ARCHETYPE: Multi-Loop & Two-Battery Circuits
Take one 12-volt battery, one 6-volt battery, and three resistors. Wire the batteries into different branches of the same circuit and something unsettling happens to your toolkit: nothing is in series with anything, nothing is in parallel, and the folding tricks of ARCHETYPE: Equivalent-Resistance Puzzles find no symmetry to fold. Yet the electrons are not confused — currents settle instantly into one definite pattern. The question is whether you can predict it under exam pressure. This setup — two or more EMFs, three or more branches — is the most reliably-set circuit in school physics. CBSE and ISC ask it as "find the current in each branch" for three to five marks; JEE-Main hides it inside "which battery is charging?"; JEE-Advanced and A-Level dress it with internal resistances and ask for the potential difference between two unhelpful-looking points. Every version yields to one drill: label branch currents with free guesses, write one junction statement and two loop walks, solve, and read negative answers calmly. Kirchhoff's Rules gives you the two laws; this gym makes them fast. The payoff goes beyond marks. The two-battery circuit is how every phone charger, car jump-start and solar installation actually works — one source deliberately overpowering another through resistance. By the end you will recognise, from a single minus sign, which battery is secretly being refilled.
- The archetype: two EMFs in different branches, three branches meeting at two junctions — no series, no parallel, no symmetry. Every board sets it.
- The drill: draw a current arrow on every branch (guess freely), write the junction statement, walk two independent loops with strict sign discipline, solve.
- A negative answer is information, not error: the true current runs opposite your arrow — often meaning one battery is being charged.
- The canonical network (12 V behind 1 Ω, 6 V behind 2 Ω, 2 Ω shared) solves to , , — the 6 V battery absorbs energy.
- Superposition-thinking: the equations are linear, so currents equal the sum of one-source-at-a-time answers (kill an EMF to plain wire, keep its internal resistance).
- Audit every answer with the power ledger: source power in = heat out + charging power stored.
- Key idea: label smart, guess freely, walk loops with one consistent sign convention, and trust the minus sign — the method never jams.
Any multi-loop circuit surrenders to the branch-current drill — one junction equation, independent loop walks with disciplined signs, and a linear solve — with guessed arrow directions carrying no risk, because a negative result simply reports the true current flowing the other way.
Where this lives today
Jump-starting a car is this article's second board example performed with 200-ampere clamps: a healthy 12.6 V battery in parallel with a sagging one, simultaneously running the starter motor and driving current backward through its weak partner — the same minus sign you computed, now smelling faintly of sulphuric acid. Every power bank charging a phone while the phone runs is the canonical three-branch circuit in miniature: source branch, storage branch, load branch, one junction statement deciding who feeds whom. At industrial scale the drill was simply automated. SPICE and its descendants assemble exactly these junction-and-loop equations for billions of elements and solve them millions of times per design; a laptop's power-management firmware juggling wall power, battery charging and load is checked against precisely the two-battery bookkeeping you now do by hand.
The setup
Here is the circuit this article masters. Two junctions, on top and below, three branches between them: on the left a 12 V battery with behind it; in the middle a plain ; on the right a 6 V battery with . Both batteries point their positive terminals toward , each trying to push current up its own branch.
Test your old tools on it honestly. Is the in series with the middle ? No — junction sits between them, bleeding current into the right branch. Are the two resistors in parallel? No — one has a battery riding in its branch. The reduction rules of Resistors in Series & Parallel need bare resistors; a battery inside a branch poisons every shortcut. This is branch-current territory, and three habits define the drill.
Label smart. Give every branch its own current symbol and arrow. Guess directions freely — the algebra is indifferent — but count before you write: 3 branches need 3 equations, and junctions supply only one independent statement (the second junction repeats the first, since what flows into must flow out of ). The rest come from loops: here, two. Smarter still, spend the junction rule at birth — call the currents , and write the middle one as immediately, working with two unknowns instead of three.
Walk loops with one contract. Choose a direction around each loop and never renegotiate mid-walk. Crossing a resistor with its current arrow costs ; against it, . Crossing a battery from its negative plate to its positive plate gains regardless of which way the current flows; the reverse crossing costs . (Why these signs work is Kirchhoff's Rules territory; here we spend them, fast.)
Trust the minus sign. If a current solves negative, your guess was backwards; the magnitude is already correct. Do not redraw arrows — that way lies a fifteen-minute spiral. Negative currents through a battery are physics: that battery is being charged, absorbing into chemistry, exactly as EMF, Terminal Voltage & Internal Resistance describes from the single-battery side.
The master solution — the canonical circuit, beaten thoroughly
We want all three currents in the hero circuit: 12 V behind on the left, 6 V behind on the right, in the shared middle branch. The strategy is the drill verbatim — guess arrows, one junction statement, two loop walks, solve, audit with the power ledger. Watch the right branch: we guess its current upward (the way its battery pushes), and the circuit will overrule us.
Step 1. Label: up the left branch, up the right, down the middle. Junction : . The junction rule is spent; everything else is loops.
Step 2. Left loop, clockwise from : up through the 12 V source from minus to plus (), through its with the current (), down the middle with (), home:
Step 3. Right loop, anticlockwise from (this walk also goes up the right branch and down the middle):
Why walk the loops in opposite rotational senses? No law requires it — any two independent walks work — but walking with the guessed arrows keeps the signs uniform and cuts slips in half. The contract, not the choice, is what matters.
Step 4. Substitute . From Step 2: ; from Step 3: . Add and equate to :
Back-substitute: and .
Step 5. Read the signs. and positive: those guesses were right. : the right branch's current actually flows downward, 0.75 A pushed backward through the 6 V battery — the 12 V source is feeding the middle resistor and charging its weaker partner.
Step 6. Audit. The 12 V source delivers . Heat: . The 6 V battery absorbs into chemistry. Ledger: . Exact — certified. Verified: node analysis agrees — yields , whence , and ; the ledger balances to the last digit.
That node-voltage cross-check deserves promotion to a habit — it is how you answer the JEE favourite "find the potential difference between and ." Potential bookkeeping along a branch: start at and add potential changes as you walk to . Up the right branch: . With (0.75 A flowing downward, against the walk):
where: the is the resistor's contribution — walking against the true current, potential rises across a resistor. Read it physically: the branch's terminals sit at 7.5 V though its EMF is only 6 V, because a charging battery's terminal voltage exceeds its EMF by — the charging twin of the discharge law. Walking the left branch instead gives : same answer, as single-valued potential demands. Two branches, one number — a third independent check, after the junction audit and the power ledger.
Superposition-thinking: one source at a time
Go deeper · Derivation B · 1 minWhy one source at a time is legal and when it is fastest
We want to justify the competition shortcut: solve the circuit with only the 12 V source alive, again with only the 6 V alive, and add the answers. The strategy is to read the structure of the loop equations — linear in the EMFs — then verify numerically on the canonical circuit. The mechanical rule to watch: killing an EMF means replacing it with plain wire, but its internal resistance stays.
Step 1. Every term in the master derivation's equations is either (EMF) or (resistance × current), resistances fixed by hardware. So the currents produced by several EMFs together equal the sum of the currents each would produce alone, the others set to zero. Setting an EMF to zero while keeping its branch's resistance is exactly "replace the source by a wire, keep ."
Step 2. 6 V dead: the right branch becomes a bare , and ordinary reduction works — middle and right in parallel (), plus the left : total . The 12 V drives up the left branch; splits it equally, down the middle, down the right.
Step 3. 12 V dead: the left branch becomes a bare ; middle and left in parallel () plus the right : total . The 6 V drives up the right branch; gives down the middle, down the left.
Step 4. Add, respecting directions (up positive in each outer branch, down positive in the middle):
All three match the master derivation exactly. Why bother, if the drill already works? Because each one-source sub-circuit is a plain series-parallel problem — no simultaneous equations. When an MCQ asks for one current in a two-source network, superposition is often the 90-second route. Its one forbidden use: power — heat goes as , and . Superpose currents first, then square. *Verified: both sub-circuits recomputed by reduction; the summed currents reproduce , , , and the ledger still certifies the total.*
Variants at a glance
| variant | what changes | the key move |
|---|---|---|
| single loop, EMFs opposing | no junctions | ; bigger EMF wins, smaller charges |
| canonical three-branch | the archetype | one junction statement + two loop walks |
| parallel batteries + load | the "jump start" | node bookkeeping: one unknown |
| find | a potential, not a current | walk any branch; all must agree |
| what EMF makes a branch dead? | design, | dead branch ⇒ terminals differ by exactly |
| one current only, MCQ | speed matters | superposition: two series-parallel sub-circuits |
Worked examples
Board level — one loop, two opposing batteries (CBSE/ISC pattern, 3 marks)
A battery of EMF , internal resistance , sits in a single loop with a battery of EMF , internal resistance , and an external , EMFs opposing. Find the current, each terminal voltage, and each battery's role.
Step 1. One loop, no junctions: walk once in the direction the 9 V pushes. .
Step 2. Solve: , so , flowing the way the stronger EMF pushes.
Step 3. Terminal voltages, by the twin laws. Discharging: . The 3 V battery has current driven into its positive terminal — charging — so .
Step 4. Ledger: the 9 V source delivers ; resistors burn ; the 3 V battery banks as chemistry. ✓. This loop is every battery charger you own.
Board level — two batteries in parallel charging one another (ISC/A-Level pattern)
Two cells, () and (), are joined in parallel (positive to positive) across a load. Find all three currents and each cell's role.
Step 1. Label the shared top node , bottom node ; set , call . Branch currents toward : from cell 1, ; from cell 2, ; down the load, .
Step 2. Junction statement at : . Multiply by 2: , so and .
Step 3. Currents: out of cell 1; — cell 2 receives 0.8 A, because the node sits above its EMF; load . Junction audit: ✓.
Step 4. Ledger: cell 1 delivers ; heat is ; cell 2 stores . ✓. Moral: parallel cells of unequal EMF quietly waste power charging the weaker one — why Combining Cells: Series & Parallel wants matched EMFs in parallel banks.
Competitive level — one current, ninety seconds (JEE-Main pattern, ~2.5 minutes)
In the canonical circuit (12 V behind ; 6 V behind ; middle branch ), find the current in the middle branch.
Step 1. Superposition route. Kill the 6 V (wire in its place, stays): the 12 V sees , drives 6 A, and the node voltage splits it equally: down the middle.
Step 2. Kill the 12 V ( stays): the 6 V sees , drawing ; node voltage sends down the middle.
Step 3. Both contributions run the same way: . Timing note: budget 2.5 minutes. Superposition wins because both sub-circuits are plain reductions — no simultaneous equations. Had the question asked for all three currents or any power, the branch-current drill is faster overall; and never superpose powers — currents first, square after.
Competitive level — the dead branch and the overfilled battery (JEE-Advanced stretch, ~4 minutes)
In the canonical topology the right-branch EMF is unknown: 12 V behind left, behind right, middle. (a) For what does the right branch carry no current? (b) For , find the right battery's terminal voltage and state.
Step 1. Part (a): a dead branch draws nothing and drops nothing across its own resistance, so its terminal pair sits at a potential difference of exactly . Meanwhile the left-plus-middle loop runs alone: , putting the node at above .
Step 2. Dead-branch condition: . (Both ways: pushes current up the right branch; lets the node push current down into it.)
Step 3. Part (b): , so current flows down the right branch — the battery charges. From the master solution that current is and .
Step 4. Terminal voltage of the charging battery: — matching , as it must. A charging battery's terminals read above its EMF; on discharge, below. Examiners flip this sign on one option every year. Timing note: ~4 minutes. Part (a) is pure node thinking — no loop equations once you see that a dead branch pins its terminal difference to its EMF. That idea (" through a source ⇒ terminals show the bare EMF") also powers the potentiometer's null method, coming later in this chapter.
Exam traps
Renegotiating the walk. Choosing clockwise, then flipping direction halfway because a battery "felt backwards" — the classic self-inflicted wound. Decide the tour, write every sign from the crossing, let wrong-feeling terms stand.
EMF signs from current direction. The sign of depends only on which plate you cross first: minus-to-plus is , even if the current inside runs plus-to-minus (a charging battery). Tying it to the current is the most common single error in board scripts.
Using both junction equations. With two junctions, the second statement repeats the first; adding it produces a degenerate system. Junctions give independent statements — here, one.
Killing a source but not keeping its . In superposition an EMF dies to a plain wire but its internal resistance stays. Deleting the whole branch changes the topology and every answer downstream.
Superposing power. Heat goes as ; currents add, their squares do not. Total current first, then square. A JEE-Main option equal to "" is bait.
Terminal voltage of a charging battery. on charge, on discharge. If your two-branch walks of disagree, this sign is almost always the culprit.
Fraction discipline. These circuits are engineered for clean fractions (, ). Keep them until the last line; a rounded for makes the power ledger miss by watts and costs the audit mark.
Origins · 1 minLéon Charles Thévenin and how this was discovered
Thévenin was a French telegraph engineer with a working engineer's complaint: solving the same multi-battery network again for every new load was a waste of a career. His 1883 theorem — any two-terminal tangle of sources and resistors behaves, from outside, like one EMF behind one resistance — turned that repetition into a single calculation done once. He published it as a modest note, unaware that Helmholtz had proved the same result thirty years earlier while studying currents in animal tissue; engineering history kept Thévenin's name on it anyway. The theorem is the industrial-strength version of this article's superposition-thinking (both are children of linearity), and it is why the " behind " battery model works for wall adapters and solar panels as faithfully as for chemical cells: every source you will ever meet is somebody's two-terminal tangle, already collapsed.
Going further
Texts. NCERT Physics Class 12, Ch. 3 · HC Verma, Concepts of Physics Vol. 2, Ch. 32 · Halliday, Resnick & Walker, Ch. 27 (multiloop circuits).
Provenance. All results recomputed from scratch. A verifier should confirm: the canonical solution , , (; ledger ); the superposition split (/ sub-circuits, sums matching); board example 1 (; /; ); board example 2 (; , , ; ); the dead-branch condition ; the charging terminal voltage .
Prerequisites. Kirchhoff's Rules · EMF, Terminal Voltage & Internal Resistance. Leads to. ARCHETYPE: Battery Problems — Terminal V, Max Power & Efficiency · Combining Cells: Series & Parallel · Redraw the Circuit (technique, coming up) · Kirchhoff's Rules (UG deep dive) — the university treatment, including where the loop rule can legitimately fail.
How do I decide which way to draw the current arrows before I know the answer?
You don't, and you shouldn't try. The arrows define what "positive" means for each branch — nothing more. Guess anything, keep every sign consistent during the walks, and the algebra reports the truth: positive means the guess was right, negative means the real current runs the other way with exactly the computed magnitude. Experienced solvers guess plausibly (current out of the biggest EMF's positive terminal) to keep minus signs rare, but the answers are identical either way.
Why do I only get one junction equation from two junctions?
Because charge that flows into the top junction must flow out of the bottom one — the two statements carry the same information. Write both and one is the other times minus one; include both and the system becomes degenerate. The general count: junctions yield independent junction statements, and loops supply the rest. For the three-branch archetype: one junction statement, two loops.
When should I use branch currents versus node potentials?
Branch currents are the marks-scheme default — boards award steps for the junction statement, each loop walk and the solve. Node bookkeeping (set , solve for ) collapses the problem to one unknown and shines when branches share just two junctions — the parallel-charging network — or when the question asks for a potential difference directly. Master both: they are the same two laws in different clothes, and agreement between them is a complete check.
Is a battery being charged really "absorbing" power? Where does it go?
Genuinely absorbing: current driven backward through a source pushes its chemistry in reverse, converting electrical energy into stored chemical energy at rate — that is what "charging" means. (The heating in its internal resistance happens on top of that and is lost.) In primary, non-rechargeable cells the reverse chemistry has nowhere to go, so the energy becomes heat and gas — which is why forcing current backward through a dry cell is a safety hazard, not a recharge.
Why is superposition forbidden for power when it works perfectly for current?
Because the loop equations are linear in currents and EMFs, but power is quadratic: . If two sources separately drive and through a resistor, the true heat involves , not — the cross term is real energy. The safe procedure is mechanical: superpose currents branch by branch, then compute every power from the totals.
My two walks for the potential difference between the junctions give different answers. Which is right?
Neither — a disagreement means an arithmetic or sign slip, because in a steady circuit the potential difference between two points is a single number no matter which branch you walk. Recheck the charging-battery sign first ( when current is driven into the positive terminal); that error causes the mismatch in nine cases out of ten. Used deliberately, this is the cheapest full-solution check available: walk two different branches and demand the same .