[hpsdr] PowerSDR vs. cuSDR bandscope width

[Jeremy McDermond]

On Nov 22, 2012, at 12:18 PM, Glenn Thomas <glennt at gbis.com> wrote:

> ***** High Performance Software Defined Radio Discussion List *****
> 
> Restating John's question, how is it possible to recover 186 KHz of bandwidth from a set of samples with a Nyquist frequency of only 98 KHz?
> 
> My uneducated guess (from the 40,000 foot level) is to observe that HPSDR isn't really using a 192 KHz sample rate. The HPSDR hardware uses a sample rate of many tens of MHz. Presumably, the cuSDR display is driven by data derived from this much higher sample rate.
> 
> As John suggested, PSDR assumes that it has an analog front end and that the sample rate really is 192 KHz, so the PSDR pan display can't be any wider than the Nyquist frequency..
> 
> Corrections or confirmations from someone who really does know?

The hardware samples RF at the ADC sample rate of 122Msps.  The FPGA then decimates this down to the 192KHz sample rate seen by the computer.  But, where the ADC is putting out real samples, the FPGA turns those into complex samples.  That means your computer sees 192K complex samples per second.  Each complex sample is essentially two samples, so you're able to recover 192KHz of bandwidth off of a 192Ksps complex signal.

> 73 de Glenn wb6w

--
Jeremy McDermond (NH6Z)
Xenotropic Systems
mcdermj at xenotropic.com




 1353617016.0