Skip to main content
Modules / Tools / Power tree and budget calculator

Power tree and budget calculator

Build a board or system power tree of rails, regulators and loads, and see each rail's current, converter losses, efficiency and total input power.

Digital and physical designFree to useHow it worksSourcesWhere it helps

Fills the inputs below with typical figures. Replace any of them with your own.

Power tree: one entry for each rail or node
  1. Rail 1
  2. Rail 2
  3. Rail 3
  4. Rail 4
Source and time in each mode

The shares set the average power and the battery life. They are scaled to 100% if they do not add up.

Battery, optional

Leave the capacity at 0 if the system runs from a supply.

Allows for cold, ageing and pulse loads.

Converter temperature
Input power, active mode
42.48 mW
14.16 mA from the source, 86.2% efficient overall
Input power, idle
2.82 mW
100.0% efficient
Input power, sleep
0.0045 mW
100.0% efficient
Worst case
65.1 mW
21.7 mA with every rail at its maximum load
Average input power
0.05389 mW
0.002939 mW of it lost in converters
Battery life
365 days
472.5 mWh usable
Where the power is lost, active
  • VDD_RADIO3.476 mW59%
  • VDD_RF_CORE2.402 mW41%
  • VDD_SENSOR0.000094 mW0%
Show the data as a table
Where the power is lost, active
ItemLoss (mW)
VDD_RADIO3.476 mW
VDD_RF_CORE2.402 mW
VDD_SENSOR0.000094 mW
Input power by mode
  • Active42.48 mW
  • Idle2.82 mW
  • Sleep0.0045 mW
  • Worst case, maximum loads65.1 mW
Show the data as a table
Input power by mode
ItemInput power (mW)
Active42.48 mW
Idle2.82 mW
Sleep0.0045 mW
Worst case, maximum loads65.1 mW
Battery life as more time is spent asleep
Line chart of Battery life (days) against Time asleep (%), one line. The same numbers are in the table under the chart.05001,0001,5002,000020406080Time asleep (%)Battery life (days)
Show the data as a table
Battery life as more time is spent asleep
Time asleep (%)Battery life (days)
04.181
508.355
8020.83
9041.46
9582.13
98199.7
99382
99.92,139

Sign in to save your inputs and come back to them later.

Rails, active
RailTypeFed fromInput (V)Output (V)Own loadOutput currentInput currentInput powerLossEfficiencyConverter temperature (degrees C)
VDD_MCUDirect connectionSource333.2 mA3.2 mA3.2 mA9.6 mW0 mW100.0%not given
VDD_RADIOBuckSource31.811 mA15 mA10.16 mA30.48 mW3.476 mW88.6%not given
VDD_RF_CORELDOVDD_RADIO1.81.24 mA4 mA4.001 mA7.202 mW2.402 mW66.7%not given
VDD_SENSORLoad switchSource330.8 mA0.8 mA0.8 mA2.4 mW0.000094 mW100.0%not given
Every mode at a glance
ModeInput powerInput currentPower to the loadsConverter lossesOverall efficiency
Active42.48 mW14.16 mA36.6 mW5.878 mW86.2%
Idle2.82 mW0.94 mA2.82 mW0.00003016 mW100.0%
Sleep0.0045 mW0.0015 mA0.0045 mW0 mW100.0%
Worst case, maximum loads65.1 mW21.7 mA56.7 mW8.402 mW87.1%

How it works

A power tree shows how energy flows from the source through regulators to the loads. This tool works from the loads up: each rail's current, what its converter loses, and the current it asks of the rail above, until the source. It gives the input power in each use mode, the worst case, where the losses are, and how long a battery lasts.

  1. Set the voltages down the tree. A rail with no parent is fed from the source (or from its own input voltage if you enter one). Any other rail takes its input voltage from the output of its parent. Rails that feed each other in a loop cannot be solved and are left out with a warning.V in (child) = V out (parent)
  2. Find each rail's output current. A rail delivers its own load plus the current its child converters draw from it. Work from the leaves up. A load given in mW is turned into a current at the rail voltage. A load duty cycle scales the typical loads, and a rail that is off in a mode draws nothing in it, and neither do the rails below it.I out = I load x duty + sum of I in of the child converters; I load = P load / V out when the load is given in mW
  3. Convert through the converter. A linear regulator passes its output current straight through and burns the voltage difference, plus its quiescent current. A switching converter (buck, boost or a PMIC channel) takes in the output power divided by its efficiency, plus the quiescent power; the efficiency may be a fixed figure or a curve through a light-load point and a heavy-load point, joined against the logarithm of the current. A load switch or a direct connection loses only I^2 x R and the quiescent power.LDO: I in = I out + I q; loss = (V in − V out) x I out + V in x I q. Switcher: P in = P out / efficiency + V in x I q. Switch: loss = I^2 x R on + V in x I q
  4. Add up the tree. The total input power is the sum of the input power of the rails fed from the source. The overall efficiency is the power the loads receive divided by that input power. The loss of every converter is its input power minus the power it delivers, and this is the heat it must shed. With a thermal resistance entered, the converter's temperature follows.overall efficiency = sum of P load / P in total; T j = T a + loss x theta JA
  5. Run every mode and the worst case. The same sums run for the active, idle and sleep modes using each mode's loads, and for a worst case that uses the maximum load of every rail with every rail on. The checks flag a linear regulator whose input is less than its dropout above its output, a converter asked for more than its rating, and a converter that runs hot.dropout check: V in − V out >= dropout; rating check: I out <= I rated
  6. Average power and battery life. The shares of time in each mode weight the input power of each mode. A battery gives its capacity times its voltage times the usable share, and the life is that energy divided by the average power. The sweep shows how the life grows as more time is spent asleep.P avg = share active x P active + share idle x P idle + share sleep x P sleep; life = capacity x V battery x usable share / P avg

Assumptions

  • Loads are constant in each mode; the duty cycle scales the average current of a load that switches on and off. Peaks that are shorter than the converter's output capacitor can ride through are not modelled.
  • A switcher's efficiency excludes the quiescent current, which is added separately. Datasheet efficiency curves often include it, so for a light load enter the curve point without it or leave the quiescent current out.
  • A converter's efficiency does not change with its input voltage, and a child rail sees its parent's nominal output voltage.
  • Battery voltage and capacity are constant. The usable share covers temperature, ageing and pulse-load losses of capacity.
  • Loads, efficiencies and ratings in the presets are editorial examples.

Limits and warnings

  • A real converter's efficiency depends on input voltage, temperature, inductor and switching frequency. Use the datasheet curve at your conditions for the points you enter.
  • Inrush, start-up sequencing and brown-out behaviour are not part of a steady-state budget.
  • A battery's voltage falls through its life and its capacity falls at high current; for a coin cell with a pulsed radio, confirm the life with a measurement.
  • Several sources with different voltages are added as power. The total input current is a simple sum and is only meaningful when every top-level rail runs from the same source voltage.

Sources

The methods and default figures rest on these public sources. The wording here is our own; read the source for its full text.

Where it helps

The lifecycle stages this tool serves: 2 stages in 2 lifecycles.

Method notev1Formula, ranges, sources and limits

Works a power tree from the loads up to the source: each rail's current, each converter's loss, the input power in the active, idle and sleep modes and the worst case, the overall efficiency and the battery life for a mode mix.

Formula

V in (child) = V out (parent); I out = I load x duty + sum of I in of the child converters; I load = P load / V out when the load is given in mW; LDO: I in = I out + I q; loss = (V in − V out) x I out + V in x I q. Switcher: P in = P out / efficiency + V in x I q. Switch: loss = I^2 x R on + V in x I q; overall efficiency = sum of P load / P in total; T j = T a + loss x theta JA; dropout check: V in − V out >= dropout; rating check: I out <= I rated; P avg = share active x P active + share idle x P idle + share sleep x P sleep; life = capacity x V battery x usable share / P avg

Inputs
InputUnitRange it holds forComes from
Source voltageV0.5 to 60Your figure
Time in the active mode%0 to 100Your figure
Time in the idle mode%0 to 100Your figure
Time in the sleep mode%0 to 100Your figure
Battery capacity (0 for none)mAhfrom 0Your figure
Battery voltageV0.5 to 60Your figure
Usable share of the capacity%10 to 100Your figure
Ambient temperaturedegrees C-55 to 125Your figure
Converter temperature limitdegrees C60 to 175Your figure

Outputs: Input power, active mode (mW); Input power, sleep mode (mW); Input power, worst case (mW); Average input power (mW); Overall efficiency, active mode (fraction); Battery life (hours).

Assumptions

  • Loads are constant in each mode; the duty cycle scales the average current of a load that switches on and off. Peaks that are shorter than the converter's output capacitor can ride through are not modelled.
  • A switcher's efficiency excludes the quiescent current, which is added separately. Datasheet efficiency curves often include it, so for a light load enter the curve point without it or leave the quiescent current out.
  • A converter's efficiency does not change with its input voltage, and a child rail sees its parent's nominal output voltage.
  • Battery voltage and capacity are constant. The usable share covers temperature, ageing and pulse-load losses of capacity.
  • Loads, efficiencies and ratings in the presets are editorial examples.

Uncertainty: The arithmetic is exact; the result is as good as the loads, efficiencies and quiescent currents entered. With the editorial presets expect a real board to differ by tens of per cent, most of all in the sleep mode where tiny currents dominate.

Known limits

  • A real converter's efficiency depends on input voltage, temperature, inductor and switching frequency. Use the datasheet curve at your conditions for the points you enter.
  • Inrush, start-up sequencing and brown-out behaviour are not part of a steady-state budget.
  • A battery's voltage falls through its life and its capacity falls at high current; for a coin cell with a pulsed radio, confirm the life with a measurement.
  • Several sources with different voltages are added as power. The total input current is a simple sum and is only meaningful when every top-level rail runs from the same source voltage.

Where the method comes from

The Hub's own bottom-up power tree arithmetic: loads to converters to source, using the loss relationships for linear and switching regulators described in the cited Texas Instruments application reports.

Reference example, checked by test: Worked example in this method note: the IoT sensor preset (four rails on a 3 V coin cell), hand-worked in the tool's unit test. Input power, active mode 42.4775; Input power, sleep mode 0.0045; Input power, worst case 65.1024; Average input power 0.053892; Overall efficiency, active mode 0.861632; Battery life 8,767.57.

Versions

  • v1 (2026-10-25): First version, built from the tool idea.

All models in the register

Terms on this page

PMIC
An integrated circuit combining several regulators, references, sequencing and supervisory functions to generate all the rails a processor or system needs.
Datasheet
The authoritative customer-facing document giving a part's electrical characteristics, absolute maximum ratings, timing, pinout, package drawing and ordering information.

The full glossary

Need help acting on this?

Tell us what you're after and leave your email: we'll connect you with the right person on the Hub.