Accurate full spectrum test robust to simultaneous non-coherent sampling and amplitude clipping

Sudani, Siva; Xu, Li; Chen, Degang

doi:10.1109/test.2013.6651920

Cited by 8 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, another method that performs accurate and robust spectral test with simultaneous noncoherent sampling and amplitude clipping was presented, thus relaxing the constraints on sampling and input amplitude [10]. The method includes estimating noncoherent, over-range fundamental and removing it from raw data to obtain residue.…”

Section: Fundamental Estimation Removal and Residue Interpolation (mentioning

confidence: 99%

See 1 more Smart Citation

A Comparative Study of State-of-the-Art High-Performance Spectral Test Methods

Sudani

Chen

2015

IEEE Des. Test

Self Cite

View full text Add to dashboard Cite

h SPECTRAL ANALYSIS IS one of the most widely used methods in several areas such as geophysics, oceanography, medical sciences, etc. It provides information about different frequency components present in a signal that are otherwise not observed in time domain. In electronics, it is extensively used to test dynamic parameters of a wide range of semiconductor circuits and systems, such as analog-to-digital converters (ADCs).ADCs are one of the most widely used mixedsignal circuits. Sufficient spectral performance of ADCs is critical for high-speed and high-resolution applications such as communications [1], [2]. To ensure high accuracy in spectral test, the IEEE standards [3], [4] recommend the test setup to satisfy a list of stringent requirements.As ADC resolution and speed become high, the IEEE requirements have become challenging. Furthermore, satisfying these requirements inevitably increases test time and test cost. Relaxing these requirements not only decreases test time/cost but also facilitates on-chip test capability. Several methods have been proposed in the past that relax one or more of these requirements. However, it is not clear which method suits what applications.We provide a comparative study of four methods that relax stringent conditions for spectral test, based on several practical criteria. This enables the user to select the most suited method for his/her situation.Ideal DFT-based spectral test Figure 1 shows the ideal discrete Fourier transform (DFT)-based spectral test setup. The IEEE recommended requirements include the following.

show abstract

Section: Fundamental Estimation Removal and Residue Interpolation (mentioning

confidence: 99%

“…A coherently sampled fundamental is generated and the information of harmonics and noise at each coherently sampled point is obtained by interpolating the residue. A detailed description of the method along with measurement results is provided in [10]. The test procedure is given as follows.…”

Section: Fundamental Estimation Removal and Residue Interpolation (mentioning

confidence: 99%

A Comparative Study of State-of-the-Art High-Performance Spectral Test Methods

Sudani

Chen

2015

IEEE Des. Test

Self Cite

View full text Add to dashboard Cite

show abstract

“…There is one solution that takes into account two of these problems, non-coherent sampling and amplitude clipping, and has been combined into an algorithm in [13]. However, this is just for reducing these two requirements and will not work properly when combined with a non-pure input signal.…”

Section: Measurement Resultsmentioning

confidence: 99%

“…All that is left in the signal's residue at this point is the errors created by the ADC. From this residue, it will be possible to reconstruct the ADC spectral characteristics using a similar strategy as is used in [13].…”

Section: Algorithm Flowmentioning

confidence: 99%

“…To begin this process, there will need to be initial rough estimates of all of the parameters. These can be found by the following equations that are originally derived in [13] and are now shown in Once all of the initial rough estimates of the parameters have been found, it will be possible to begin the approximation iterations to obtain better estimates of all of the parameters. To begin the residue of the estimate can be expressed by (3.16 The iterative algorithm for the four-parameter least-square sine-wave fit can be given by (3.17-3.22).…”

Section: Fundamental Identificationmentioning

confidence: 99%

See 1 more Smart Citation