RE: [Discuss-gnuradio] USRP Update

[Weber, Michael J. \(US SSA\)]

Ah, that makes sense! Thanks for the clue-in. 

BTW, I didn't see your response come though the list... does the list
engine not send one emails already cc'd to one?

Thanks,
Mike

> -----Original Message-----
> From: Matt Ettus [mailto:address@hidden 
> Sent: Tuesday, December 06, 2005 4:51 PM
> To: Weber, Michael J. (US SSA)
> Cc: discuss gnuradio
> Subject: Re: [Discuss-gnuradio] USRP Update
> 
> 
> Weber, Michael J. (US SSA) wrote:
> > Matt, can you elaborate on the 20MHz bandwidth figure for 
> the Flex400 
> > and give some USRP application guidance? I have this "8MHz maximum 
> > supportable bandwidth" figure in my head which I think came 
> from the 
> > sampling/decimation rates + USB2 throughput limitation discussed in 
> > the USRP documentation..? Some folks have seen the release 
> below and 
> > are asking me if we can do XYZ in terms of bandwidth and 
> I'm not sure 
> > what to say at this point.
> > 
> > Is there a way to move 20MHz-wide signals between the host 
> machine and 
> > the Flex400?
> 
> 20 MHz bandwidth indicates the width of the baseband filters. 
>  Thus, you can tune digitally anywhere +/- 10 MHz from the 
> LO.  If you use 16 bit samples, you can't get that much over 
> the USB bus at once.
> 
> So why is 20 MHz useful?  Here are 3 example scenarios:
> 
> - You could simultaneously receive or transmit 2 separate 
> signals which are 4 MHz wide each, which are separated by 12 MHz.
> 
> - You can use 8 bit samples, and get 16 MHz of bandwidth.  If 
> you were a radio astronomer, you could use 4-bit samples and 
> get 32 MHz of bandwidth.
> 
> - You could do a wideband mod and demod in the FPGA.  You 
> would be able to deal with much wider signals, and send a 
> lower bit rate over the bus.
> 
> Matt
>