Strain Gauge Circuit Diagram: Wheatstone Bridge Wiring
A strain gauge is a resistive sensor whose resistance changes in proportion to mechanical deformation. The change is small -- a typical metal foil gauge has a gauge factor of about 2, meaning a 0.1% strain produces a 0.2% change in resistance. For a 120Ω gauge, 0.2% is 0.24Ω. To measure a 0.24Ω change on a 120Ω baseline you cannot just connect it to a voltage divider and call it done. You need a Wheatstone bridge.
Why a Wheatstone Bridge?
A Wheatstone bridge has four resistive legs. At balance, the differential voltage across the bridge midpoints is zero. When one (or more) legs change value -- like a strain gauge under load -- the bridge goes out of balance and a small differential voltage appears. This differential voltage is proportional to the resistance change.
The advantage over a simple voltage divider: the bridge rejects the large common-mode component (the DC excitation voltage) and amplifies only the difference. A precision instrumentation amplifier amplifies this differential signal to a readable level.
Quarter Bridge, Half Bridge, and Full Bridge
The Wheatstone bridge has four arms. Replacing one, two, or four of those arms with active gauges determines the bridge type.
Quarter Bridge
One active gauge (R_sg), three fixed precision resistors (all equal to gauge nominal resistance):
+Vex
|
R3 ──┤── R_sg
| |
Vout+ Vout-
| |
R4 ──┤── R2
|
GND
The output voltage at small strains: Vout ≈ (Vex / 4) × (ΔR / R)
For Vex = 5V, ΔR/R = 0.002 (0.2% change): Vout ≈ 5V/4 × 0.002 = 2.5mV
This is the signal level you are dealing with -- millivolts. A gain of 100--500 is needed before feeding an ADC.
Quarter bridge is the simplest to wire and is fine for measuring bending or axial strain where temperature compensation is not critical. The three fixed resistors should be matched metal-film types (0.1% tolerance or better) to keep the bridge balanced at zero strain.
Half Bridge
Two active gauges in adjacent arms of the bridge. The typical arrangement puts one gauge in tension and one in compression (as in a bending beam), which doubles the output compared to a quarter bridge and provides inherent temperature compensation -- both gauges see the same temperature change, which cancels out.
Vout ≈ (Vex / 2) × (ΔR / R)
For the same 0.2% change: Vout ≈ 5V/2 × 0.002 = 5.0mV
This arrangement is common in commercial load cells and torque sensors.
Full Bridge (Wheatstone Bridge with Four Active Gauges)
All four arms are active gauges. Two in tension, two in compression, arranged so opposing gauges are in opposite arms.
Vout ≈ Vex × (ΔR / R)
For 0.2% change: Vout ≈ 5V × 0.002 = 10mV
Full bridge configurations are used in precision force and weight sensors. Virtually all commercial load cells -- the kind you find in kitchen scales and industrial weigh platforms -- use a full bridge inside a machined aluminum or steel body.
Load Cell Color Codes
Commercial load cells typically use a 4-wire system:
| Wire Color | Connection |
|---|---|
| Red | +Vex (excitation positive) |
| Black | -Vex / GND (excitation negative) |
| White | Signal + (Vout+) |
| Green | Signal - (Vout-) |
Some manufacturers use Red/Black/White/Blue. A few use a 6-wire configuration with additional Sense+ and Sense- wires that allow the amplifier to compensate for excitation voltage drops in long cable runs (Kelvin sensing). When a 6-wire load cell feeds a 4-wire input, tie Sense+ to Exc+ and Sense- to Exc-.
Excitation Voltage
The bridge needs a stable, clean excitation voltage. Noise on the excitation supply directly appears as noise on the output signal.
Typical excitation voltages:
- 5V: Common when the system runs from USB or a 5V regulated supply. Works well with the HX711.
- 3.3V: Lower output signal but fine for short-range analog or when the ADC reference is also 3.3V.
- 10V: Higher signal-to-noise ratio; used in precision industrial systems.
Use a precision voltage reference (LM4040, REF3025) or the regulated supply from a low-noise LDO rather than a raw microcontroller VCC pin, which carries switching noise from the digital circuitry.
Amplifying the Signal: HX711 vs. INA125
HX711
The HX711 is the most popular load-cell amplifier for maker/hobbyist use. It combines a precision instrumentation amplifier with a 24-bit delta-sigma ADC, communicates over a simple 2-wire serial interface (CLK/DOUT), and includes an internal voltage reference.
Wiring a load cell to an HX711:
- Red (Exc+) → E+ pin
- Black (Exc-) → E- pin
- White (Sig+) → A+ pin
- Green (Sig-) → A- pin
- VCC: 2.7--5.5V (from 3.3V or 5V microcontroller supply)
- GND: Common ground
- CLK, DOUT → Arduino digital pins
Gain is set by channel selection: channel A at gain 128 (default), channel A at gain 64, or channel B at gain 32. At gain 128 with 5V excitation, the full-scale input range is ±20mV -- fine for most load cells.
Arduino library: SparkFun's HX711 library or bogde's HX711 library both work well.
INA125 / INA128
For analog output (to feed a microcontroller ADC directly), an instrumentation amplifier like the INA125 (includes internal voltage reference and bridge excitation) or INA128 is a cleaner approach.
Gain is set by a single external resistor: G = 1 + (49.4kΩ / Rg)
For G = 500: Rg = 49.4kΩ / (500-1) ≈ 99Ω → use 100Ω.
Output voltage at full load: Vout = G × Vbridge = 500 × 10mV = 5V (matches 5V ADC reference).
Noise and Layout Considerations
At gain 128, the HX711 amplifies everything 128 times -- including 50/60Hz interference, power supply noise, and radio-frequency pickup.
Key practices:
- Run the four bridge wires together (twisted pair or flat cable with shield)
- Keep the amplifier board physically close to the load cell
- Use short wires from the load cell to the amplifier; long runs pick up noise
- Decouple the HX711 VCC pin with a 100nF capacitor as close to the pin as possible
- Average multiple readings in firmware -- 10 reads averaged out reduces random noise effectively
Simulating a Strain Gauge Circuit
Sketch the Wheatstone bridge in CircuitDiagramMaker with the four resistors and an instrumentation amplifier symbol. Set one resistor to R+ΔR to simulate the loaded condition and run DC analysis. The simulator will show the exact differential voltage across the bridge midpoints, which lets you confirm the amplifier gain setting before ordering parts.
Create Your Own Strain Gauge Circuit Diagram
- Place the four Wheatstone bridge arms, labeling the active gauge(s)
- Add the excitation voltage source and instrumentation amplifier
- Show the HX711 or INA125 connections with pin labels
- Document wire colors from the load cell connector
- Export as a reference for assembly and troubleshooting
Create your own strain gauge circuit diagram -- free
Common Strain Gauge Failure Modes and How to Diagnose Them
Most strain gauge circuit problems fall into a handful of categories, and a multimeter (with the bridge unpowered) will identify most of them before you start chasing amplifier settings or firmware.
| Failure mode | Symptom | Multimeter check | Likely cause |
|---|---|---|---|
| Broken gauge (open circuit) | No output, or output pinned at one rail | Resistance across the gauge reads infinite/OL instead of nominal (e.g., 120Ω or 350Ω) | Gauge grid physically broken, a snapped lead wire, or a cracked solder joint |
| Bonding failure | Output drifts over time, or jumps erratically under a steady, unchanging load | Resistance may still read close to nominal -- a bonding failure often will not show up on a simple resistance check | Adhesive has debonded from the substrate, so strain in the part is no longer transferred to the gauge grid |
| Bridge imbalance | A nonzero output voltage with no load applied (an offset) | Compare resistance across all four bridge arms; a mismatched arm reads noticeably different from the nominal gauge resistance | Fixed-resistor tolerance, unequal lead wire lengths, or a gauge with a different nominal resistance than the others |
| Gauge shorted to the structure | Output pinned low or erratic, sometimes only under vibration or moisture | Resistance from a gauge lead to the metal structure (case) is low instead of very high (megohms) | Damaged or missing gauge coating letting moisture or a conductive surface bridge the gauge to ground |
Always disconnect excitation before probing a bridge with a multimeter in resistance mode -- measuring resistance on a live, powered bridge gives a misleading reading and can affect the measurement circuit.
Gauge Factor and a Worked Strain Calculation
Gauge factor (GF) is defined as the ratio of the relative resistance change to the strain that caused it:
GF = (ΔR / R) / ε
Rearranged, the resistance change for a known strain is ΔR/R = GF × ε.
Worked example: A 350Ω gauge with a gauge factor of 2.1 is bonded to a steel beam. Under load, a multimeter reads the gauge resistance as 350.29Ω, an increase of 0.29Ω.
- ΔR/R = 0.29 / 350 = 0.000829 (0.0829%)
- ε = (ΔR/R) / GF = 0.000829 / 2.1 = 0.000395, or about 395 microstrain (µε)
That result is a typical working strain level for structural steel, which usually runs from a few hundred to a few thousand microstrain under normal service loads.
Lead Wire Resistance and the 3-Wire Quarter Bridge
In a basic 2-wire quarter bridge, the two wires connecting a remote gauge to the bridge circuit add resistance in series with the active gauge arm only. That added resistance changes with cable temperature and length, and because it sits in just one arm of the bridge, it unbalances the bridge and shows up as apparent strain even when nothing has changed at the gauge itself.
The 3-wire hookup fixes this by splitting the lead resistance across two adjacent arms of the bridge instead of concentrating it in one arm: one wire feeds excitation to the gauge, a second wire returns from the gauge to the active-arm corner, and the third wire runs from that same gauge terminal to the adjacent completion-resistor corner. Because equal changes in adjacent bridge arms cancel out in the bridge's differential output, the lead resistance -- including its own drift with temperature -- no longer shows up in the measurement.
The 3-wire hookup is the standard recommendation for any quarter-bridge installation where the gauge sits more than a few feet from the bridge completion circuitry, such as gauges on rotating shafts, remote structural test points, or embedded sensors with long cable runs back to a data acquisition unit.
Key Takeaways
- Strain gauges produce tiny resistance changes (fractions of an ohm); a Wheatstone bridge converts this to a differential millivolt signal.
- Quarter bridge uses one active gauge; half bridge uses two; full bridge uses four. More active gauges = higher output and better temperature compensation.
- Commercial load cells are full bridges with standard color coding: Red = Exc+, Black = Exc-, White = Sig+, Green = Sig-.
- The HX711 is the standard low-cost solution -- 24-bit ADC plus instrumentation amp in one chip, communicates over a simple 2-wire serial protocol.
- Use a clean, stable excitation voltage; noise on the excitation supply appears amplified in the output.
- Layout matters: keep wiring short, keep bridge wires together, and add firmware averaging to reduce noise.
Frequently asked questions
What size strain gauge should I use for a given application?
Gauge length depends mainly on the material's grain structure and stress gradient. Fine-grained, uniform materials like steel and aluminum commonly use small gauges, often 3-6mm active length. Coarse materials such as concrete or composites need longer gauges, sometimes 10mm or more, so the gauge averages strain over enough material to give a representative reading.
Can a strain gauge be reused after removal from a test piece?
No. Strain gauges are bonded to the test surface with adhesive, and removing a gauge damages the thin metal foil grid or its backing. Once debonded, the gauge no longer reads accurately even if reattached. Strain gauges are treated as single-use, single-installation components, not reusable instruments.
What's the difference between a strain gauge and a load cell?
A strain gauge is the individual resistive sensing element that changes resistance with strain. A load cell is a complete engineered assembly -- typically four gauges wired as a full bridge inside a machined metal body -- designed to convert an applied force into a calibrated electrical output, ready to bolt into a weighing or force-measurement system.
Can strain gauges be used in wet or underwater environments?
Yes, but the gauge, lead wire connections, and solder joints need a waterproofing treatment such as a butyl rubber pad, silicone coating, or a potting compound rated for submersion. Without protection, moisture can bridge the gauge circuit to the test structure or corrode the fine foil grid, causing drift or a shorted reading.
Do strain gauges need to be calibrated before use?
The manufacturer's stated gauge factor is usually accurate enough for general-purpose measurements. For precision force or weight measurement, though, the complete assembly -- gauge, bonding, wiring, and amplifier -- is typically calibrated against a known reference load, since bonding quality and lead resistance both affect the system's overall sensitivity.
Can I connect multiple strain gauges to the same amplifier channel?
Not directly as separate bridges. One amplifier channel reads one Wheatstone bridge, which may already contain one, two, or four active gauges wired as a quarter, half, or full bridge. To read multiple independent bridges, use a separate amplifier channel for each, or a multiplexer that switches one channel between several bridges in sequence.