1. Home
  2. Field-Programmable Gate Arrays
  3. Digital AM/FM Receiver - Systemic Question

Digital AM/FM Receiver - Systemic Question

Hi There! I did start a post earlier with the title " Digital AM/FM Receiver", and I'm starting one again in continuation to that (the reason for starting another post is selfish....get more visibility....I know...I know...I should be a better person, but I'll wait for next year to make that resolution!!) Anyways, the responses to my earlier post were very informative and with them I've been able to get to a point where I am downconverting from IF (thanks for the suggestion Ray), doing decimation (CIC + invSinc filter) and basic lpf to sharpen the band of interest. Now I plan to use CORDIC to demodulate FM. I was reading up on the CORDIC core and it seems that for ATan(y/x) it takes inputs with 1QN and generates phase outputs as 2QN. The output of my LPF is 42 bits wide 2's complement data and that needs to be fed as the X & Y inputs (obviously 1 lpf for I and one for Q), I want to operate the CORDIC with 16-20 bits of input, I was wondering how I should truncate/round. Do I have to convert since the data is already in 2's complement format and if I just truncated....can I just use the upper bits of the lpf output as the input to the CORDIC?

Ideas...a*@ kicking much appreciated!! cheers

-M

Reply to
morpheus
Loading thread data ...

Depends. simple truncation will introduce a DC offset on your signal. You can use symmetric rounding instead (round x.5 toward or away from 0) to eliminate the bias, or in some cases you can subtract out the bias later. For CORDIC, you can keep extra LSBs through the CORDIC processor and then round them off when done to greatly reduce the bias rather than going through the expense of doing a symmetric round at each stage.

Reply to
Ray Andraka

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.