Data Acquisition System (DAQ) Block Diagram
This is a free printable data acquisition system block diagram: download the diagram as SVG or open it and print to paper or PDF.
A data acquisition system block diagram maps the signal path from physical sensors through conditioning, analog-to-digital conversion, and processing to storage or display, showing how real-world measurements become digital data.
A data acquisition system (DAQ) is an assembly of hardware and software that measures physical phenomena — temperature, pressure, voltage, strain, flow, vibration — converts them to digital values, and passes those values to a computer or embedded processor for logging, analysis, or control. The block diagram reveals the functional chain: each block represents a stage in signal transformation.
The first block is the transducer or sensor. A thermocouple converts temperature to a millivolt-level differential voltage; a piezoelectric accelerometer converts vibration to charge; a strain gauge converts mechanical deformation to a resistance change. The transducer's output is always an analogue signal, often at a very low level and mixed with noise.
The second block is signal conditioning. Because raw sensor signals are weak, noisy, or in a form unsuitable for conversion, conditioning amplifies (instrumentation amplifiers for differential signals), filters (low-pass anti-aliasing filters are mandatory before sampling), converts (charge amplifiers for piezoelectric sensors, Wheatstone bridge excitation for strain gauges), and isolates (galvanic isolation protects the DAQ from high common-mode voltages or ground loops). Anti-aliasing filters must have a cut-off frequency below half the sampling rate (Nyquist criterion) or aliased artefacts corrupt the data irreversibly.
The third block is the multiplexer (MUX). Most DAQ systems measure multiple channels with a single ADC to reduce cost. The MUX switches through channels sequentially. Sequential scanning introduces a time skew between channels, which matters for phase-sensitive measurements; simultaneous-sampling DAQ systems provide a separate sample-and-hold per channel to eliminate this.
The fourth block is the sample-and-hold (S/H) circuit. It captures the analogue value at an instant and holds it steady while the ADC converts it, preventing aperture errors.
The fifth block is the analog-to-digital converter (ADC). Resolution (bits) and sample rate (samples per second, Sa/s) are the primary specifications. A 16-bit ADC resolves 1 part in 65 535 of the input range. Common architectures are successive approximation register (SAR) for medium speed and high resolution, and delta-sigma for very high resolution at lower speeds.
The final blocks are the digital interface (USB, PCIe, Ethernet, SPI/I2C for embedded), the data buffer or FIFO, and the host computer running acquisition software. The block diagram may also show a digital-to-analogue converter (DAC) on a feedback or stimulus output path, making the system capable of both measurement and excitation.
How to wire data acquisition system block diagram
- Define the measurement requirements List every physical quantity to be measured, its expected range (e.g., 0–150°C, ±500 mV, 0–10 bar), required accuracy, and the highest frequency component of interest. These parameters drive every subsequent hardware choice. Underestimating frequency content leads to aliasing; overestimating requires unnecessarily expensive hardware.
- Select appropriate transducers for each channel Match the transducer type to the measurand: thermocouples or RTDs for temperature, load cells or strain gauges for force, piezoelectric accelerometers for vibration, current transformers for AC current. Note the output type (voltage, current 4–20 mA, resistance, charge), output range, and supply requirements for each transducer.
- Design the signal conditioning stage For each channel, specify amplification gain to map the transducer's output range to the ADC input range (e.g., 0–5 V or ±10 V). Add a low-pass anti-aliasing filter with cut-off at no more than 40% of the sample rate. For 4–20 mA current loop sensors, add a precision shunt resistor. For thermocouples, add cold-junction compensation.
- Choose the ADC and multiplexer configuration Select sample rate at least 2× (preferably 5–10×) the highest signal frequency. Choose bit depth to match required resolution after accounting for noise floor. Decide between multiplexed scanning (lower cost, introduces inter-channel time skew) and simultaneous sampling (required for phase-coherent multi-channel measurements, for example vibration analysis or power quality).
- Select the digital interface and host platform USB (USB 2.0 full-speed: 12 Mbit/s) suits benchtop DAQ at moderate sample rates. PCIe or PXI suits high-speed, deterministic applications. Ethernet (LXI standard) suits networked or remote acquisition. Embedded SPI/I2C ADCs connect directly to microcontrollers. Consider the required sustained data throughput: a 16-channel, 16-bit, 100 kSa/s system produces 3.2 MB/s of raw data.
- Wire and ground the system Use shielded twisted-pair cable for all analogue signals; connect the shield at one end only (the DAQ end) to avoid creating a ground loop. Route analogue and digital cables in separate cable trays. Use a single-point star ground for all analogue commons. Power analogue and digital sections from separate regulated supplies where possible.
- Calibrate and validate Apply known reference inputs (precision voltage source, calibrated temperature reference) at the transducer input and verify the digital readings at the host. Record the calibration date, reference instrument serial number and calibration certificate number, and the resulting offset/gain correction coefficients. Set a recalibration interval based on the drift specifications of the ADC and signal conditioning components.
Specifications
| Typical ADC resolution range | 12–24 bit (industrial DAQ commonly 16 bit) |
|---|---|
| Sample rate range (per channel) | 1 Sa/s (slow data logger) to >1 MSa/s (high-speed DAQ) |
| Nyquist minimum sample rate | ≥2× highest signal frequency; practical design uses ≥5× |
| Common input voltage ranges | ±10 V, ±5 V, ±1 V, ±100 mV (user-selectable on most instruments) |
| Common-mode rejection ratio (CMRR) | ≥80 dB at 50/60 Hz for instrumentation amplifier inputs |
| 4–20 mA current loop sensor range | 4 mA = 0% of range; 20 mA = 100% of range; 250 Ω shunt gives 1–5 V |
| Isolation voltage (isolated inputs) | Typically 60 V to 1 000 V channel-to-channel, per product specification |
| Relevant standards | IEC 61010-1 (measurement equipment safety), IEEE 1057 (ADC testing), IEC 60751 (RTDs) |
Safety warnings
- Connecting a DAQ input to a live mains circuit without appropriate isolation barriers or high-voltage differential probes is extremely dangerous and will destroy the DAQ hardware. Always verify that the signal voltage is within the DAQ input specification and that adequate isolation is provided for any mains-referenced measurement.
- For any fixed measurement installation connected to mains-powered equipment, a licensed electrician must install and connect the system in accordance with applicable electrical codes (NEC/NFPA 70, BS 7671, IEC 60364, AS/NZS 3000). Always isolate the circuit under test and verify it is dead before connecting signal leads.
- Current transformer (CT) secondary windings must never be open-circuited while the primary carries current. An open CT secondary produces dangerous high voltage. Always short-circuit or load the CT secondary before opening the measurement circuit.
- Electrostatic discharge (ESD) can permanently damage ADC and instrumentation amplifier inputs. Handle PCBs at an ESD-controlled workstation, use wrist straps, and ground the chassis before connecting signal cables.
Tools needed
- Calibrated multimeter (CAT III or CAT IV rated for the voltages involved)
- Precision DC voltage source or calibrator (for ADC calibration)
- Oscilloscope (for verifying signal conditioning bandwidth and noise floor)
- Spectrum analyser or FFT software (for aliasing and noise analysis)
- Shielded cable and appropriate connector tooling (crimper, soldering iron)
- ESD wrist strap and antistatic mat
- Torque screwdriver for terminal connections
- Calibration certificate documentation and logbook
Common mistakes
- Omitting the anti-aliasing filter before the ADC: aliased frequency components fold back into the signal band and cannot be distinguished from real data, corrupting measurements permanently.
- Grounding the cable shield at both ends, creating a ground loop that injects 50/60 Hz noise into the measurement.
- Setting ADC input range wider than necessary (e.g., ±10 V for a sensor that only outputs 0–100 mV), which wastes dynamic range and increases effective noise by the ratio of unused range.
- Ignoring cold-junction compensation for thermocouple measurements: the cold junction (where the thermocouple wire connects to copper at the terminal block) generates its own thermoelectric voltage that must be measured and compensated.
- Using a multiplexed (scanned) DAQ for phase-coherent multi-channel measurements: the time skew between channels is proportional to sample rate divided by channel count and will introduce phase error in vibration, power, or transient analysis.
Troubleshooting
- 50 Hz or 60 Hz noise on all channels
- Cause: Ground loop between sensor, cable shield, and DAQ chassis; or insufficient common-mode rejection Fix: Connect cable shield at DAQ end only. Verify differential input mode is selected in software. Add 10–100 nF bypass capacitors from each input pin to chassis ground at the DAQ connector. If the problem persists, add galvanic isolation to the signal conditioning stage.
- Readings saturate at maximum or minimum ADC value for some channels
- Cause: Signal exceeds ADC input range, or wrong channel gain setting Fix: Measure the actual sensor output with a calibrated multimeter before connecting to DAQ. Adjust signal conditioning gain so the peak sensor output maps to approximately 80% of ADC full-scale range, leaving headroom for transients.
- Apparent frequency components in the measured signal that are not present in the real signal
- Cause: Aliasing — signal contains frequency content above half the sample rate; anti-aliasing filter missing or cut-off too high Fix: Add or lower the cut-off frequency of the anti-aliasing filter to below Fs/2 (preferably Fs/5 to allow for filter roll-off), or increase the sample rate. Verify by temporarily increasing sample rate to confirm whether the spurious components shift in frequency (aliased) or remain fixed (real).
- Channel-to-channel cross-talk on multiplexed DAQ
- Cause: Multiplexer settling time too short; high source impedance causes charge injection from adjacent channel to bleed through Fix: Increase the MUX settling time in software (reduce scan rate). Add a unity-gain buffer amplifier (voltage follower) between the signal conditioning output and the MUX input to reduce the source impedance presented to the switch.
Frequently asked questions
What is the Nyquist criterion and why does it matter for DAQ design?
The Nyquist-Shannon sampling theorem states that a signal must be sampled at least twice the highest frequency component to reconstruct it without aliasing. In DAQ practice, an analogue low-pass anti-aliasing filter must remove signal content above half the sample rate before the ADC stage, because once aliasing occurs it cannot be removed in software.
What is the difference between a DAQ board and a data logger?
A DAQ board (such as a PCI or USB instrument card) relies on a host PC for processing, storage, and software; it offers high sample rates and flexibility. A data logger is a self-contained instrument with on-board memory, battery, and often a display; it is optimised for unattended, long-duration field measurement rather than high-speed laboratory acquisition.
Why do DAQ systems use instrumentation amplifiers rather than ordinary op-amps?
Instrumentation amplifiers (in-amps) have a differential input with very high common-mode rejection ratio (CMRR, typically >80 dB), very high input impedance, and precision gain set by a single resistor. These properties are essential when amplifying millivolt-level sensor signals in the presence of mains-frequency common-mode interference and long cable runs.
What causes ground loops in a DAQ system and how are they prevented?
A ground loop occurs when two parts of a measurement circuit are connected to earth at different physical points, creating a loop. Mains hum or RF interference induces a circulating current in the loop that adds noise to the measurement. Prevention methods include using differential or isolated inputs, routing all grounds to a single star point, and using galvanic isolation in the signal conditioning stage.
How many bits of resolution does a DAQ system need?
Resolution depends on the required measurement precision. A 12-bit ADC resolves 1 in 4 096 (0.024%); 16-bit resolves 1 in 65 535 (0.0015%); 24-bit resolves 1 in 16.7 million. For most industrial measurements 12–16 bits is sufficient. Audio and precision instrumentation often use 24-bit delta-sigma converters. Always account for noise, which effectively reduces usable resolution below the raw bit count.
Related diagrams
- battery management system block diagram
- block diagram of digital communication system
- 12 volt fuse block wiring diagram
- 24 volt charging system diagram
- 3 phase solar system wiring diagram
- battery ignition system diagram