I want to use FFT to process the audio signal(AUX interface), so as to get the amplitude of 7 fixed frequency points(63Hz, 160hz, 400Hz, 1kHz, 2.5KHz, 6.25khz, 12khz). Can s12zvl32 realize this application? Please provide the sample code if possible.
we are very sorry but we were not able to find anything for this MCU related to FFT.
I have contacted more department and guys and there is nothing everywhere...the reason is that this MCU was designed for different purpose than you want to use so application notes and libraries do not contain any support for FFT.
I believe you can use whatever MCU, even S12Z, but unfortunately there is no of-the-shelf code to share. Once you find the right example in C, I believe it's not so difficult to port it.
There are 3 demo apps (meterlibFFTXph_test) in the SDK 2.9 that I mentioned below. The metering functions use the basic FFT functions. So, you can extract their API from headers and use them in your app.
Generally, you can use any other MCUs, but for example KM3x has MMAU unit which accelerates the FFT computing.
S12Z core is capable of MAC operations thanks to its instruction set: see CPU S12Z Reference Manual - Reference Manual
However, the core performance is basically given by the clock settings and S12ZVL is not one of those high-end MCUs. But this is up to you anyway.
The biggest challenge from my side would be to move the entire (e.g. MATLAB-based) FFT algorithm to Q1.15 fractional format. Therefore, I'd recommend to either look for Q1.15 implementation example or look for another MCU with hardware floating point unit (e.g. S32K14x). Then it would be mainly about the total system costs (discrete solution with more complex PCB vs integrated solution).
Appnote: AN4255 (basic), AN4847 (optional)
Projects: 1-and 2-phase FFT-based power meter RDs. Unfortunately, the application SW (including FFT library) was removed from the project web page several years ago, I don't know why: https://www.nxp.com/design/designs/two-phase-power-meter-reference-design:TWO-PHASE-POWER-METER
Anyway, you can obtain the FFT-library (binary form + API) from the latest SDK_2.9.0_MKM35Z256.
Be aware that: