- Spurious Free Dynamic Range (SFDR)
- The targeted purity of the tone produced by the DDS. This sets the output width as well as internal bus widths and various implementation decisions.
- Frequency Resolution
- Specified in Hz, this specifies the minimum frequency resolution and is used to determine the phase width, as employed by the phase accumulator and its associated phase increment (PINC) and phase offset (POFF) values. Small values give high frequency resolution and require larger accumulators. Larger values reduce hardware resources. Depending upon the choice of noise shaping, the phase width can be increased, and the frequency resolution higher than that specified. For rasterized mode, the frequency resolution is fixed by the system clock, the number of channels, and the modulus selected. See Theory of Operation for details.
- Noise Shaping
- This parameter controls whether phase truncation, dithering, or Taylor
series correction is used. The options are:
- None
- Phase truncation DDS is produced.
- Dithering
- Phase dither is used to improve SFDR at the expense of increased noise floor. See Phase Dithered DDS.
- Taylor Series Corrected
- Sine/cosine values are interpolated using the otherwise discarded bits from phase truncation (Integer data type). See Taylor Series Corrected DDS.
- Partitioning
- Sine/cosine values are reconstructed from a combination of course and fine tables (Floating Point data type).
- Auto
- Noise shaping is automatically determined, based on system parameters such as SFDR. The selected noise shaping option is presented in the core summary pages. Auto is only available when the parameter selection is System Parameters.
The availability of particular noise shaping options depends upon the configuration option selected and the parameter selection method. System parameter entry automatically constrains whether a particular noise shaping option is possible. When hardware parameter entry is selected, the options summarized in the following table are made available, and the choice of the noise shaping option then constrains the hardware parameter ranges to those supported by the selected option.
Table 1. Availability of Noise Shaping Options for Hardware Parameters Setting DDS Part Phase Generator Part SIN/COS LUT Part None Available Available Available Dithering Available Partitioning Available Available Taylor Available Available Auto Available
Based on the system parameters entered and the noise shaping selected, the minimum phase width and output width are derived by the Vivado IDE in the following way. The phase width can be increased to enable a particular noise shaping option. For example, Taylor series correction requires a minimum phase width of 12 bits.
| Noise Shaping | Output Width (Integer) |
|---|---|
| None and Dithering |
|
| Taylor |
|
|
|
The following figure shows the regions of SFDR and phase width over which each noise shaping option operates. There are three overlapping regions for None, Phase Dithering and Taylor Series Correction, and deeper levels of shading have been used to show where regions overlap. The darkest region is where all three regions overlap and all three noise shaping options are possible. The lower dashed line signifies that Taylor series correction is only valid for SDFR > 66.0 dB (and not 66.0 dB). Phase width can be increased to maximize the number of noise shaping options for a particular SFDR target.