The Wheatstone bridge is the definitive resistor diagram circuit for precision analog sensing. Unlike a basic series string or a single-ended voltage divider, a bridge topology uses four resistive arms arranged in a diamond to measure unknown resistance or detect minute changes in a sensor—like a strain gauge, load cell, or thermistor. It achieves this by nulling the differential voltage between two midpoints, allowing you to extract high-resolution data even from noisy environments.
In this guide, we will break down the exact node topology, select real-world component values for a 10kΩ NTC thermistor, analyze failure modes, and walk through a bench-verified breadboard testing sequence.
Topology and Node Labels: Reading the Resistor Diagram Circuit
To design or troubleshoot this network, you must first map the standard node labels. The classic bridge consists of two parallel voltage dividers sharing a common excitation source and ground.
- Node A (Top): Excitation Voltage ($V_{EX}$ or $V_{IN}$). This is your stable reference voltage, typically 3.3V or 5.0V.
- Node B (Right): Output High ($V_{OUT+}$). The midpoint of the sensor leg.
- Node C (Bottom): Ground Reference ($GND$). The common return path.
- Node D (Left): Output Low ($V_{OUT-}$). The midpoint of the fixed reference leg.
The resistive arms connect these nodes: R1 bridges A to B, R2 bridges B to C (this is typically your sensor), R3 bridges A to D, and R4 bridges D to C. The differential output voltage is measured between Node B and Node D, calculated as:
$V_{OUT} = V_{IN} \times [ (R2 / (R1 + R2)) - (R4 / (R3 + R4)) ]$
When the ratio of R1/R2 perfectly matches R3/R4, the bridge is 'balanced' and $V_{OUT}$ is exactly 0V. Any physical change in the sensor (R2) unbalances the bridge, generating a proportional millivolt signal.
Component Selection and Behavior Matrix
Let us design a practical bridge for a 10kΩ NTC thermistor (R2) measuring ambient temperature, using a 5.0V excitation source. We will set R1, R3, and R4 to 10.00 kΩ.
The 10kΩ Sweet Spot
Why choose 10kΩ for the fixed arms? If you use 100Ω resistors, the 5V excitation pushes 25mA through the thermistor. This dissipates over 60mW of heat inside the sensor bead, causing severe self-heating errors that can skew your reading by several degrees. Conversely, if you use 1MΩ resistors, the current drops to 2.5µA. At this level, Johnson-Nyquist thermal noise increases, and the input bias current of a standard microcontroller ADC (often 1-10µA during sample-and-hold) will pull the node voltage, ruining your accuracy. 10kΩ limits current to 250µA per leg, keeping self-heating below 0.1°C while maintaining a low enough impedance to drive most ADCs cleanly.
For R1, R3, and R4, do not use standard 5% carbon film resistors. Specify 0.1% tolerance metal film resistors with a 25ppm/°C temperature coefficient, such as the Vishay PR02 or Yageo MFR-25 series.
| State | R2 Value (NTC) | V_OUT+ (Node B) | V_OUT- (Node D) | Differential V_OUT | Total Current |
|---|---|---|---|---|---|
| Balanced (25°C) | 10.00 kΩ | 2.500V | 2.500V | 0.000 mV | 0.50 mA |
| Hot (50°C) | 3.60 kΩ | 1.323V | 2.500V | -1177 mV | 0.68 mA |
| Cold (0°C) | 27.28 kΩ | 3.658V | 2.500V | +1158 mV | 0.41 mA |
| Shorted Sensor | 0 Ω | 0.000V | 2.500V | -2500 mV | 1.00 mA |
| Open Sensor | ∞ Ω | 5.000V | 2.500V | +2500 mV | 0.25 mA |
As shown in the behavior matrix above, the differential voltage swings symmetrically around 0V as the temperature changes, providing a clear, bipolar signal for your microcontroller to process.
Failure Modes: What Breaks at the Extremes?
To understand the robustness of this resistor diagram circuit, we must contrast it with a simple series topology. In a basic series string of resistors or LEDs, a single open connection drops the current to zero everywhere. The entire circuit dies silently, and a microcontroller monitoring a downstream node might just read 0V, unable to distinguish between a legitimate low signal and a broken wire.
The Wheatstone bridge handles extremes with diagnostic clarity:
- Open R2 (Sensor wire snaps): Node B is no longer pulled to ground. R1 pulls Node B all the way up to $V_{IN}$ (5.0V). The differential output rails to +2500 mV. Your firmware immediately reads a maxed-out ADC value and can trigger a definitive 'Sensor Open' fault code.
- Short R2 (Sensor wires shorted together): Node B is pulled directly to ground (0V). The differential output rails to -2500 mV. The firmware registers a 'Sensor Short' fault.
- Open R1 (Excitation leg breaks): Node B is pulled to ground via R2. The output rails negative, again providing a clear, detectable fault state rather than a silent failure.
Step-by-Step Breadboard Testing and Verification
Breadboarding a bridge circuit introduces parasitic contact resistance (often 0.1Ω to 0.5Ω per spring clip). Follow this exact sequence to verify your circuit without chasing phantom errors.
- De-energize and Populate: With the power supply off, insert your 0.1% fixed resistors (R1, R3, R4) and the NTC thermistor (R2) into the breadboard. Ensure the nodes A, B, C, and D are physically separated on different terminal strips to avoid accidental shorts.
- Wire Excitation: Connect Node A to your 5.0V precision reference (or a clean LDO output, not a noisy switching buck converter). Connect Node C to the common ground.
- Single-Ended Verification (Crucial Step): Before measuring the differential output, power the circuit and measure Node B to GND, then Node D to GND. Both should read approximately 2.50V. If Node D reads 2.45V and Node B reads 2.50V, your fixed resistors are mismatched, or breadboard contact resistance is skewing the reference leg. Fix this before proceeding.
- Differential Measurement: Switch your multimeter to the millivolt DC range. Place the red probe on Node B and the black probe on Node D. At room temperature (approx 25°C), you should read between -5mV and +5mV.
- Thermal Perturbation: Pinch the thermistor bead with your fingers. The differential voltage should immediately swing negative (as resistance drops) by several hundred millivolts. Release it and watch it decay back to zero.
Why a Bridge Over a Simple Voltage Divider?
A common question from hobbyists is why they cannot simply use a single voltage divider (one fixed resistor, one sensor) referenced to ground, as detailed in basic tutorials like those on SparkFun's Voltage Divider guide. The answer lies in Common-Mode Rejection Ratio (CMRR) and power supply noise.
In a single-ended divider, your output voltage is strictly referenced to ground. If your 5V rail has 50mV of high-frequency ripple from a switching regulator, that ripple couples directly into your ADC reading, manifesting as jitter in your temperature or weight data. You would need heavy software filtering or complex analog low-pass filters to clean it up.
The Wheatstone bridge outputs a differential signal. If the 5V rail jumps by 50mV, both Node B and Node D jump by approximately 25mV simultaneously. When you feed these two nodes into a differential amplifier (like the classic INA128) or a differential-input ADC (like the Texas Instruments ADS1115), the hardware subtracts Node D from Node B. The 50mV of common-mode noise cancels out mathematically, leaving only the pure microvolt-level signal from the sensor. As noted in All About Circuits' DC textbook, this ratiometric cancellation is what makes the bridge topology the undisputed standard for load cells and precision strain gauges in industrial environments.
The ADC Interface Reality
Do not connect a Wheatstone bridge directly to a standard Arduino Uno (ATmega328P) analog pin expecting high resolution. The Uno's ADC is single-ended and only offers 10-bit resolution (4.8mV per step at 5V). A bridge outputting a 20mV swing will only use 4 steps of the ADC, rendering the precision useless. Always pair this resistor diagram circuit with a 16-bit differential ADC like the ADS1115 over I2C, or an instrumentation amplifier to boost the millivolt signal to a 0-5V range before it hits the microcontroller.
