Binary Spectrally Sparse Sequence for High-Speed Transfer Function Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for measuring the complex transfer function of systems and substances face challenges with low excitation energy due to high crest factors in multisine signals and uniform energy distribution in broadband measurements, leading to measurement uncertainties and inefficiencies.
Innovation Solution
A method and device that generate a binary spectrally sparse sequence with energy concentrated on distinct frequencies, maintaining a crest factor of unity, allowing for higher dynamic range and signal-to-noise ratio by optimizing the magnitude spectrum and using a comb-like form in the frequency domain.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If simultaneous multisine excitation is used for high speed parallel measurements, then measurement speed is improved, but crest factor becomes very high (up to 10 or more) causing low RMS level and insufficient excitation energy at each frequency
Solution Approach 1:
The frequency spectrum is segmented into multiple discrete frequency lines rather than continuous excitation. The binary sequence is designed to have spectral components only at specific frequencies of interest, concentrating energy at these discrete points while maintaining low crest factor through binary nature (CF=1).
Solution Approach 2:
Excitation energy is concentrated locally at specific frequencies of interest rather than uniformly distributed. The magnitude spectrum is optimized to provide higher energy at frequencies where measurement is needed, while the binary nature ensures low crest factor and high RMS level overall.
2Adaptability or versatility
If pseudorandom rectangular waveforms (MLS) are used to cover wide frequency band, then bandwidth coverage is improved, but energy is distributed almost equally over the whole frequency band leading to low power spectral density at individual frequencies
Solution Approach 1:
The excitation signal provides different energy levels at different frequencies based on measurement requirements. The magnitude spectrum is optimized to concentrate energy at frequencies of interest, providing high power spectral density where needed rather than uniform distribution across all frequencies.
Solution Approach 2:
The spectral characteristics of the excitation signal are dynamically optimized for the specific measurement task. The magnitude spectrum can be tailored to emphasize frequencies of interest, making the measurement process adaptive to the specific system under test and measurement objectives.
3Productivity
If MLS excitation is used, then measurement speed is improved, but useful excitation bandwidth extends only up to 0.45f with significant energy (near 40%) falling onto higher frequencies outside the measurement bandwidth
Solution Approach 1:
The binary sequence is designed with a comb-like magnitude spectrum that concentrates energy only at frequencies within the measurement bandwidth of interest. By carefully selecting the set of excitation frequencies and optimizing their relative magnitudes, energy is localized where needed, eliminating waste at frequencies outside the band of interest.
4Use of energy by moving object
If rectangular waveforms are used to provide more excitation energy, then excitation energy is improved, but crest factor increases significantly above unity value
Solution Approach 1:
The excitation signal parameters are changed from conventional rectangular waveforms to an optimized binary sequence with comb-like spectrum. The binary nature maintains crest factor of unity (CF=1) while the optimized magnitude spectrum provides sufficient excitation energy at all frequencies of interest through proper amplitude distribution across spectral lines.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A method and device for high speed broadband testing of systems and substances using a binary, spectrally sparse sequence (SSS) as a periodic excitation waveform. The sequences with controllable frequency and magnitude spectra content are designed by component manipulation method or by edge manipulation method. The excitation waveform is typically pre-calculated, and kept in waveform memory, from where it is shifted out into digital to physical quantity converter (DQC). The sparse spectrum of the SSS makes it easy to create plenty of uncorrelated frequency sets with adjacent, but sufficiently different frequencies to form multi-path test systems, where all the paths can be measured simultaneously. The response of the sample under test (SUT) is sampled and the complex transfer function is calculated directly or indirectly via Impulse Response by Discrete Fourier Transform technique and its derivatives. The sequence bit interval and sampling interval have a predetermined ratio.