Wire voltage sources (like batteries) in series to increase total voltage, and in parallel to increase total capacity (amp-hours). Wire loads (like LEDs, resistors, or heaters) in series to force them to share the same current, and in parallel to ensure they all receive the same voltage. The decision always comes down to a single question: is your power supply voltage-constrained or current-constrained?
If you have a 12V power supply and need to drive 24V loads, you must wire the loads in parallel and step up the voltage, or wire two 12V batteries in series. If you need to drive three 3V LEDs from a 12V source, you wire the LEDs in series. Below is the bench-level breakdown of how these topologies behave, how they fail, and how to design them with real component values.
Topology Breakdown: Nodes, Current, and Voltage
To understand circuit behavior, we map connections using nodes. A node is any continuous conductive path where two or more components meet.
Series Topology: Components are connected end-to-end, forming a single path for current. Imagine a 12V battery connected to Resistor 1 (R1), which connects to Resistor 2 (R2), which returns to the battery.
Node A is the positive terminal. Node B is the junction between R1 and R2. Node C is the negative terminal. Current ($I$) is identical through all components, while voltage ($V$) divides across them based on their resistance ($V_{total} = V_{R1} + V_{R2}$).
Parallel Topology: Components are connected across the same two nodes, creating multiple paths. Both R1 and R2 connect directly between Node A (positive) and Node B (negative). Voltage is identical across all branches, while current divides based on branch resistance ($I_{total} = I_{R1} + I_{R2}$).
Behavior Table: What Changes When One Element Changes?
Understanding how a circuit reacts to component drift or failure is critical for troubleshooting. Here is what happens if the resistance of R1 increases (or fails entirely).
| Parameter Changed | Series Circuit Reaction | Parallel Circuit Reaction |
|---|---|---|
| R1 increases in value | Total current drops. Voltage across R1 increases; voltage across R2 decreases. | Total current drops slightly. Current through R1 drops; current through R2 remains completely unchanged. |
| R1 fails OPEN | Total current drops to 0A. The entire circuit stops functioning (single path broken). | Current through R1 drops to 0A. R2 continues operating normally on its independent path. |
| R1 fails SHORT | Total resistance drops. Current spikes. R2 receives the full source voltage (likely causing an overvoltage cascade failure). | Total resistance drops to near 0Ω. Massive current spike trips the main breaker, blows a fuse, or sags the power supply to 0V. |
Failure Modes: What Breaks at the Extremes?
Abstract theory assumes ideal components. On the workbench, components fail. Here is the failure-mode contrast you must design around.
The Series Cascade: In a series string of LEDs, if one LED fails short (a common failure mode for high-power emitters), the string's total forward voltage drops. The constant-current driver or current-limiting resistor now pushes the same current through fewer LEDs, but the remaining LEDs must absorb the excess voltage. This over-drives them, generating excess heat and causing them to fail in rapid succession—a cascade failure.
The Parallel Thermal Runaway: If you wire LEDs in parallel without individual current-limiting resistors, they share current based on their exact forward voltage ($V_f$). Because $V_f$ has a negative temperature coefficient, as one LED gets hot, its $V_f$ drops. It then hogs more current, gets hotter, drops further, and eventually burns out. When it fails open, the remaining LEDs are forced to absorb its share of the current, accelerating their demise.
Design Walkthrough: Sizing Real Components for a 12V LED Array
Let’s apply this to a real design scenario. You need to power three Cree XLamp XP-E2 LEDs from a standard 12V DC bench supply. The datasheet specifies a typical forward voltage ($V_f$) of 2.9V and a nominal drive current ($I_f$) of 350mA.
Option A: Pure Series Configuration
- Total $V_f$ for 3 LEDs = $2.9V \times 3 = 8.7V$.
- Voltage to be dropped by the current-limiting resistor = $12V - 8.7V = 3.3V$.
- Required Resistance ($R = V / I$) = $3.3V / 0.35A = 9.42\Omega$.
- Standard E12 Resistor Value: 10Ω.
- Resistor Power Dissipation ($P = I^2R$) = $0.35^2 \times 10 = 1.225W$.
- Component Choice: One 10Ω, 2W (or 3W) metal film resistor.
- Total Current Draw: 350mA.
Option B: Pure Parallel Configuration
- Each LED requires its own dedicated resistor to prevent thermal runaway.
- Voltage to drop per branch = $12V - 2.9V = 9.1V$.
- Required Resistance per branch = $9.1V / 0.35A = 26\Omega$.
- Standard E12 Resistor Value: 27Ω.
- Resistor Power Dissipation per branch = $0.35^2 \times 27 = 3.3W$.
- Component Choice: Three 27Ω, 5W wirewound resistors.
- Total Current Draw: $350mA \times 3 = 1.05A$.
The Verdict: Why choose series over parallel here? The series topology wins decisively. It requires only one 2W resistor, wastes less power as heat, and draws only 350mA from the 12V supply. The parallel topology requires three bulky 5W resistors, generates massive localized heat, and draws over 1 Amp. You only choose parallel for loads when your source voltage is lower than the load's required forward voltage.
Breadboard Testing: Step-by-Step Verification
Before soldering the series LED string designed above, verify it on a breadboard. High-power LEDs like the Cree XP-E2 will overheat in seconds without a heatsink, so we will test using standard 5mm indicator LEDs (2.0V $V_f$, 20mA $I_f$) scaled to the same 12V topology for safe bench testing.
- De-energize the board: Ensure your 12V bench supply is turned off and unplugged. Never build or modify a live circuit.
- Place the components: Insert three 5mm LEDs in series across the breadboard rails, noting the anode/cathode polarity. Insert a 470Ω, 1/4W current-limiting resistor in series with the string.
- Verify continuity (Dead Test): Set your multimeter (e.g., Fluke 117) to continuity mode. Probe from the positive rail to the negative rail. You should read an open loop (OL). If it beeps, you have a short circuit.
- Configure the meter for current: Move the multimeter's red probe to the mA or A jack. Set the dial to DC Amps. Warning: Never measure current in parallel with a voltage source; the meter's internal shunt will short the supply and blow the meter's fuse.
- Break the circuit to measure: Disconnect the positive jumper wire. Place the multimeter probes in series to bridge the gap (red probe to the 12V source, black probe to the LED anode).
- Energize and read: Turn on the 12V supply. The LEDs should illuminate. The meter should read approximately 20mA to 22mA. If it reads 0mA, check for an open LED or backward polarity. If it reads >30mA, your resistor value is too low.
Frequently Asked Questions
Should I wire my 18650 battery pack in series or parallel?
It depends on your inverter or load voltage. If you are building a 12V system, wire 3.7V nominal 18650 cells in parallel (e.g., 4P) to increase amp-hours, then wire those parallel groups in series (e.g., 3S4P) to reach the required voltage. According to Battery University, parallel strings require careful cell matching to prevent cross-currents, whereas series strings require a BMS to monitor individual cell balancing during charging.
Can I mix different value resistors in series or parallel?
Yes, but the math changes. In series, total resistance is simply the sum ($R_{total} = R_1 + R_2$). In parallel, the total resistance is always lower than the smallest individual resistor, calculated as $1 / (1/R_1 + 1/R_2)$. However, in a series circuit, the resistor with the highest value will dissipate the most power ($P = I^2R$), so ensure its wattage rating is sized accordingly, not just the total string wattage.
Why do my parallel LEDs burn out one by one?
This is the classic parallel thermal runaway described earlier. If you wired multiple LEDs in parallel using only a single shared current-limiting resistor, slight manufacturing variations in their forward voltage ($V_f$) cause them to draw unequal current. The one drawing the most current gets the hottest, which lowers its $V_f$ further, causing it to draw even more current until it fails. Always use a dedicated series resistor for every individual parallel LED branch.
How do I calculate total wattage for speakers in series vs parallel?
Speaker wiring affects impedance (AC resistance), which changes how much power the amplifier delivers. Wiring two 8Ω speakers in series yields 16Ω; the amplifier delivers less power, but the wattage is split evenly between the speakers. Wiring them in parallel yields 4Ω; the amplifier delivers roughly double the power (if it is stable at 4Ω), and again, the power splits evenly. Always check your amplifier's minimum impedance rating before wiring speakers in parallel to avoid blowing the output transistors.
