Re: [Discuss-gnuradio] working (in theory) phasing SSB modulator

[cswiger]

On Fri, 18 Feb 2005, Thomas Traber wrote:

> Chuck,
>
> > [...] looks more intensive than the
> > weaver method for now.
>
> I don't own a USRP (yet).
> Isn't it possible just to output the
> complex signal at [combine1] to the USRP
> and to IQ upconversion in the front end?
>
> >                                          +---\   /----+
> >                   (float)     (complex)  |    \ /     |        (complex)
> > [src]-->[bandfilter]-->[hilbert]-->[split1]    X     [combine1]---->
> >                   (af)        (af)       |    / \     |        (af)
> >                                          +---/   \----+
>
>
> O.K. for generation of short wave
> signals you don't need any front end at
> all, when you do upconversion in
> software.
>
>

Don't know. I'm not sure if the usrp can do the funcation required
w/o re-doing the fpga code. You definitely need to bump up the sampling
rate from audio (32k) to something like 1,600,000 or 2,000,000 and
the usrp can get it to DAC rate of 128,000,000 from there.

I played around with the sink.set_mux(0x98) - it looks like the
0x10 bit is the input and the 0x80 and 0x08 are two outputs. If you
set DUC at 3900e3, and send it a signal at 30Khz you get an
output at 3930e3 with mux(0x98). If mux(0x90) or mux(0x18) you get
a signal at 3930e3 and also 3900e3-30e3 (3870e3). If mux(0x00),
mux(0x80) or mux(0x08) you get no signal.

I'll try it and see what happens!

Final note, this amp circuit:
http://www.qsl.net/wa1ion/amp/w7iuv_amp.htm
will boost the usrp tx up to over 10mW.

--Chuck