To work out total resistance in a parallel circuit, calculate the reciprocal of the sum of the reciprocals of each individual branch resistance. The universal formula is 1/R_total = 1/R1 + 1/R2 + ... + 1/Rn. For exactly two resistors, use the product-over-sum shortcut: R_total = (R1 × R2) / (R1 + R2). Regardless of the number of branches, the calculated total equivalent resistance will always be strictly lower than the smallest individual resistor in the network.
The Parallel Topology: Nodes, Paths, and the Core Formula
A parallel circuit is defined by its nodes. Imagine a primary input junction (Node A) and a primary return junction (Node B). Every component in the network connects directly across Node A and Node B. Because they share the exact same two electrical nodes, the voltage drop across every parallel branch is identical, dictated by Kirchhoff’s Voltage Law (KVL).
While voltage remains constant across branches, current divides. According to Kirchhoff’s Current Law (KCL), the total current entering Node A equals the sum of the currents flowing through each individual branch. According to Georgia State University HyperPhysics, this inverse relationship between resistance and current means the branch with the lowest resistance will draw the highest proportion of the total current.
If you are using n resistors of the exact same value (R) in parallel, skip the reciprocal math. Simply divide the single resistor value by the number of resistors:
R_total = R / n. Three 300Ω resistors in parallel will always yield exactly 100Ω.
Parallel vs. Series: Why Choose Parallel and What Breaks
Why wire resistors in parallel instead of series? In a series topology, current has only one path. If a single series component fails open, the entire circuit is broken, and current drops to zero. Parallel topology provides independent paths. If one parallel branch fails open, the remaining branches continue to operate, albeit with a slightly higher total network resistance.
However, parallel circuits introduce a distinct catastrophic failure mode: the short circuit. If a single resistor in a parallel network fails short (drops to near 0Ω), the total equivalent resistance of the entire network instantly plummets toward zero. This creates a massive current spike that will either trip a breaker, blow a fuse, or melt your PCB traces if unprotected. Series circuits do not suffer from this specific short-circuit cascade; a shorted series resistor simply removes its resistance from the total, causing a moderate current increase rather than a dead short across the supply.
Behavior Matrix: What Happens When One Element Changes
Understanding how a parallel network reacts to physical changes or faults is critical for designing robust protection and sensing circuits. Here is the behavior matrix for a standard DC parallel resistor network:
| Circuit Event | Effect on Total Resistance | Effect on Total Current Draw | Effect on Remaining Branches |
|---|---|---|---|
| Add a new resistor branch | Decreases | Increases | No change (voltage is constant) |
| Remove an existing branch | Increases | Decreases | No change |
| One branch fails OPEN | Increases slightly | Decreases slightly | No change in current per branch |
| One branch fails SHORT | Drops to ~0Ω | Spikes to maximum (fault) | Voltage collapses; branches starve |
| Increase value of one resistor | Increases slightly | Decreases slightly | No change |
Design Walkthrough: Sizing a 120Ω Dummy Load
Let us apply this theory to a physical build. Suppose you need to test a 12V DC bench power supply and want to build a dummy load that draws exactly 100mA. Using Ohm’s Law (R = V / I), your target total resistance is 12V / 0.1A = 120Ω.
The total power dissipated by this load will be P = V × I = 12V × 0.1A = 1.2W. You could use a single 120Ω, 2W power resistor. However, single high-wattage resistors run hot, and their resistance drifts significantly as temperature rises (due to their positive temperature coefficient). A better thermal design is to use multiple standard 1/2W resistors in parallel to spread the heat across a larger physical area.
The Math: We need 120Ω total. Using the identical resistor shortcut (R_total = R / n), if we use three resistors (n = 3), each resistor must be 120Ω × 3 = 360Ω.
The Power Check: The 1.2W total load is divided equally among the three branches. Each 360Ω resistor will dissipate 1.2W / 3 = 0.4W.
The Derating: A standard 1/2W (0.5W) resistor running at 0.4W is at 80% capacity. For reliable bench operation, we want to stay below 70% capacity to keep the component cool to the touch. Therefore, we step up to 1W resistors. Running 0.4W through a 1W resistor is a safe 40% load.
Concrete Part Pick: We select the Yageo CFR-50JB-52-360R (a standard 360Ω, 1/2W carbon film resistor) but realize we need the 1W variant for thermal headroom. The correct part number is the Yageo CFR-100JB-52-360R (360Ω, 1W, 5% tolerance). We buy three of them.
Breadboard Testing: Step-by-Step Verification
Before soldering your parallel network into a permanent enclosure, verify the math on a solderless breadboard. Breadboards introduce parasitic contact resistance (typically 0.1Ω to 0.5Ω per junction), which can skew low-resistance measurements.
- Zero your DMM: Turn your multimeter (e.g., Fluke 115 or Brymen BM235) to the lowest Ohms range. Short the red and black probes together. Note the lead resistance (usually 0.2Ω). You will subtract this from your final reading.
- Establish Nodes: Insert one leg of all three 360Ω resistors into a single shared 5-hole row on the breadboard (this is Node A). Insert the remaining three legs into a second shared row further down the board (Node B).
- Measure Unpowered: Place your DMM probes directly onto the metal lead wires inside the breadboard holes for Node A and Node B. Do not measure the power rails; measure the component leads directly to bypass breadboard track resistance. Your reading should be approximately 120.2Ω (120Ω network + 0.2Ω DMM lead offset).
- Energize and Verify KVL: Connect a 12V supply to Node A and Node B. Measure the voltage directly across the resistor leads. It should read 12.0V. Measure the current using the DMM in series; it should read 0.1A (100mA).
- Thermal Check: Let the circuit run for 5 minutes. Carefully touch the resistors. They should be barely warm. If they are hot to the touch, your breadboard contacts may be oxidized and adding series resistance, or your power supply voltage is sagging.
Decision Tree: Picking Your Resistor Network
When designing a circuit that requires a specific resistance and power profile, use this decision matrix to determine your topology and component selection.
| Design Requirement | Recommended Topology | Action / Calculation |
|---|---|---|
| Need to dissipate high heat (>1W) without using a single massive power resistor. | Parallel | Divide target resistance by n. Multiply target wattage by 1.5 for derating. Pick n standard resistors. |
| Need an odd, non-standard resistance value (e.g., 315Ω) using standard E12/E24 parts. | Parallel (or Series-Parallel) | Use the product-over-sum formula in reverse, or use an online parallel resistor calculator to find two standard values that yield the target. |
| Need to drop a large voltage across a single sensing path. | Series | Sum the resistances. Ensure the voltage rating of each individual resistor is not exceeded. |
| Need circuit redundancy (if one part fails open, the system must keep running). | Parallel | Use two higher-value resistors in parallel. Size them so that if one fails open, the remaining single resistor can handle the full circuit current. |
For general-purpose prototyping, dummy loads, and current-sharing applications where thermal stability is required, always default to a parallel network of 1W metal or carbon film resistors rather than a single 2W+ wirewound resistor. The parallel approach costs pennies more, runs significantly cooler, and provides inherent open-circuit redundancy. For the 120Ω / 1.2W dummy load detailed above, purchase three Yageo CFR-100JB-52-360R (360Ω, 1W) resistors. This is the most robust, cost-effective configuration for the workbench.






