MODELLING OF PREDICTIVE SIGMA-DELTA CONVERSION |
|---|
| F. Sandu, W. Szabo |
- Abstract:
- The idea of acquiring not the proper samples at the input of an ADC but the difference ("Delta" and its sign "Sigma") between the new sample and the previous one (the derivative of the signal, in finite-difference sense) led to an important decrease of samples’ dynamic, provided the over-sampling frequency goes towards the slew-rate of the conversion circuits divided by the last significant bit. Considering the previous sample as the simplest (1-st order) estimate of the new sample, the authors modeled a more efficient way (in terms of speed and memory requirements): the acquisition of the difference between the new sample and a better estimate of it (higher order estimate, adapted to signal’s statistic).
The Levinson-Durbin algorithm is justified and implemented into an Itakura lattice. Computational similarities with the Schur-Cohn tests of stability are underlined. The authors implemented this computation scheme in a LabView “virtual ΣΔ ADC”.
Modelling by Mason graphs of the Itakura finite-impulse-response (FIR) of the predictive ΣΔ-ADC is followed by its symmetric flow-graph for the infinite-impulseresponse (IIR) DAC. - Keywords:
- linear prediction, sigma-delta converters, flow-graphs
- Download:
- IMEKO-WC-2000-EWADC-P645.pdf
- DOI:
- -
- Event details
- Event name:
- XVI IMEKO World Congress
- Title:
Measurement - Supports Science - Improves Technology - Protects Environment ... and Provides Employment - Now and in the Future
- Place:
- Vienna, AUSTRIA
- Time:
- 25 September 2000 - 28 September 2000