1 / 30

Review of Progress in Quantitative NDE 2007 Colorado School of Mines Golden, Colorado

Review of Progress in Quantitative NDE 2007 Colorado School of Mines Golden, Colorado July 22 – July 27, 2007 Digital Audio Signal Processing and NDE: an unlikely but valuable partnership Patrick Gaydecki School of Electrical and Electronic Engineering The University of Manchester

solomon-henson
Download Presentation

Review of Progress in Quantitative NDE 2007 Colorado School of Mines Golden, Colorado

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Review of Progress in Quantitative NDE 2007 Colorado School of Mines Golden, Colorado July 22 – July 27, 2007 Digital Audio Signal Processing and NDE: an unlikely but valuable partnership Patrick Gaydecki School of Electrical and Electronic Engineering The University of Manchester PO Box 88 Manchester M60 1QD United Kingdom patrick.gaydecki@manchester.ac.uk [UK-44] (0) 161 306 4906 www.eee.manchester.ac.uk/research/groups/sisp/research/dsp www.signalwizardsystems.com

  2. Characteristics of Real-Time DSP Systems • DSP offers flexibility, allowing a single platform to be rapidly reconfigured for different applications • Operations such as modulation, phase shifting, signal mixing and delaying are simply performed in software • System performance is far more accurate than equivalent analogue systems • However, considerable intellectual investment is required to design and program DSP platforms

  3. The Sound Transduction Process: sound energy  electrical signal  binary  processed binary  electrical signal  sound energy PC… amplifier 10100100100100100111111010101010010010101011010101001010110110101001011010010101010010100101001110100101010100101010010101 digital to analogue converter analogue to digital converter amplifier loudspeaker system …or DSP system

  4. FIR Performance Figures for DSP56309 and DSP56321

  5. DSP Systems at UoM: Generations II and III Processing Repertoire Generation II Generation III • Standard filters, e.g. Butterworth etc. • Arbitrary FIR filters • Arbitrary IIR filters • Adaptive filters • Inverse filters • Echo • Real-time gain control • Mixing • Phase delays • Time delays • Real-time FFT and waveform capture • Sine and arbitrary wave synthesis • Standard filters, e.g. Butterworth etc. • Arbitrary FIR filters • Arbitrary IIR filters • Adaptive filters • Inverse filters • Echo • Real-time gain control • Mixing • Phase delays • Time delays • Real-time FFT and waveform capture • Hilbert transform • Quadrature Signal Processing • Envelope detection • Sine and arbitrary wave synthesis • Tone generation • Shaped noise generation • Modulation, encoding and decoding • Vocoding • JTAG support • 3rd party software support (C++ design tools)

  6. Signal Wizard 2 Hardware Concept Serial Interface DSP Core Control system Flash memory 24-bit Dual Channel Codec

  7. Signal Wizard 2 Hardware

  8. Signal Wizard 2 Software Graphical display of filter FIR and IIR design area Hardware control: download, gain, adaptive, delay, mixing etc.

  9. Signal Wizard 3 Hardware Concept JTAG Interface USB Interface Parallel Interface Control system DSP Core Flash memory S/PDIF interface 24-bit multichannel (6 in 8 out) 200 kHz Codec

  10. Signal Wizard 3 550 million multiplications and additions per second

  11. Signal Wizard 3 Software

  12. Basic Linear Filter Theory: Property / Filter type FIR IIR

  13. Noise in audio can be classified into two types: narrowband (easy to remove) or broadband (difficult) easy narrow band noise amplitude audio signal frequency difficult amplitude audio signal broad band noise frequency

  14. Finite Impulse Response (FIR) Filter Design using the Frequency Sampling Approach Arbitrary filter impulse response h[n] obtained by: Output signal y[n] obtained by discrete-time convolution:

  15. Phase Control Precise phase control for each harmonic, to a resolution of 0.0001 degree. Application: real-time Hilbert transform.

  16. Infinite Impulse Response (IIR) Filter Design

  17. Digital Emulation of Analogue Networks I: Laplace to z-domain mapping s-domain z-domain

  18. Digital Emulation of Analogue Networks II: The Bilinear z-transform (BZT) Substitutions: Final difference equation:

  19. Digital Emulation of Analogue Networks III: Real-Time Performance Design Realised

  20. IIR Comb Filters and Pole-Zero Placement

  21. Use of Super Narrowband Filters in the Detection of Low Amplitude Ultrasonic Pulses Propagated through Seawater via a Steel Structure (Rito Mijarez) 7 m structure being lowered into the dock at Liverpool, UK Location of transmitter

  22. digital gain (x 2048) DSP super narrowband filter Experimental Configuration Digital Oscilloscope Instrumentation amplifier (x 400) Microcontrolled pulser 100 m Receiving transducer Transmitting transducer

  23. Typical Results (a) (c) • Detail of original received signal degraded by noise. • Detail of received signal, recovered by super narrowband filter. • Complete tone burst signal detected after transmission through water, recovered using a super narrowband IIR filter. (b)

  24. Signal Shape Reconstruction Essential Equations Describing Time Domain Deconvolution (Inverse Filtering) for Real-Time and Off-Line Processing Simple Fourier domain scheme (rarely successful): Fourier domain scheme with noise estimate: Time domain scheme with noise estimate (surprisingly useful): Finally:

  25. Signal Shape Reconstruction in Practice Signal comprising three impulses Signal after low-pass distortion Signal after inverse filtering in real-time

  26. Loudspeaker Equalisation • System operation: • Signal Wizard software generates a swept sine or white noise test signal and downloads it to the hardware. • The hardware sends the signal to the speaker, simultaneously recording its response. • Signal Wizard software analyses the response and generates an equalization filter based on deconvolution. • The hardware convolves the filter with the test signal and the analysis is repeated. DSP system amplifier loudspeaker amplifier

  27. Loudspeaker Equalisation Results Before After

  28. Noise in audio can be classified into two types: narrowband (easy to remove) or broadband (difficult) easy narrow band noise amplitude audio signal frequency difficult amplitude audio signal broad band noise frequency

  29. Recovered signal Adaptive Filters Input signal + noise + _ FIR filter section with tap modifiers (LMS) Adaptive filter theory is complex but its implementation simple. It is ideally suited to real time DSP systems. amplitude broadband noise frequency

  30. Conclusions • Flexibility ensures that real time DSP is suitable for many applications in both real time NDE and audio signal processing • Software-realized processing yields very significant improvements with regard to stability, precision and repeatability • New generation DSP devices are extending the range of signal frequencies over which real time discrete processing can be applied

More Related