Forum Discussion
Altera_Forum
Honored Contributor
16 years agoHello Kaz,
I got the answer from Altera support. The guy who answered me, said, that my workaround can not work for every signal input, as you said and as I thought. Further he said I should find the maximum of my input samples, either by scaling up to a certain maximum or by just looking htrough the samples. Then if I have read the samples at the FFT output I have to calculate the magnitude of the vector (sqrt(RE²+IM²)) of each bin, serach through the block for the maximum magnitude nad compare it to the maximum input amplitude. With the factor between these two maxima I can scale up every calculated magnitude. This is really frustrating and also slow, because I first have to calulate the magnitude before scaling up. Doing this in hardware is not so easy for me, because i only now the basics of VHDL. It's just more complicated. Thanks for the help kaz.