| 摘要 |
This description relates to digital-analog conversion permitting high amplitude precision even at a low number of quantisation stages. This is achieved by the super-sampling and filtering of the quantisation noise. The quantisation precision which is generally limited by process tolerances is compensated in the design proposed here by an additional logic. This permits the accurate linearity correction of an inaccurately given quantisation characteristic. The design is also made flexible by the fact that d/a conversion can be performed at a relatively low signal frequency with high resolution while even a high signal frequency with a coarser quantisation can be converted. The facility to perform any necessary analog filtration with steep edges digitally to economise on analog components is an additional advantage. |
