Re: [Discuss-gnuradio] IF sampling in ISM band

[Eric Blossom]

On Fri, Jul 04, 2008 at 12:21:25PM +1000, Phaysal Khan wrote:

> Hi List,I am using an RFX2400 d'board, for IF sampling a signal
> which is 20MHz wide, centered at 2.4125GHz. I know it's not possible
> to sample whole 20Meg, but I am happy with a 16MHz version at this
> stage (which we belive is the widest we can go). I am using

> python usrp_rx_cfile.py -f 2404.5M -s --no-hb -8 -N 40000 data1.dat

I'm pretty sure that the 8-bit support is still missing from the FPGA.
Until somebody fixes that, the highest rate you'll be able to run at
is 8MS/s complex.  That's what you'll get with -d 8.

> 1. Am I correct in using 2404.5M to downconvert the IF signal at
> 10MHz? I need "real" IF samples at 10MHz, do I need to change
> something here?

You set the frequency that you want to appear at DC in the complex
baseband.

> 2. this frequency in the above command is LO frequency or incoming
> signal frequency?

The -f <freq> argument is the RF frequency that you want to appear at
zero in the complex baseband.

> 3. My signal is in the range 2402.5-2422.5MHz range, which is well
> with in defined range of RFX2400 (2.4-2.483GHz), do I still need to
> cut lines for ISM filter on-board, as one of my project mates was
> advised to do. If yes, can't I use the other signal input on the
> board, which hasn't got any filter around it?

> 4. when I look at the spectrum of the IF samples collected using
> above command, they appear to have lesser bandwidth (4MHz) than the
> expected bandwidth(20MHz).

With -d 8 you'll get 64M/8 = 8MS/s complex.  That'll be +/- 4MHz.

> Sorry if any/all of these questions sound stupid, but consider me
> novice :S

Welcome!

Eric