Re: [Discuss-gnuradio] subdevice selection in USRP2

[Eric Blossom]

On Tue, May 12, 2009 at 04:28:44PM -0400, Eric Matlis wrote:
> From: Eric Blossom address@hidden
> Sent: Tuesday, May 12, 2009 4:16 PM
> To: Eric Matlis
> Cc: address@hidden
> Subject: Re: [Discuss-gnuradio] subdevice selection in USRP2
> 
> On Tue, May 12, 2009 at 03:30:09PM -0400, address@hidden wrote:
> > Hi all-
> >
> > as part of my on-going efforts to acquire 2 channels (two independent
> > signals) on my USRP2 with an LFRX board, I would like to know if there is
> > an equivalent to the subdevice setup code that exists for the USRP1.  For
> > example:
> 
> 
> >         if not self.u.set_nchannels(nchan):
> >             sys.stderr.write('set_nchannels(%d) failed\n' % (nchan,))
> >             raise SystemExit
> >
> >         self.subdev = self.u.db[0] + self.u.db[1]
> >
> >         self.u.set_mux(gru.hexint(0xf0f0f1f0))
> >
> > I already understand from a previous response on this list that there is
> > only a single DDC in the USRP2 fpga image, and I can at least verify that
> > set_nchannels, db, or set_mux is not defined for the USRP2.  Are there
> > equivalents with different syntax for the USPR2?  Without them or their
> > equivalents, is it even possible to acquire two channels with the USRP2
> > currently?
> 
> It is not currently possible to acquire more than a single channel
> with the USRP2.
> 
> > My thought was I could get arround the existence of only 1 DDC by tuning
> > to the midpoint of my two signals and shifting them independently to
> > baseband in the host pc, but it seems to me I need more than this to
> > properly setup the two signal paths.
> 
> If the two signals are within the same IF passband, then yes, you can
> split them out on the host.  Use the gr.freq_xlating_fir_filter_ccf.
> 
> I have been trying gr.freq_xlating_fir_filter_ccf, but again, physically I 
> have two independent signals,
> not two frequencies in one signal.  Ie, I have two coax cables going to the 
> front of my USRP2, 
> one carrying the AM signal, the other the non-AM signal.  My objective is to 
> capture both.
> 
> So as I understand it this is not possible currently?  

Yes, you understand correctly.

This should be possible after the "great refactoring" that will occur as
part of the USRP2 VRT work.

Eric