February 11, 2020
The Distortion Introduced by Sampling lesson lists three sources of distortion related to sampling:
- Distortion due to signal bandwidth, as discussed in Section 2
- Distortion due to DAC interpolation, as discussed in Section 3
- Distortion due to quantization, as discussed in Section 4
In each case, increasing the sample rate Fs can improve several characteristics of sampling at the possible expense of other factors, such as system cost and power consumption. To be more specific:
- Increasing the sample rate allows signals to have larger bandwidths.
- Increasing the sample rate results in less error at the output of a sample-and-hold DAC.
- Increasing the sample rate while maintaining an acceptable SNQR decreases the number of required bits per sample.
These statements are true for standard PCM consulting.
The Distortion Introduced by Sampling lesson also introduced a tool for reducing SQNR by encoding the difference (delta) between consecutive samples rather than encoding each sample as-is. This encoding method could be justifiably called “delta encoding” or “delta modulation.”
This section builds upon the methods of over-sampling and delta modulation. It also introduces a third tool: noise shaping.
Noise shaping pushes a large portion of the quantization noise energy into the higher frequency bands while keeping the desired signal energy in the lower frequency bands. The system can then low-pass filter the output. This yields a high energy signal and low quantization noise—that is, a high SNQR.
The delta-sigma encoder is a simple method for accomplishing a high SNQR, as discussed in the next lesson.