Re: [Discuss-gnuradio] gr-usrp library

[Eric Blossom]

On Sun, Apr 09, 2006 at 01:55:33PM -0700, Ges wrote:
> Hi Eric,
> 

> I tried using the gnuradio blocks in gmsk (the old python script :
> gmsk-test.py) and sending a file across from one usrp (using the
> basic daughter cards) to the other using SMA cable - the same
> connection setup I had mentioned earlier.
> 
> I am simply modulating a char stream using the blocks in
> gmsk-test.py and sending it out using usrp_sink and using
> usrp_source to receive and then demodualting using the blocks in
> gmsk-test.py and dumping to a file.

Have you tried plotting the raw received signal using usrp_oscope?
Does the signal look OK?

> I am not sure what is happening, but the general_work() of the
> simple_correlator() simply returns 'zero'. And hence I have no
> output in the file_sink(). I have set a tx freq of 29.32e6 and a rx
> freq of the same value. I had tried with a gain of 0 on both sides n
> then with a gain of 20. But the results seem to be the same....

The transmit and receive gain setting have different allowable ranges.
Try setting them both to their maximum value.  You can query any usrp
daughterboard for its allowable range of gains.   Many of the examples
to this.  grep for gain_range.


>From db_base.py, the abstract base class for all daughterboards:

    def gain_range(self):
        """
        Return range of gain that can be set by this d'board.

        @returns (min_gain, max_gain, step_size)
        Where gains are expressed in decibels (your mileage may vary)
        """
        raise NotImplementedError

    def set_gain(self, gain):
        """
        Set the gain.

        @param gain:  gain in decibels
        @returns True/False
        """
        raise NotImplementedError



> Is it because of the simple_correlator() or is somehting else
> possibly going wrong? Please help....

simple_correlator works reasonably well for blocks of data that are
up to a couple of hundred bytes long.  This assumes that you're feeding
it 8x oversampled data, etc.

> I will be trying out the new blocks in the gmsk2 implementation. I
> did not want to do a CVS update due to some other issues. I guess I
> will grab those blocks and recompile gnuradio-core.

Good idea.  They are known to work with the Flex 400 cards.

> Thanks a lot

You're welcome.

Eric