[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] An interesting application for my new USRP... som

From: Larry Doolittle
Subject: Re: [Discuss-gnuradio] An interesting application for my new USRP... some input requested
Date: Tue, 24 May 2005 14:10:28 -0700
User-agent: Mutt/1.5.9i

John -

> [chop], there 
> will be three or four signals, on different frequency bands (near 3.58, 
> 7.08, 14.08, and 21.08 MHz).  The signal will be an unmodulated carrier 
> (well, there will be some CW identification) and will be transmitted for 
> about 10-15 minutes.
> It seems to me that with two of the Basic RX boards, I could monitor all 
> four frequencies simultaneously and save the data streams to disk

It sounds like a single acquire channel would be enough, the signals
are low enough to acquire without aliasing at 65 MS/s.

> for 
> processing with a deep FFT to extract the average frequency of each 
> signal over the run.

FFTs suck for this kind of precision work.
Instead, think of it as a curve-fitting excercise.
Test your process ahead of time with some simulated signals.

> Since I'll know the frequency ahead of time to 
> within a few kHz, I won't need to capture a wide swath -- probably 10kHz 
> on each RX input will do.

Right, but do the filtering digitally instead of analog.

> Since Matt tells me the USRP can be clocked by an external reference, I 
> should be able to extract very accurate frequency information

In principle, you could dedicate a second channel to recording the
GPS band, and extract near perfect frequency calibration from that.
Unfortunately for your time-frame, GPS reception in itself is a
serious project that has not been fully solved by the GNU Radio
community (i.e., Krzysztof).

> 2.  If I clock the USRP from a highly stable external reference, is 
> there any other source of possible frequency error in the system (since 
> we're at HF, I shouldn't need to use a downconverter)?  In other words, 
> will the measured FFT frequency have any error source other than errors 
> in the external reference?

Not in the long term.  Any acquisition hardware adds phase noise,
but that mostly averages away for the measurement you want to make.

> This sounds like fun.  Any ideas/suggestions would be greatly appreciated.

Keep us posted.

    - Larry

Attachment: signature.asc
Description: Digital signature

reply via email to

[Prev in Thread] Current Thread [Next in Thread]