Hi >From what I can see of the block internals;
The phase slope generated is in the form phase(channel_index) = slope*channel_index + offset slope = theta_fract offset = theta_fs + samples/fft_fs So setting theta_fract to 0 and theta_fs starts the phase at 0 but the phase will slowly increase (see Fringe_Theat_Cnt1). Regards Andrew On Fri, Feb 20, 2015 at 3:48 PM, Wang Jinqing <[email protected]> wrote: > Hi, thank you > > But I have set the theta_fract zero,this means that the phase changing > every 4095 FFT cycles is zero, So the output phase should be stable. But > now is not like that. > > Best Regards. > > Oliver Wang > > > -----原始邮件----- > *发件人:* "Andrew Martens" <[email protected]> > *发送时间:* 2015年2月20日 星期五 > *收件人:* "Wang Jinqing" <[email protected]> > *抄送:* "[email protected]" <[email protected]> > *主题:* Re: [casper] question about finedelay_fstop_prog > > > Hi Oliver > > From what I can see, that port determines how quickly the phase offset > applied to all channels is incremented. So your value tells it to increment > the phase offset every 4095 cycles. I am not sure if there is some way to > turn off fringe-stopping. > > There is some good documentation at > https://casper.berkeley.edu/wiki/Finedelay_fstop_prog which may help. > > Regards > Andrew > > On Fri, Feb 20, 2015 at 3:06 PM, Wang Jinqing <[email protected]> wrote: > >> Hi, >> >> I want to use the core finedelay_fstop_prog to design a FX >> correlator( two stations ) after FFT as the appendix. I have set >> theta_fract and theta_fs all zeros, fft_fs 4095. I think in this design the >> the correlated phase should be constant with time. But when I input a >> sine wave signal then devides into two channels( as two stations' >> signal), the correlated phase is changing with time. I have tested the >> model after I removed the finedelay_fstop_prog core and I found >> the correlated phase is stable. So I think I may set the >> finedelay_fstop_prog's input parameters not correct. But how to set the >> parameters correctlly? >> >> Best Regards. >> >> Oliver Wang >> >> >> >> > > > > >
