Hi Jason, In addition to Terry's suggestions, I would add that in order to decrease the number of channels in your spectrometer you probably have to make a few changes to your vector accumulator logic as well.
- Andrew On 4/8/09 1:32 PM, "Terry Filiba" <[email protected]> wrote: > Hi Jason, > What version of the library are you using? I haven't had any trouble > reconfiguring the green 7.1 blocks in my packetized astronomy signal > processor design. > Have you tried simulating the design by putting in a test tone through > the adc and a scope on the pfb and fft output? If it won't simulate > correctly you could try replacing the blocks with fresh versions > copied from the Casper library (sometimes the block become corrupt > when they redraw). > Also you should double check the sync period for the new design. There > is a memo on the Casper website that describes how to determing the > sync period for the fft. > -Terry > > On Apr 8, 2009, at 12:48 PM, Jason Ray <[email protected]> wrote: > >> All, >> >> Randy and I are working on a low resolution version of our GUPPi DSP >> designs. Basically, we want to take our existing (and widely used) >> 2048 channel design and reduce it to 128 channels, for the sake of >> faster readout times. We did this by simply changing the mask >> parameters of the PFB and FFT blocks to give us 128 channels. This >> seems like a very simple thing but with the 128 channel design, we >> are getting bad spectra. >> >> Does anyone have any thoughts on this? >> >> Thanks, >> Jason >> >> >> >
