Multi audio frequencies detector

It is intended to build on every (enough large) FPGA and ASICs as he design is highly parametric. No multiplication nor division library is required.

For easy validation, especially the carries propagation, the additions (and subtractions) are proceed N bits at a time. Some shift registers handle the full length. However, there is no test planned for now with N greater than 1.

The Cordic algorithm is split into stages that proceed only one angle/arc-tangeant. That improves the sample rate, but increases the latency. 

It is designed using modules. More or less modules have to be used according with the requiered properties. The schema choices mostly include the amplitude of signals to be detected, with which "neighbour" signals, in which environement (noise, high amplitude signals etc...). For more information, see the README.modules file.

It detects all the 12 notes of N octaves. The edges of the notes are precise. N can be large if the master clock is high. Fclk=396.N.SampleRate

It is NOT based on the poly-phase which requires a linear progression of the frequencies.

The latency is as short as possible, according to the note (short of high frequencies, long for low frequencies).

This project is in an early stage. The documentation may refer to document not yet written.

The design is a DevOps style. It goes forward to get a working end to end. The detailed tests are going to be done after.

The initial development tests are performed, using a wave viewer. VHDL modules perform extended tests. The run of the tests is done, using scripts rather than mouse actions in a GUI.