Electrical Trade & Electronics DC Circuit Fundamentals Kirchhoff's current law and Ohm's law

Current Divider Calculator

Enter the current arriving at a parallel node and the resistance of each branch, and this calculator shows how that current splits. It returns the current in every branch, the equivalent resistance of the group, the voltage across all of them and the power each resistor dissipates. Use it to size an ammeter shunt, to check that a parallel load bank shares current the way you intend, or to work a circuit-theory problem and follow each arithmetic step. Set a branch to zero to remove it, or to a very large resistance to model a leakage path.

Calculator

This calculator runs in your browser. Enable JavaScript for live results — the inputs, formula and worked example below remain fully readable without it.

Inputs this calculator takes, with typical values
InputWhat to enterExample
Total current into the nodeThe current arriving at the parallel junction, from a source or an upstream series element.6 A
Branch 1 resistance (R1)Resistance of the first parallel branch, measured or read from the colour code.2 Ω
Branch 2 resistance (R2)Resistance of the second parallel branch. Enter 0 if this branch does not exist.3 Ω
Branch 3 resistance (R3)Resistance of the third parallel branch. Enter 0 if you only have two branches.6 Ω

It returns

  • Current in branch 1 — I1 = I x Req / R1.
  • Current in branch 2
  • Current in branch 3
  • Equivalent resistance
  • Voltage across the parallel group

The formula

I1=IReqR1
1Req=1R1+1R2+1R3
I1=IR2R1+R2

In plain text: I₁ = I × R_eq / R₁, where 1/R_eq = 1/R₁ + 1/R₂ + 1/R₃

  • I₁Current in branch 1 (A)
  • ITotal current arriving at the parallel node (A)
  • R₁Resistance of branch 1 (Ω)
  • R_eqEquivalent resistance of all branches in parallel (Ω)

The rule follows from Kirchhoff's current law and Ohm's law: every branch shares one node voltage V = I x R_eq, so each branch current is that voltage divided by the branch resistance.

Updated Category DC Circuit Fundamentals Verified against published test cases Reading time 12 min

What a current divider is and when you need one

A current divider is any group of components wired between the same two nodes. Current arriving at the first node has to leave through one branch or another, and how much goes down each path is fixed entirely by the branch resistances. You reach for the divider rule whenever you know the total current but need a single branch current: sizing an ammeter shunt, working out how much fault current returns through a parallel ground path, checking that four paralleled power resistors in a dummy load share evenly, or answering a circuit-analysis question that hands you the source current rather than the source voltage.

The physical statement behind it is Kirchhoff's current law: charge does not pile up at a node, so the currents leaving equal the current arriving. The second statement is that parallel elements share one voltage. Put those together and each branch current is that shared voltage divided by that branch's resistance. Nothing else enters the calculation — not the source voltage, not what sits upstream, not the wire between the node and the resistor, provided that wire's resistance is small compared with the branch it feeds.

The dual of this problem is the voltage divider, which splits a voltage between series elements. Series and parallel are mirror images: in series the current is common and the voltage splits in proportion to resistance; in parallel the voltage is common and the current splits in inverse proportion to resistance. If you are combining resistances rather than splitting current, use the parallel resistance calculator or the series resistance calculator.

The formula explained, term by term

Start with what the branches have in common. Because every branch spans the same pair of nodes, they all see one voltage V. The group as a whole behaves like a single resistor of value R_eq, defined by adding conductances rather than resistances:

1/R_eq = 1/R₁ + 1/R₂ + 1/R₃

Conductance, measured in siemens, is the reciprocal of resistance, and it is the natural quantity here because conductances in parallel simply add — each extra branch opens another path. The node voltage then follows from Ohm's law applied to the group: V = I × R_eq. Ohm's law applied to one branch gives that branch's current:

I₁ = V / R₁ = I × R_eq / R₁

Read that last expression as a fraction of the total. The factor R_eq / R₁ lies between 0 and 1, because R_eq is always smaller than the smallest branch resistance. Add the factors for all branches and you get exactly 1, which is Kirchhoff's current law reappearing as a bookkeeping check.

With only two branches the algebra collapses to the form most people memorise. Substituting R_eq = R₁R₂/(R₁+R₂) gives I₁ = I × R₂/(R₁+R₂). Notice the opposite resistance sits on top. That inversion is the most common slip in the whole topic: in a voltage divider the resistor you are asking about appears in the numerator, and in a current divider it does not.

Everything above assumes pure resistance. At DC, or at frequencies low enough that reactance is negligible, that is exactly right. When branch impedance is complex you replace each R with an impedance and do the arithmetic with phasors — see the RLC impedance calculator for how magnitude and angle combine.

Worked example: 6 A into 2 Ω, 3 Ω and 6 Ω

A current source pushes 6 A into a node feeding three parallel resistors of 2 Ω, 3 Ω and 6 Ω. Find each branch current.

  1. Add the conductances. 1/2 + 1/3 + 1/6 = 0.500000 + 0.333333 + 0.166667 = 1.000000 S.
  2. Invert to get the equivalent resistance. R_eq = 1 / 1.000000 = 1.0000 Ω. It is smaller than the smallest branch, 2 Ω, as it must be.
  3. Find the node voltage. V = I × R_eq = 6 × 1.0000 = 6.0000 V. Every resistor has 6 V across it.
  4. Divide that voltage by each resistance. I₁ = 6 / 2 = 3.0000 A. I₂ = 6 / 3 = 2.0000 A. I₃ = 6 / 6 = 1.0000 A.
  5. Check the sum. 3 + 2 + 1 = 6 A, the current you started with.

The shares are 3/6 = 50.00%, 2/6 = 33.33% and 1/6 = 16.67%. They are exactly the conductance shares 0.5/1, 0.3333/1 and 0.1667/1. Power follows from V²/R: 36/2 = 18 W, 36/3 = 12 W and 36/6 = 6 W, totalling 36 W — the same as I × V = 6 × 6. Those three resistors need serious wattage ratings, which the calculator flags for you.

Now the two-branch shortcut on the same numbers. Delete the 6 Ω branch and keep 2 Ω and 3 Ω with 6 A arriving. I₁ = 6 × 3/(2+3) = 6 × 0.6 = 3.6 A, and I₂ = 6 × 2/5 = 2.4 A. The 2 Ω branch takes more, because for one shared voltage the smaller resistance passes the larger current. Removing the third path also raises R_eq from 1.0000 Ω to 2×3/5 = 1.2000 Ω, and the node voltage rises with it to 6 × 1.2 = 7.2 V.

How to read the result

Look first at the share each branch takes. A branch's share of the total equals its share of the total conductance, so a branch of half the resistance takes twice the current of the branch it is compared with. If your design intends four paralleled resistors to share a load equally, a share that departs from 25% by more than the resistor tolerance points at a wiring or measurement problem rather than a rounding one — in a 100 W load bank, four watts of imbalance is the difference between a part that survives and one that does not.

Next look at the equivalent resistance. It is always lower than the smallest branch, and adding a branch always lowers it further. That matters upstream: if the parallel group is fed from a real source rather than an ideal current source, a lower R_eq changes the loop current, and the node voltage reported here assumes the stated total current genuinely arrives.

Then look at the power column. Dissipation is why most current-divider designs fail in hardware rather than on paper. A shunt carrying 9.95 A through 0.005 Ω dissipates 9.95² × 0.005 = 0.495 W — modest, but its resistance drifts as it warms, and that drift changes the split. Metal-foil and manganin shunts exist precisely because their temperature coefficient is small. For the resistor-heating side of the problem, the electrical power calculator works the same numbers from any two of volts, amps, ohms and watts.

Finally, sanity-check the whole group with Ohm's law: V divided by R_eq must return the total current you entered. If you would rather solve the same circuit from a known source voltage, the Ohm's law calculator gets you the total current first.

How two parallel resistors split a current

Branch shares for a two-resistor divider with R₁ = 100 Ω. The share taken by R₁ is R₂/(R₁+R₂), and the equivalent resistance is R₁R₂/(R₁+R₂).
R₂ (Ω)R₂ / R₁Share in R₁Share in R₂R_eq (Ω)
100.19.09%90.91%9.09
250.2520.00%80.00%20.00
500.533.33%66.67%33.33
1001.050.00%50.00%50.00
2002.066.67%33.33%66.67
4004.080.00%20.00%80.00
100010.090.91%9.09%90.91

The share column and the R_eq column carry the same digits because R_eq = R₁ × R₂/(R₁+R₂) and R₁ is 100 Ω in every row. The pattern is scale-free: only the ratio R₂/R₁ sets the split.

Mistakes that put the current in the wrong branch

  • Putting the branch's own resistance on top. For two branches the numerator is the other resistance: I₁ = I × R₂/(R₁+R₂), not R₁/(R₁+R₂). Reversing it produces an answer that is still plausible, which is why the error survives review.
  • Extending the two-resistor shortcut to three branches. R₂/(R₁+R₂+R₃) is not the share of a three-way divider. Go back to I × R_eq/R₁, which is correct for any number of branches.
  • Ignoring the resistance of wiring and connections. When branch resistances are milliohms — shunts, busbars, paralleled battery cables — a few milliohms of lug and cable resistance is a large fraction of the branch and shifts the split badly. Enter the measured branch resistance including its leads.
  • Forgetting that resistance moves with temperature. Copper changes about 0.39% per °C. Two nominally identical paralleled paths at different temperatures do not share equally, and for copper the hotter path takes the smaller share, which is self-stabilising; for materials with a negative coefficient it is not.
  • Applying the DC rule to an AC circuit with reactance. With inductance or capacitance present the branch currents differ in phase, so their magnitudes do not add arithmetically and the resistive ratio no longer predicts the split.
  • Assuming the source really delivers the stated current. A voltage source behind a source resistance is not a current source. Adding a branch lowers R_eq, which changes the node voltage and the total current — solve the loop first, then divide.

Key terms

Conductance (G)
The reciprocal of resistance, measured in siemens (S). Conductances in parallel add directly, which is why the divider formula is cleaner written in terms of G.
Equivalent resistance (R_eq)
The single resistance that would draw the same current from the same node voltage as the whole parallel group. Always smaller than the smallest branch.
Shunt
A low-value precision resistor deliberately placed in parallel with a meter movement or sense circuit so that a known small fraction of the current flows through the measuring element.
Kirchhoff's current law
The sum of currents entering a node equals the sum leaving it. It is conservation of charge, and it is why the branch shares total 100%.

Where the current divider sits among the other tools

The divider rule is a shortcut, not a separate law. Any circuit it solves can also be solved by node-voltage analysis, and for anything more tangled than one parallel group — bridges, multiple sources, dependent sources — node analysis is the method that scales. The value of the divider rule is that it removes an unknown: you never have to find the source voltage to get a branch current.

In instrumentation the rule is the design equation for an ammeter shunt. A moving-coil movement that reads full scale at 50 mA across 1 Ω becomes a 10 A ammeter when you parallel it with a shunt sized so that the rest bypasses the movement. Rearranged, the shunt resistance is R_shunt = R_movement × I_movement / (I_total − I_movement), which is the two-branch divider solved for a resistance instead of a current. Modern current-sense amplifiers use the same idea with a much smaller shunt and gain applied afterwards.

In power work the same arithmetic explains paralleled conductors and paralleled transformers. The National Electrical Code permits paralleled conductors only under strict conditions on length, size, material and termination, and the physical reason is this formula: unequal branch impedance means unequal current, so one conductor of a parallel set can be overloaded while the set as a whole looks fine. If you are sizing those conductors, start with the wire size and ampacity calculator and then confirm the split here.

The rule also governs unintended paths. Current returning through a grounding conductor divides between that conductor and every other metallic path in parallel with it — conduit, water pipe, building steel. Objectionable current on a water pipe is a current divider working exactly as physics requires, on a circuit nobody drew.

Frequently asked questions

Why is the other resistor on top in the two-resistor formula?

Because each branch current is the shared node voltage divided by that branch's resistance, so a branch's share rises as its own resistance falls. When you write the two-branch case out, I₁ = I × R_eq/R₁ with R_eq = R₁R₂/(R₁+R₂), the R₁ cancels and leaves R₂ on top. It is algebra, not a special rule to memorise, and it is the opposite of the voltage divider.

How do I model a branch that is not there?

Enter 0 for that branch and the calculator drops it from the parallel group and reports its current as a dash. If you would rather see the effect of a very high but finite path — board leakage, a voltmeter across the node — enter the real value, for example 10 MΩ. Any branch more than about a thousand times the smallest one carries under 0.1% of the total, and the calculator says so.

Do the branch currents always add up to the total?

Yes, exactly, for any number of purely resistive branches. That is Kirchhoff's current law. If you sum the reported branch currents and do not recover the total you entered, the discrepancy is display rounding rather than physics. In an AC circuit with reactance the magnitudes no longer add, because the branch currents are out of phase and must be summed as phasors.

What size shunt do I need to extend an ammeter range?

Rearrange the two-branch divider: R_shunt = R_meter × I_meter / (I_total − I_meter). A 1 Ω movement reading full scale at 50 mA, extended to 10 A, needs 1 × 0.05 / 9.95 = 0.005025 Ω. Enter 0.005 Ω and 1 Ω as the two branches with 10 A total to see the split; the movement receives 49.75 mA. Choose a shunt with a low temperature coefficient, because its drift becomes a direct scale error.

Why is the equivalent resistance always smaller than the smallest branch?

Because adding a branch adds a path, and paths add conductance. R_eq = 1/(G₁+G₂+…), every term in that sum is positive, so the total conductance exceeds the largest single conductance and its reciprocal falls below the smallest single resistance. Three 6 Ω resistors in parallel measure 2 Ω, and n equal resistors of value R measure R/n.

Does the current divider rule work with capacitors and inductors?

Only in its impedance form, and only with complex arithmetic. Replace each resistance with the branch impedance at the frequency of interest and the structure of the formula is unchanged, but both magnitude and phase matter, so branch currents no longer add arithmetically. For a single frequency, compute each branch's impedance first — the capacitive reactance and inductive reactance calculators give the reactance terms.

What current imbalance between paralleled conductors is normal?

There is no single normal figure, because it depends on how closely the parallel paths match in length, size, material and routing. The useful test is the one this calculator performs: measure each path's end-to-end resistance, enter them as branches, and compare the predicted shares with what a clamp meter reads. A measured share that differs from the prediction points at a loose termination or a path you have not accounted for.

Can a branch current be larger than the total current?

Not in a purely resistive divider fed from one source: every share lies between 0 and 1 and they sum to 1. A branch current larger than the total means either another source is injecting current at that node, or there is reactance and you are comparing magnitudes of quantities that are out of phase — in a parallel resonant circuit the circulating branch currents genuinely exceed the line current.

References

  • Fundamentals of Electric Circuits, 6th ed. — McGraw-Hill Education (Alexander & Sadiku)
  • The Art of Electronics, 3rd ed. — Cambridge University Press (Horowitz & Hill)
  • IEEE Standard Dictionary of Electrical and Electronics Terms, 7th ed. — Institute of Electrical and Electronics Engineers