Using Low-Cost Transducers for Ultrasonic Sensing Applications

Signal conditioning prepares signals before acquisition by removing noise and adjusting characteristics. Noise originates from internal equipment or external environments, including 60 Hz power line interference common in North America. Two primary noise reduction techniques exist: amplification (increasing signal magnitude so noise becomes negligible) and filtering (removing unwanted frequencies). Low-pass filters allow low frequencies while blocking high frequencies; high-pass filters do the opposite; bandpass filters permit only specific frequency ranges. Signal conditioning occurs at multiple stages throughout the data acquisition process, involving sensors, microcontrollers, and later data analysis.

Noise can be specified as peak-to-peak (worst-case, time-dependent), RMS (quadratic mean, better for comparisons), or Power Spectral Density (frequency distribution). For Gaussian noise, RMS equals standard deviation, enabling peak-to-peak estimation (±3.3σ for 99.9% confidence). PSD data in datasheets enables noise estimation in specific frequency bands by calculating area under the curve. Filters reduce noise by attenuating signals outside the passband, with key considerations including filter steepness, transition characteristics, and cut-off frequencies (3dB attenuation points). Advanced signal conditioning devices can perform acceleration compensation by combining force and acceleration sensor data with setup mass to eliminate spurious force signals from movement, allowing true measurement signals to remain unaffected.

Signal conditioning prepares signals for digitalization through amplification, filtering, and level shifting. Common noise sources include 60Hz power line interference, electromagnetic interference, poorly installed transducers, and internal circuit electronics. Fourier series states that any periodic signal can be decomposed into sine and cosine waves at different frequencies. The Fourier transform extends this to non-periodic signals, transforming them from the time domain to the frequency domain. Filtering must be performed before digitalization using analog filters, as ideal filters do not exist. Simple filters use resistors and capacitors, while steeper roll-off requires active filters. Filter complexity affects cost, temperature stability, and potential signal distortion near the cutoff frequency.

Signal conditioning addresses two critical challenges: amplifying weak transducer outputs and eliminating noise. Amplification increases signal levels to match data acquisition requirements but also amplifies noise. Filtering techniques (band-pass, notch filters) selectively pass desired frequencies while rejecting interference. Ground loops create 50 Hz interference when instruments have different ground potentials; solutions include using DC power supplies, ensuring common grounding, and maintaining proper earthing. These techniques ensure clean, usable measurement signals.

Advanced signal conditioning operations include excitation (providing power to passive transducers like strain gauges and RTDs for force, pressure, temperature measurements), linearization (converting non-linear input-output relationships to predictable linear forms), filtering (using low-pass filters to remove noise and improve signal-to-noise ratio), isolation (protecting circuits from transient voltages), and impedance matching (ensuring high-impedance transducers deliver signals without distortion). These techniques enable accurate data acquisition across diverse measurement applications.

Acoustic impedance matching improves transducer performance by adding intermediate layers between the piezoelectric element and load, ideally quarter-wavelength thick. Acoustic impedance equals density multiplied by wave speed, with boundaries causing transmission and reflection. The optimal single-layer impedance targets the geometric mean of the square root of the product of impedances. Electrical impedance matching is equally critical for power transfer, as transducers rarely present ideal 50 ohm impedances. Matching uses passive components (inductors/capacitors) to correct reactance or transformers to adjust resistance. Since impedance is frequency-dependent, perfect matching cannot occur across all frequencies—designers must choose between narrow-band matching (excellent at one frequency) or broadband matching (reasonable across wide range). Drive method selection depends on application: sine waves produce steady-state resonance for high-power outputs, while impulses reveal short pulse characteristics for imaging. Beam patterns exhibit distinct near-field and far-field regions, with drive regime significantly affecting beam shape—continuous wave preserves complex near-field patterns while single-cycle driving reduces variations.

Impedance matching improves transducer performance by adding intermediate layers between the PZT and load. Acoustic impedance equals density multiplied by wave speed, with transmission and reflection determined by impedance ratios. For optimal single-layer matching, target impedance should be the geometric mean of PZT and load impedances. Quarter-wavelength layers are ideal, but practical thicknesses (e.g., 375 microns at 1 MHz in water) require precise fabrication. Similarly, electrical impedance matching is essential for efficient power transfer from typical 50 ohm sources. Electrical impedance is complex (resistance plus reactance), with reactance frequency-dependent. Passive components or transformers correct reactance and adjust resistance. A fundamental trade-off exists: narrowband matching achieves perfect 50 ohm match at one frequency, while broadband matching provides reasonable matching across a wide range.

PCB discontinuities disrupt ideal return current paths and increase EMI. Edge discontinuities require critical signals to stay 20× separation distance from board edges. Layer changes should maintain same ground reference; when unavoidable, connect divided planes with vias near crossings. Connectors removing copper from ground planes force longer return paths. Proper connector design maintains continuous ground planes. Component-level shielding (5-sided enclosures) connects to ground through multiple vias for low impedance. Floating metal parts can receive and radiate EMI through parasitic capacitance. Cable shielding requires 360-degree chassis contact, avoiding ties/pigtails. Twisted pairs with common-mode chokes at entry points provide filtering. Cables should run close to chassis walls and away from seams/apertures.

Characteristic impedance is the ratio of electric to magnetic fields surrounding a transmission line and must remain constant throughout propagation. Factors affecting impedance include track dimensions, distance to reference plane, dielectric properties, and adjacent structures. Signals propagate as electromagnetic fields in dielectric spaces, not within copper. Connector placement strategy significantly impacts EMI: placing all connectors on one board side minimizes antenna structure formation. Cavities formed between stacked boards can resonate at signal harmonics, acting as unintended antennas. Shielded connectors require 360° low-impedance chassis connections to maintain faraday cage effects.

To properly design a PCB with an antenna, the feed line must be designed for 50 ohm characteristic impedance (not resistance) using transmission line theory, typically achieved through microstrip or coplanar waveguide configurations with specific width, height, and spacing parameters calculated using software like AppCAD; this ensures maximum power transfer between the RF transceiver and antenna, and proper ground clearance zones and stitching vias are essential for optimal performance.

Non-intrusive fluid identification uses acoustic fingerprinting to determine vessel contents through metal walls without opening them. Military origins enabled detection of explosives and hazardous materials; now applied industrially for fluid identification. Tank overfill protection systems mount wirelessly at 90% marks, adding 12% capacity without new tanks. Pipeline transmix reduction detects product transitions externally, reducing waste by 50%. These technologies enable instantaneous measurements without invasive sampling lines, transforming operational efficiency across industries.

Flow meters are used in various industries including chemical industry, paper and mining industry, pharmaceutical and water treatment, HVAC, oil and gas field. Specific applications include liquid chemical and pharmaceuticals, steam measurements, inflow of liquid suspension, custody transfer of natural gas metering, general water application, detection of sulfur dioxide and nitrogen oxide in greenhouse gas emissions, measurement of air-to-fuel ratio in boilers and steam generators, monitoring of gas flares in natural gas and biogas plants, and testing of gas valves and pressure regulators for leaks.

This video explains five common methods used in the industrial process industry to measure the concentration of liquid components: density (mass/volume ratio), refractometry (light refraction changes), ultrasonic (sound speed variations), conductivity (electrical current measurement for ionic solutions), and photometry (light absorption analysis). These techniques measure bulk fluid properties arising from collective atomic, ionic, or molecular interactions, with photometry additionally enabling identification of specific compounds through light-matter interactions.

Industrial flow sensors employ three primary measurement principles—thermal dispersion (using heat transfer to measure gas flow), ultrasonic transit time (measuring sound pulse travel time for liquid flow), and magnetic induction (detecting voltage induced by conductive liquid flow)—each offering non-moving part designs for reliable monitoring of liquids and gases in various industrial applications.

This comprehensive discussion covers industrial flow measurement technology, beginning with AW Lake's company history and flow meter categories. Flow meters range from variable area meters (24% accuracy, visual indicators) to Coriolis meters (0.1% accuracy, mass measurement). Coriolis meters work by splitting flow into vibrating tubes where mass creates specific frequencies, enabling measurement regardless of viscosity changes. Applications include fuel quality control (detecting contamination through density changes), battery testing across extreme temperatures (-40°F to 250°F), and chemical manufacturing requiring 0.05% accuracy. Turndown ratio (e.g., 250:1) indicates measurement range capability, with instruments performing best in the middle 80% of their range. Modern data collection provides multi-metric outputs (flow, temperature, density) via pulse, Modbus, or Ethernet signals, replacing multiple instruments with single solutions.