PowerElecHub

LDO Power Dissipation and Junction Temperature Calculator

Calculate the power a linear regulator dissipates, its junction temperature from θ_JA, the maximum output current the package allows, efficiency, dropout headroom at minimum input, and how much dissipation a series resistor can move out of the LDO without causing dropout.

Voltages and current

Thermal

Datasheet θ_JA is measured on a standard JEDEC board; your board's copper area can make the real value much higher or lower. A series resistor in front of the LDO takes part of the voltage drop and its heat; leave 0 for none.

LDO dissipation

2.53 W

at V_in = 13.2 V

Junction temperature

212 °C

limit 125 °C

Maximum current, thermal

124 mA

at V_in,max without R

Efficiency at V_in,max

37 %

46 % at V_in,min

  • At maximum input the junction reaches 212 °C, above the 125 °C limit. Reduce the voltage drop, add copper to lower θ_JA, share the loss with a series resistor, or use a switching regulator.

Other operating points

LDO dissipation at 10.8 V
1.79 W
Junction temperature at 10.8 V
168 °C
Resistor dissipation at V_in,max
0.00 W
LDO input at V_in,max / V_in,min
13.20 / 10.80 V

Limits

θ_JA needed for the limit
25.7 °C/W
Headroom above dropout at V_in,min
5.30 V
Largest R that keeps regulation
17.4 Ω

Junction temperature against output current

Junction temperature rises in a straight line with current, with a small offset from the ground current. The steeper line at maximum input is the one that sets the limit.

Steady-state DC model with a single θ_JA. Transient thermal capacity, θ_JA variation with board layout and airflow, and the temperature dependence of dropout and ground current are not included. Confirm with the case temperature on a prototype.

How it works

A linear regulator works by dropping the difference between input and output across its pass transistor, so the power it burns is simply that voltage difference times the load current, plus the input voltage times its own ground current. Nothing about the design changes that: a 13.2 V input regulated down to 5 V at 300 mA dissipates about 2.5 W whichever LDO you choose. The only questions are whether the package can shed that heat without the junction overheating, and whether a different arrangement would be better.

The junction temperature follows from T_J = T_A + P·θ_JA. The catch is θ_JA itself. Datasheets measure it on a standardized JEDEC test board, and the copper on your own board can make it considerably higher or lower; TI's notes on thermal metrics make the point that θ_JA is a property of the package and board together, not of the part alone. Use the datasheet number as a starting point and enter the value you expect for your layout.

The calculator also works the problem backwards. It gives the largest current the package can deliver at the highest input voltage before reaching the junction limit, and the θ_JA you would need for the current you want. If both look hopeless, one common fix is a resistor in series with the input. It takes part of the voltage drop and part of the heat, and the LDO only needs enough input to stay out of dropout at the lowest input voltage. That sets an upper limit on the resistor, R ≤ (V_in,min − V_out − V_DO)/I, which the page computes and checks.

Efficiency is reported for both input extremes. It is roughly V_out/V_in and no amount of thermal design improves it; a series resistor only moves the loss to a part that tolerates heat better. If the efficiency or the total heat is the real problem, a switching regulator is the answer.

Design procedure

  1. 1Take the highest and lowest input voltages, the output voltage and the maximum load current.
  2. 2Read the ground current and the dropout voltage at that current from the LDO datasheet.
  3. 3Estimate θ_JA for your package and copper area, and enter the highest ambient temperature.
  4. 4Check the junction temperature at the highest input. If it is too high, add copper, choose a larger package, or add a series resistor within the computed limit.
  5. 5Check the headroom at the lowest input, then confirm the case temperature on a prototype.

Formulas

LDO dissipation
P = (V_in − V_out)·I_out + V_in·I_GND
Junction temperature
T_J = T_A + P·θ_JA
Maximum current for T_J,max
I_max = ((T_J,max − T_A)/θ_JA − V_in·I_GND)/(V_in − V_out)
Series resistor
V_in,LDO = V_in − (I_out + I_GND)·R, R ≤ (V_in,min − V_out − V_DO)/(I_out + I_GND)
Efficiency
η = V_out·I_out/(V_in·(I_out + I_GND))

Worked example

The defaults describe a 5 V, 300 mA rail taken from a 12 V supply that can vary by ±10 %, in a package with an assumed θ_JA of 60 °C/W, a 60 °C ambient and a 125 °C junction limit. At 13.2 V the LDO dissipates 2.53 W and its junction would reach 212 °C, far above the limit. Even at 10.8 V it is 1.79 W and 168 °C. At the maximum input this package can only supply 124 mA, or it would need a θ_JA of about 26 °C/W, which takes a large copper area or a bigger package. The efficiency is only 37 % at 13.2 V. A series resistor of up to 17.4 Ω keeps the LDO out of dropout at 10.8 V. With 15 Ω, the LDO input falls to 8.6 V at maximum input, its dissipation to 1.13 W and its junction to 128 °C, still slightly over the limit, while the resistor takes 1.40 W and needs a 2 to 3 W rating. The input power, output power and both losses balance exactly. These figures were checked with a separate Python calculation that works from the energy balance.

Common mistakes

  • Checking only the nominal input voltage when the highest input sets the temperature.
  • Taking the datasheet θ_JA for granted on a board with much less copper than the JEDEC test board.
  • Adding a series resistor so large that the LDO drops out at minimum input and full load.
  • Forgetting the ground current of older bipolar LDOs, which can be several milliamps.
  • Expecting a better LDO to run cooler; the dissipation is set by voltage and current, not by the part.

Frequently asked questions

Why is my LDO hot even though the load current is small?

Because its dissipation is the voltage difference times the current. Dropping 8 V at 300 mA is 2.4 W, which is more than a small package can shed. The ground current adds V_in·I_GND on top, which matters at light load for older LDOs with milliamp ground currents.

Which θ_JA should I use?

Start with the datasheet value and correct it for your board. Datasheet θ_JA comes from a JEDEC standard test board with defined copper; a tiny board with little copper can be much worse, a large copper pour with thermal vias much better. Some datasheets give θ_JA for several copper areas or a curve, which is the better source. Confirm by measuring the case temperature on a prototype.

Does a series resistor really help?

It moves part of the dissipation from the LDO into the resistor, which can be a larger part spread elsewhere on the board. It does not reduce the total loss or improve efficiency. Its value is limited by dropout at the lowest input voltage and the highest current; above that limit the LDO stops regulating. The calculator gives that limit and the resulting temperatures.

When should I switch to a buck converter?

When the voltage difference times the current is more than the package can handle with realistic copper, or when efficiency matters for battery life or total heat. As a rough guide, if the LDO would dissipate more than about 1 W in a small package, or the efficiency is below about 50 %, a switching regulator is worth the extra parts.

How were the numbers checked?

A separate Python program computes the LDO loss as the input power into the LDO minus the output power, rather than with the voltage-difference formula, and the results match. The page also checks that input power equals output power plus the LDO and resistor losses.

References

  1. Texas Instruments, application report SLVA118, "Digital Designer's Guide to Linear Voltage Regulators and Thermal Management" (LDO power dissipation, junction temperature, θ_JA).
  2. Texas Instruments, application report SPRA953, "Semiconductor and IC Package Thermal Metrics" (what θ_JA means, why it depends on the board, and the ψ metrics).
  3. The values on this page were recomputed independently in Python.

Results are engineering estimates. Verify your design with datasheets, simulation and measurement.

Related calculators