Received signal is DANCING

I am now building a flow graph for ASK modulation,during the test,

Here is my flow graph,
http://picasaweb.google.com/lh/photo/-4xnI_UgFeSNENH3UGHgxw?feat=directlink

And this is the result plot,
http://picasaweb.google.com/lh/photo/QYy5NJEezHCqaMX-8AIFww?feat=directlink

The bottom is the RX signal. As you can see, RX is not similar with TX,
even
worse, it is dancing!

On Thu, Nov 18, 2010 at 6:34 AM, Songsong G.
[email protected]wrote:


Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page

What does the resulting plot look like if you remove the bandpass
filter?
Those discontinuities when the orig signal abruptly turns on/off is
essentially high-frequency info, which when filtered out by whatever
means,
will result in a rx signal that is distorted from the original.

Also, why do you have a throttle? You have a USRP sink, let its Interp
factor be the rate-determiner.

-Steven

Also, why do you have a throttle? You have a USRP sink, let its Interp
factor be the rate-determiner.

-Steven
The whole flow-graph is suggestive of deep confusion.

The bandpass filter is being told to operate at a sample rate of 7K, but
in fact the samples are coming from a USRP with decimation=256,
which gives 250KHz sample rate. The filter bandwidth is quite
narrow, so if it was told the correct sample rate, it would likely
distort
that signal even more.

As previously observed, the throttle blocks are quite unnecessary.
Also, why the “repeat” block? The file source is already set to
“repeat”,
so I don’t understand what that repeat block is doing there.

As previously observed, the throttle blocks are quite unnecessary. Also,
why the “repeat” block? The file source is already set to “repeat”,
so I don’t understand what that repeat block is doing there.

the repeat acts like an interpolation, where the file source repeat
option just replays the files. -Josh