Simcenter Testing Solutions Single Ended vs Differential Inputs

Ever hear the terms “single ended” and “differential input”? Ever wondered what the difference is? Or how the signal is affected?

It is important to know the difference between single ended and differential inputs as it can greatly affect the quality of your data.

A single ended input measures the voltage difference between a wire and the ground. This difference is then amplified to provide the output.

Single ended inputs can suffer from noise as the wire that carries the signal picks up electrical background noise. The signal on a single ended input could also be subject to ground loops.

For example, a transducer wire that is near a power cable (carrying 110V at 60Hz) could be subject to the EMF (Electric and Magnetic Field) noise from that power cable.

Here an accelerometer wire is connected to the Simcenter SCADAS mobile front end and mounted on an air tank. A power wire is lying on top of the transducer wire.

The graphic below shows what is happening on the wire. On the left, we have the wire (grey box) and the EMF coming in from the side (red arrow). Inside the wire, we see the desired signal on the wire (purple) and the EMF (red, 60Hz) on the wire. On the right we see the output of the wire.

Left: The grey box represents the wire connecting the transducer to the Simcenter SCADAS front end. The purple is the desired signal, the red is the EMF. Right: The output (shown on the Simcenter SCADAS data acquisition) system is composed by the addition of the desired signal and the noise. Bottom Right: An FFT of the output: it is composed of the signal and noise (60Hz) components.So, single ended inputs are susceptible to noise.

Remember, that noise is relative to the signal level. When using transducers that have a high output voltage (accelerometers) the EMF noise may not impact the output. However, if the transducer has a relatively low voltage output (strain gauges), then the noise will have a great impact on the output.

A differential input has no reference to ground. Rather, the differential input carries the signal on two wires, a (+) signal wire and a (-) signal wire. Look at the figure below. The (+) and (-) signal wires are carried along the cable and the EMF is introduced along the side of the cable.

The grey box on the left represents the wire with the (+) and (-) signal wires. The purple waves on the left are the desired signals on the wires. The red waves are EMF noise. The red arrow represents the EMF from a power cord. The purple wave on the right is the resultant after subtracting the signals on the two wires from one another.

The output value is the difference between the signals on the two wires (common mode rejection). Notice the output eliminates the noise and the desired signal doubles. This is because the signal on the (+) wire will be the opposite of the signal on the (-) wire (because the signal on the (-) wire is being multiplied by -1). Therefore, by subtracting the (-) signal wire from the (+) signal wire, the desired signal doubles. The EMF induced noise is the same on each wire (because it comes in from the side). By taking the difference between the two signal wires, much of the EMF noise is rejected.

Demonstration time!

Two strain gauges are affixed to the same bar and hooked up to the Simcenter SCADAS Mobile frontend. One gauge is a full bridge gauge (differential input) and the other is a quarter bridge gauge (single input). See this post on strain gauges.

Question 1: Which gauge reading will have more noise in the time domain? The full bridge or quarter bridge?

Question 2: When a power brick (110V, 60Hz) is held near the gauge wires, which signal will be impacted more by the EMF from the power brick? The full bridge signal or quarter bridge signal?

As the video demonstrates, the quarter bridge gauge (single ended) produces a much noisier reading in the time domain.

The quarter bridge gauge (single ended) is also affected more by the EMF while the differential input gauge rejects the noise.

The graph below shows the FFT of the two strain readings.

The 60Hz EMF and its harmonics affecting the FFT of the signals from the two gauges.

Looking closer, we can see the amplitude of the 60Hz power brick noise is much less on the full bridge than on the quarter bridge.

The single ended input is affected by the power brick noise much more than the differential input.

Using differential input is a great way to reduce EMF noise!

Channel Setup Tips

Need a hint as to whether your sensor is single ended or differential? Under channel setup, add the Coupling column (Tools --> Channel Setup Visibility). Depending on what InputMode you set, the coupling will be automatically applied. Full bridge strain gauges are differential while half bridge gauges, quarter bridge gauges, and ICP sensors are single ended.

If you are setting the InputMode to Voltage AC, Voltage DC, or Active sensor, the coupling will be automatically set to single ended. If you are using a differential sensor you must change the coupling to differential.

Enjoy the benefits of differential inputs!

Questions? Email charles.rice@siemens.com or download the Simcenter SCADAS Brochure.

DSP links:

• On-Demand Webinar: Digital Signal Processing Fundamentals
• Index of Testing Knowledge Articles
• Gain, Range, and Quantization
• Anti-Aliasing Filters
• Overloads
• Averaging Types: What's the difference?
• What is Fourier Transform?
• Time-Frequency Analysis: Wavelets
• Spectrum versus Autopower
• Autopower Function...Demystified!
• Power Spectral Density
• Shock Response Spectrum (SRS)
• Windows and Leakage
• Window Types
• Window correction factors
• Exponential Window Correction Factors
• RMS Calculations
• The Gibbs Phenomenon
• Introduction to Filters: FIR and IIR
• Digital Data Acquisition and Signal Processing Seminar

More Acquisition links:

• Simcenter SCADAS
• Simcenter SCADAS Mobile and SCADAS Recorder Hardware
• Simcenter SCADAS: Main-Secondary Daisychain
• AC DC Coupling
• Ground Loops
• Introduction to Strain Gauges
• Measuring strain gauges in Simcenter Testlab
• Simcenter Testlab: Measuring a String Pot
• How Long Can Simcenter SCADAS Run on Battery?
• Simcenter Testlab Signature
• Hands-free acquisition (almost)
• Capturing Transient Events in Simcenter Testlab
• Cool Channel Setup tricks for Triaxial Accelerometers
• Simcenter Testlab: Working with Charge Accelerometers
• Data Acquisition Based on GPS Location
• How to calculate a VECTOR SUM
• Tips and tricks for acquiring torsional orders
• No more tach! Use OBDII to get engine and vehicle speed...
• Record Raw CANBUS Data and Decode Offline!
• Single Plane Balancing in Simcenter Testlab
• Correcting a bad calibration factor
• Simcenter Testlab Calibration
• Simcenter Testlab SCADAS and Long Strain Cables
• Selecting the Excitation Voltage for a Strain Gauge