Hi CASPER, I've recently been designing the Corner Turner module.Converts the data shape from [input channel, frequency] to [frequency, input channel].
My understanding of FFT module is not enough, especially the relationship between data and frequency points. fft_biplex_real_2x block:Fft biplex real 2x - Casper (berkeley.edu) <https://casper.astro.berkeley.edu/w/index.php?title=Fft_biplex_real_2x&oldid=7999> This is my understanding of fft_biplex_real_2x block: The FFT of the real number input will generate complex numbers, but with a Hermitian symmetry on the Fourier components, i.e. the negative frequency part is just the complex conjugate of the corresponding positive frequency component, only half of the complex numbers are independent. To avoid wasting memory and data transportation, the input data block of two input channels are combined as the real and imaginary components respectively to form a complex input, and then FFTed together. An EQ module is placed between FFT module and Transpose module. I don't think I understand the 'frequency point' enough, so the following question may itself be wrong. My question is: 1.The fft_biplex_real_2x block sets fft size = 2^11. Its two outputs can be viewed as the output of four real streams, each containing 1024 frequency points. Is that right? 2.I designed 4 transpose modules, each responsible for a quarter frequency point of 1024. Each module corresponds to a 10gb module that packages and outputs the data. I'm curious, the fft output contains 1024 frequency points.How to divide the real stream containing 1024 frequency points? Examples: transpose_1:0-255, transpose_2:256-511, transpose_3:512-767, transpose: 768-1023. Does anyone know the F-engine design article? It is better to have fft and cornerturn module design ideas, I would like to read it! I really hope someone can communicate with me, I would appreciate it! BW! Wang -- You received this message because you are subscribed to the Google Groups "casper@lists.berkeley.edu" group. To unsubscribe from this group and stop receiving emails from it, send an email to casper+unsubscr...@lists.berkeley.edu. To view this discussion on the web visit https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/5d706944-3833-4641-adfd-fca688aee6bfn%40lists.berkeley.edu.
