO, and switching regulators

A linear voltage regulator selection guide is the decision reference engineers and buyers use when the board needs a clean, simple voltage rail and the choice sits between the regulator families. A linear voltage regulator is an integrated circuit that produces a fixed or adjustable output voltage by dissipating the difference between the input and the output as heat, and its simplicity, low noise, and low cost make it the default for a large share of power rails on every board. The selection decision is rarely about linear alone, because the designer also weighs the low-dropout regulator, the LDO, and the switching regulator, and the choice between the three is the first and most consequential decision of the rail's design. This guide explains how linear regulators work, how they compare to the LDO and the switching regulators, how to read the specifications, the voltage, the current, the dropout, the quiescent current, and the thermal, how to select the package and the family, and when the switching regulator is the better answer, so that a voltage rail is chosen on the evidence rather than on habit.

The reason a selection guide matters is that the linear regulator's weakness, its efficiency, is also the source of its strengths, and the designer has to match the trade to the application. A linear regulator converts the excess input-to-output voltage into heat, so the efficiency drops as the drop-out gap grows, and the thermal design becomes the limit of the current it can deliver. But the same simplicity delivers the cleanest output, the lowest noise, the fastest response, and the lowest cost, which is why the linear regulator still dominates the low-current and the noise-sensitive rails. The selection is a matching exercise: the current and the drop decide the heat, the heat decides the package, and the application decides whether the heat is acceptable. This guide walks that matching exercise end to end, so that the next rail on the board is chosen by the numbers, the family, the package, and the thermal budget, rather than by the part that happened to be in the drawer.

How a Linear Voltage Regulator Works

A linear voltage regulator produces a stable output voltage by passing the input current through a pass transistor and dissipating the difference between the input and the output as heat, and understanding that one fact explains every selection trade in this guide. The regulator senses the output, compares it to an internal reference, and adjusts the pass transistor's conduction to hold the output at the set voltage, and the current that flows through the pass element is the same current the load draws, so the power dissipated in the regulator is the input-to-output voltage difference multiplied by the load current. The efficiency is therefore the output voltage divided by the input voltage, roughly, and it drops as the drop between the input and the output grows. The dissipated power appears as heat in the package, and the package and the board must move that heat away or the regulator's junction temperature exceeds its rating. The simplicity of the linear regulator, one pass element and a control loop, delivers the low noise, the low cost, and the fast response that the more complex switching regulator cannot match, and the heat is the price of that simplicity.

The internal architecture also explains the two families that bracket the linear regulator: the standard linear regulator and the low-dropout regulator. The standard linear regulator, the 78xx family being the classic, uses a pass transistor arrangement that requires a minimum input-to-output difference, the dropout voltage, of roughly 2 volts for the traditional parts, so a 5-volt 7805 needs about 7 volts in to regulate cleanly. The LDO, the low-dropout regulator like the AMS1117 and the LM1117, uses a different pass element arrangement that drops the minimum difference to a few hundred millivolts, so a 3.3-volt LDO can regulate from a 3.6-volt or a 5-volt input with little loss. The dropout voltage is the specification that separates the families, and it is the first number the selection reads, because it decides the minimum input the rail can tolerate and the heat the rail will make. The three families, linear, LDO, and switching, are the selection's map, and the dropout is the boundary between the first two.

What the dropout voltage means when selecting a regulator

The dropout voltage is the smallest difference between the input and the output at which the regulator still holds its specified output, and it is the number that decides how much headroom the rail needs. A traditional linear regulator with a 2-volt dropout needs the input to sit at least 2 volts above the output at all times, including during the input ripple and the load transients, so a 5-volt rail needs the input held above about 7 volts. An LDO with a 0.3-volt dropout can regulate a 3.3-volt output from a 3.6-volt input, which is why the LDO is the part of choice for the battery rails and the rails that run close to the input. The dropout also shapes the heat, because the dropout is the minimum waste and the actual input-to-output difference is the real waste, and a rail with a 2-volt dropout running from a 12-volt input wastes far more than the minimum. The selection reads the dropout against the input source, the battery, the adapter, the previous rail, and the load's transient, and it chooses the family whose dropout fits the headroom and the heat the rail can tolerate.

Linear vs LDO vs Switching Regulators

The selection between the linear, the LDO, and the switching regulators is the core decision of the rail, and the three families map to the application's priorities. The standard linear regulator wins on simplicity, cost, and noise, and it is the choice for the small-current rails and the rails where the input-to-output drop is small, because the heat is manageable and the circuit is one part and two capacitors. The LDO wins on the low headroom and the low drop, and it is the choice for the rails that run close to their input, the battery rails and the rails derived from a higher rail with little margin, where the switching regulator is overkill and the standard linear regulator cannot fit. The switching regulator wins on efficiency, and it is the choice for the rails with a large input-to-output difference or a high current, where the linear and the LDO would dissipate heat that no package and no board could remove. The selection is a triangle of noise, simplicity, and efficiency, and the application places the rail at the vertex it can afford.

The comparison is best read as a table of the three families across the dimensions that matter. The efficiency of the linear and the LDO is the output divided by the input, and it collapses at large drops, while the switching regulator's efficiency holds at 80 to 95 percent across the range. The output noise of the linear and the LDO is low, microvolts of ripple, while the switching regulator adds the switching ripple and the EMI. The cost and the simplicity favor the linear and the LDO, one part and two caps, against the switching regulator's inductor, diode, feedback, and layout care. The response favors the linear, whose control loop is fast, against the switching regulator's slower loop. The choice follows the numbers: the small-current and the low-drop rail takes the linear or the LDO, and the high-current or the large-drop rail takes the switching regulator, and the guide's remaining sections read the specifications that make the choice precise.

Dimension Linear regulator LDO regulator Switching regulator
Efficiency Output/input, falls at large drop Output/input, best at low drop 80–95%, holds across range
Output noise Low, no switching ripple Low, no switching ripple Switching ripple and EMI
Cost and parts One part, two capacitors One part, two capacitors Inductor, diode, feedback, layout
Dropout needed ~2 V for classic 78xx ~0.1–0.6 V N/A, boost/buck architecture
Best use Small current, small drop, noise-sensitive Low headroom, battery rails High current, large input-output difference

Key Specifications When Selecting a Linear Regulator

The linear regulator selection reads a set of specifications that each decides a part of the design, and knowing them prevents the failures that the headline voltage and current hide. The output voltage is the first and the simplest, fixed for the 78xx and the AMS1117 variants or adjustable for the LM317 and the adjustable LDOs, and it must match the rail's target within the tolerance the load allows. The output current is the next, and it must cover the load's worst case plus the headroom for the transients and the capacitor charge, because a regulator run at its current limit collapses under the surge. The dropout voltage, read as a specification, decides the minimum input and the heat, as the earlier section described. The quiescent current is the current the regulator draws to run itself, and it matters for the battery rails where every microamp counts and for the always-on rails. The line and the load regulation state how well the output holds against the input and the load changes, and the temperature coefficient states the drift of the output across the range. The specifications are the datasheet's answer to the rail's requirements, and the selection reads each against the design.

The selection also reads the ratings that decide the part's survival, and they deserve their own attention. The absolute maximum input voltage decides the surge and the reverse conditions the regulator must survive, and the input must be protected with the capacitors and the protection diodes the datasheet requires. The junction temperature rating and the thermal resistance decide the heat the package can move, and the junction temperature equals the ambient plus the dissipated power times the thermal resistance, a number that must stay below the rating. The short-circuit and the thermal protection decide the behavior under the fault, the foldback and the shutdown that the classic regulators provide, and the current-limit behavior the design should tolerate. The package's thermal resistance, the junction-to-ambient and the junction-to-case values, decides the heatsink and the board copper the design needs. The ratings are the layer of the selection that the headline numbers hide, and the engineer who reads them selects the part that survives, not just the part that regulates.

How to read the thermal specification of a linear regulator

Reading the thermal specification is the calculation that decides whether a linear regulator is feasible at the current the design needs. Compute the dissipated power, the input-to-output difference times the output current, and add the quiescent power, and compute the junction temperature, the ambient temperature plus the dissipated power times the junction-to-ambient thermal resistance. Compare the junction temperature to the part's maximum junction rating, and when it exceeds the rating, the options are the lower current, the larger package, the heatsink, the board copper, the lower ambient, or the switching regulator. The junction-to-ambient resistance is the package's ability to shed heat into the still air, and the junction-to-case resistance plus the case-to-heatsink resistance model the heatsink path for the TO-220 and the power packages. The calculation is the gate of the selection, and it is the reason the high-current and the large-drop rails leave the linear family for the switching family, because no package can shed the watts. The thermal read is the discipline that turns a regulator selection from a voltage and current choice into a thermal design.

Selecting by Voltage, Current, and Dropout

The selection by the voltage, the current, and the dropout is the practical heart of the method, and it runs in a clear order. Set the output voltage first, the fixed value for the fixed parts or the resistor network for the adjustable parts, and confirm the tolerance against the load. Set the output current second, the worst-case load plus the headroom, and confirm the part's rating and its thermal feasibility at that current and the input-to-output drop. Set the dropout third, the minimum input the rail must tolerate, the battery's low cell, the adapter's sag, and the previous rail's tolerance, and choose the family whose dropout fits, the standard linear for the generous headroom and the LDO for the tight. The three settings narrow the field to the family and the rating, and the remaining specifications, the noise, the quiescent current, and the protection, decide the specific part. The selection order is the method's skeleton, and the rest of the specifications hang on it.

The selection also reads the application's special requirements, because the voltage and the current do not capture every rail. A rail for the analog front end reads the noise and the PSRR, a rail for the wireless module reads the ripple and the transient response, a rail for the memory or the logic reads the tolerance and the sequencing, and a rail for the battery reads the quiescent current and the dropout at the low cell. The application requirements are the tiebreakers that select between the parts that pass the voltage and the current filters, and they are the reason two designs with the same 3.3-volt rail can legitimately select different parts. The selection method is therefore a funnel, the voltage and the current and the dropout narrow the field, and the application's special requirements pick the winner, and the funnel is the discipline that keeps the selection grounded in the design rather than in the habit.

How to choose a fixed versus adjustable linear regulator

The fixed-versus-adjustable choice is a small decision with a clear rule. Choose the fixed regulator when the rail is a standard voltage, the 5-volt, the 3.3-volt, the 2.5-volt, and the like, and the design needs the fewest parts and the tightest startup, because the fixed part is one component and its internal reference and divider are trimmed at the factory. Choose the adjustable regulator, the LM317 and the adjustable LDOs, when the rail is a non-standard voltage, when the design needs a value that the fixed catalog does not offer, or when the same part must serve several rails on the board by changing a resistor, because the adjustable part turns one device into a family of outputs. The adjustable part carries the cost of the external resistor divider and its tolerance, which adds to the output tolerance, and it carries the risk of the resistor drift and the layout noise on the adjust pin. The rule is to prefer the fixed part for the standard rails and the adjustable part for the non-standard and the shared rails, and the choice is the small print of the selection that the voltage requirement decides.

Thermal Design and Package Selection

The package selection is the thermal design made physical, and it is the half of the selection that most failures come back to. The package's thermal resistance, the junction-to-ambient and the junction-to-case values, decides how much heat the part can shed, and the package choice follows the dissipated power: the small SOT-23 and the SOT-89 for the milliwatt rails, the SOT-223 and the SO-8 and the DPAK for the watt-range, and the TO-220 and the larger power packages for the watts that need the heatsink. The board is the second heatsink, and the copper area and the vias under the thermal pad carry the heat away, so the layout of the thermal pad is part of the thermal design. The ambient temperature and the enclosure complete the picture, because the junction temperature is the ambient plus the watts times the resistance, and the enclosure that traps the heat raises the ambient the junction sees. The package and the board and the enclosure are the thermal system, and the selection sizes the system to the dissipated power the rail demands.

The thermal design also reads the load's duty and the worst case, because the heat follows the current and the current follows the application. A rail that runs at full current continuously needs the full thermal analysis, and a rail that runs at full current briefly and idles is served by the average power and the thermal time constant of the board. The worst case, the maximum input voltage, the maximum load, and the maximum ambient, is the condition the thermal design must survive, and the selection uses the worst case rather than the typical. The thermal design is the gate that the linear family respects, and the rail whose watts exceed the package's ability to shed them is the rail that belongs to the switching family, whatever the noise and the simplicity suggest. The thermal read is the honest judge of the selection, and it is the reason the guide's earlier comparison left the high-power rails to the switching regulator.

How to pick a regulator package for the power the rail needs

Picking the package for the power the rail needs is a calculation and a map. Compute the dissipated power at the worst case, and read the package's thermal resistance against the ambient to find the junction temperature, then match the power to the package's ability to shed it. For the sub-watt rails, the SOT-23 and the SOT-89 shed the milliwatts with the board copper. For the one-to-two-watt rails, the SOT-223 and the SO-8 and the DPAK carry the watts with the generous board copper and the thermal vias. For the higher watts, the TO-220 and the DPAK with a heatsink or the large copper area are the answer, and the heatsink's case-to-ambient resistance adds to the model. The rule is to leave headroom, because the actual board, the enclosure, and the airflow differ from the datasheet's free-air numbers, and the part that fits at the limit in the datasheet overheats in the product. The package selection is the thermal design's physical decision, and the calculation and the headroom are what keep the junction inside its rating across the product's life.

Linear Regulator Families: 78xx, LM317, LM1117, and Their Replacements

The linear regulator families are the working vocabulary of the selection, and knowing them makes the replacement and the sourcing of the selected part straightforward. The 78xx family, the 7805, the 7812, and the rest, is the classic fixed positive regulator with a generous 2-volt dropout, a TO-220 and SO-8 and a range of packages, and a pinout so standardized that the family is pin-compatible across the vendors, the LM7805, the L7805, the KA7805, and the field of domestic parts sharing the three-pin layout. The LM317 is the classic adjustable regulator, its output set by two resistors, and it is also standardized across the vendors with the same pinout in the TO-220 and the SO-8. The LM1117 and the AMS1117 are the LDO family, the 3.3 and the 5-volt low-dropout parts that regulate from a few hundred millivolts of headroom, and they share a pinout that the market reproduces across the vendors. The families are the catalog the selection draws from, and the pin compatibility across the vendors is the reason a selected part is also a replaceable part.

The replacement angle is worth making explicit, because the selection of a linear regulator is also the selection of a part that will be sourced and eventually substituted. The 78xx family's pin compatibility means a board designed around an LM7805 can be supplied by an L7805, a KA7805, or a domestic 7805, on the same footprint and the same test, which is the classic example of the pin-compatible replacement this site's catalog trades in. The AMS1117 and the LM1117 share the LDO pinout, and the LM317 family shares the adjustable pinout, so each family's selection opens the family's replacement market. The guide's selection method therefore ends where the sourcing begins, the selected family, the pin compatibility, and the replacement options, and it is the reason the linear regulator is one of the friendliest parts in the catalog to both select and to replace. The families are the vocabulary, and the pin compatibility is the grammar, and the selection and the replacement speak the same language.

How to substitute a 78xx or LDO regulator safely

Substituting a 78xx or an LDO regulator safely is the pin-compatible method applied to the family. Confirm the package and the pinout first, the TO-220 and the SO-8 and the SOT-223 variants are different footprints, and the pin assignment must match the board. Confirm the output voltage and the current second, the 7805 replacement must be a 5-volt part rated for the load, and the LDO replacement must match the output and the dropout the rail needs. Confirm the dropout and the thermal third, because a standard linear replacement for an LDO rail cannot regulate from the LDO's tight headroom, and a part with a worse thermal package overheats at the current the rail draws. Confirm the input range and the protection fourth, the absolute maximum and the reverse protection matching the board's conditions. And test the replacement on the board, the output accuracy, the load step, and the thermal across the temperature, and record the comparison. The substitution is the selection method run on the family, and the discipline is what keeps a pin-compatible replacement quiet, which is the standard this site's catalog applies to every part it ships.

Noise, PSRR, and the Precision Applications

The linear regulator's low noise is its main advantage over the switching regulator, and the precision and the analog applications are where that advantage decides the selection. The output noise of a linear regulator is specified as the spectral density and the total integrated noise, and it matters for the analog front ends, the audio paths, the sensor rails, and the reference rails, where the supply noise couples into the signal. The power supply rejection ratio, the PSRR, states how well the regulator attenuates the noise and the ripple on its input, and it matters for the rails derived from a noisy source, a switching rail, or a mains adapter. The precision applications select the low-noise and the high-PSRR parts, and the selection reads the noise and the PSRR curves against the signal's requirement rather than against the datasheet's single number. The noise and the PSRR are the specifications that the headline voltage and current do not capture, and they are the reason two rails with the same output select different parts.

The noise and the PSRR also connect the selection to the rest of the board, because the rail's cleanliness is a system property rather than a part property. The input capacitor and the output capacitor shape the noise and the stability, the bypassing of the load matters, and the ground layout couples or isolates the noise. The selection therefore pairs the low-noise part with the decoupling and the layout the application needs, and the datasheet's application section is the guide to the capacitor values and the layout. The precision rail is a system, and the selection is the first element of the system rather than the whole of it. The noise and the PSRR are the linear regulator's reason to exist, and the application that needs them is the application that the linear family, and not the switching family, is selected to serve.

How to read the PSRR curve of a linear regulator

Reading the PSRR curve is reading how the regulator rejects the input noise across the frequency, and it is the number that the precision design actually uses. The PSRR is plotted as the rejection in decibels against the frequency, and it is high at the low frequencies, tens of decibels, and it falls as the frequency rises, with the curve shaped by the regulator's control loop and its internal compensation. The design reads the PSRR at the frequencies where the input noise lives, the switching frequency of the previous rail, the mains and its harmonics, and the broadband noise, and it compares the rejection at those frequencies against the signal's tolerance. A regulator with a high low-frequency PSRR but a drooping high-frequency curve rejects the 100-hertz ripple well and the megahertz switching ripple poorly, and the design pairs it with the filtering that covers the frequencies it rejects weakly. The PSRR curve is the precision selection's instrument, and it is the reason the low-noise rail is selected by the curve rather than by the headline number.

When a Switching Regulator Is the Right Choice

The linear and the LDO regulators have their limits, and the switching regulator is the right choice where the limits bite. Choose the switching regulator when the current is high, when the input-to-output difference is large, or when the efficiency and the heat matter more than the noise, because the linear family dissipates the excess as heat that no package can shed and the switching family converts it into output with an 80-to-95-percent efficiency. Choose the switching regulator when the rail must step down from a high bus, a 24-volt or a 48-volt rail, to a low logic rail, where the linear waste would be catastrophic. Choose the switching regulator when the board's thermal budget is already spent, and when the battery life depends on the conversion efficiency rather than on the quiescent current. The switching regulator pays for its efficiency with the noise, the ripple, the EMI, the cost, and the layout care, and the design that needs the efficiency accepts the cost. The selection between the linear and the switching families is the selection between the clean and simple rail and the efficient rail, and the numbers decide which the application can afford.

The switching choice also respects the hybrid designs, because the two families work together rather than exclusively. A board that uses a switching regulator to step the high bus down to an intermediate rail and a linear or an LDO to clean that rail for the analog and the sensitive loads gets the efficiency of the switch and the noise of the linear in the same design, and the cascade is one of the most common power architectures in electronics. The cascade reads the switching ripple at the linear's PSRR, and it selects the linear or the LDO whose rejection covers the switch's frequency. The selection method therefore chooses the family per rail, the switch for the high-power step and the linear for the clean rail, and the cascade is the system that the two families build together. The linear regulator selection guide is the half of the power design, and the switching choice is the other half, and the two halves meet in the cascade.

Worked Example: Selecting a Rail for an Industrial Control Board

Selecting a rail for an industrial control board is the worked example that shows the whole selection on a real design, and a board that runs from a 24-volt industrial supply and needs a 5-volt logic rail and a clean 3.3-volt rail for an analog section is the case. The first rail, the 5-volt logic, comes from the 24-volt bus through a switching regulator, because the 19-volt drop at the logic current would dissipate watts that no linear package could shed, and the switching regulator steps the 24 volts down to a 5-volt intermediate rail with the 80-percent-class efficiency. The second rail, the 3.3-volt clean rail, comes from the 5-volt intermediate through an LDO, because the 1.7-volt drop at the analog current is a small wattage that the LDO sheds, and the LDO's low dropout and its low noise give the analog section the clean supply that the switching rail cannot. The selection reads the family per rail, the switching for the high-drop and the high-current step and the LDO for the clean low-drop step, and the cascade is the architecture the switching section described. The worked example shows the selection as a per-rail decision, and it is the decision the workflow formalizes.

The worked example also reads the checks the industrial board adds, because the industrial environment changes the selection's weights. The 24-volt bus's transients and the reverse conditions read the input protection, the input capacitor and the TVS that the datasheet recommends, and the temperature range, the industrial ambient, reads the thermal calculation at the worst case rather than at the bench. The analog section's precision reads the noise and the PSRR of the LDO, and the digital section's rail reads the tolerance and the load. The industrial board is the model of the selection's real weights, and it is the model that the workflow and the next section's checklist turn into a procedure, which is the practical completion of the linear regulator selection guide.

A Workflow for Selecting a Linear Voltage Regulator

The full selection method fits into a workflow that an engineer runs in the same order. First, set the rail's requirements, the output voltage, the worst-case current, the input range and its sag, the tolerance, the noise, and the thermal budget. Second, choose the family, the linear for the small-current and the low-drop and the noise-sensitive rail, the LDO for the tight-headroom rail, and the switching for the high-current and the large-drop rail, and record the choice. Third, read the specifications, the dropout, the quiescent current, the regulation, the noise, the PSRR, and the protection, against the rail's requirements. Fourth, calculate the thermal, the dissipated power at the worst case, the junction temperature, and the package that sheds it, and choose the package and the heatsink. Fifth, select the part from the family, the fixed or the adjustable, the package, and the vendor, and confirm the pin compatibility and the replacement options for the sourcing. Sixth, verify on the board, the output accuracy, the load step, the transient, the noise, and the thermal across the temperature, and record the measurements. That sequence is the linear voltage regulator selection method, and it is the same method whether the rail is a 3.3-volt logic rail or a 5-volt analog rail.

The workflow is the reason a voltage rail is chosen on the evidence rather than on the habit. With the requirements set, the rail is defined. With the family chosen, the trade is explicit. With the specifications read, the part fits the design. And with the thermal calculated and the part verified, the rail survives the product's life. The entire value of the method is that the linear regulator, the simplest power part on the board, is selected with the same care as the most complex, because the simple part's heat and its dropout and its noise carry the rail's reliability, and the discipline of the selection is what keeps the board running. An engineer who runs this workflow selects a linear voltage regulator with a family choice, a specification read, and a thermal calculation, which is the entire point of a linear voltage regulator selection done well.

Next Step

Specify your hotel project with our engineers

Send your room count, ceiling type, and protocol preference. We will return a sample plan and quote within 24 business hours.

  • Move from general guidance into a product or application discussion.
  • Use RFQ when pricing, drawings, MOQ, or launch timing needs structure.
  • Keep a direct contact path visible for fast clarifications and handoff.
Ready for RFQ

Share your product requirements and get a practical next step

Send your drawings, target quantity, and timeline. Our sales engineering team will respond within 24 business hours with a practical next step, a quote, or a sample plan.

Send a quick inquiry

Tell us the part numbers, quantities and timeline — we reply within 24 business hours.

A clear list of part numbers and quantities lets the team reply within one business day with matched alternatives.