SLAA889A March   2019  – July 2021 MSP430FR6041 , MSP430FR6043 , MSP430FR60431 , MSP430FR6045 , MSP430FR6047 , MSP430FR60471

 

  1.   Trademarks
  2. 1Flow Metering Background
    1. 1.1 Water Flow Velocity Estimation
      1. 1.1.1 Velocity of Ultrasound in Water is Known
      2. 1.1.2 Velocity of Ultrasound in Water is Unknown
  3. 2Algorithms for Ultrasonic Water Flow Metering
    1. 2.1 Correlation Based Technique for Differential TOF Estimation
    2. 2.2 Absolute Time-of-Flight (AbsTOF) Measurement (Tup and Tdn)
      1. 2.2.1 AbsTOF Calculation
        1. 2.2.1.1 Acquisition Algorithm for AbsTOF
        2. 2.2.1.2 Tracking Algorithm for AbsTOF
  4. 3References
  5. 4Revision History

Algorithms for Ultrasonic Water Flow Metering

In both of the methods described in Section 1.1.1 and Section 1.1.2 the knowledge of the differential TOF as given in Equation 3 is required. There are two techniques for differential TOF measurement:

  • Zero-crossing based using a time-to-digital converter (TDC): This technique uses an initial threshold crossing for the signal, followed by zero crossings of the signal
    GUID-044797D5-56FC-4FF6-B02B-DDD8E3AEAEBE-low.pngFigure 2-1 Zero-Crossing Based Time-to-Digital Converter
  • Correlation based using an analog-to-digital converter (ADC): In this technique the whole waveform is captured using ADC and stored for the up/dn stream and then post processing is done on the waveform to derive the differential time of flight.

This document describes the ADC-based technique for ultrasonic flow metering. This approach was chosen because of the following advantages over TDC techniques:

  • Performance: The correlation acts as a digital filter to suppress noise (implemented efficiently on the low energy accelerator (LEA) on the MSP430FR6047 MCU). This results in 3 to 4 times lower noise standard deviation. The correlation filter also suppresses other interference like line noise.
  • Robustness to signal amplitude variations: The algorithm is insensitive to the received signal amplitude as can occur with high flow rates, transducer-to-transducer variation, and temperature variation.
  • The envelope of the signal is obtained naturally in ADC-based processing: This enables tuning to the transducer frequencies because slow variations in the envelope across time can be used to detect aging of transducers or meters.