[Top][All Lists]

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

Re: [Discuss-gnuradio] signal distortion in simple code

From: cswiger
Subject: Re: [Discuss-gnuradio] signal distortion in simple code
Date: Thu, 30 Dec 2004 14:08:16 -0500 (EST)

On Thu, 30 Dec 2004, Krzysztof Kamieniecki wrote:

> Your real input seems to be centered at and wrapping around 2^15. Could this
> cause a problem since you are using shorts [-2^15,2^15) at the input?

The real ADC is 12 bits and creates data from 0-4095, centered on 2048.
(the displayed sig has been zero-centered, then converted back to
unsigned which put it at zero=32768 - sorry for the confusion)

But, yes I have tried all the dc offset stuff - sub'ing 2048 from the
real signal or offsetting the software signal doesn't change things.

Further research into a script I've been using turns up the bizarre
signal at the bottom of this page:


?? When those were run cpu was at about 50%.

Speaking of which, I've recently discovered that Pentium-M notebooks
don't always run at full speed, but have something called Intel
"SpeedStep" that throttles back cpu depending on idle time. To force
it to full speed always, I'm using (Fedora Core 3 on a Dell Inspiron)

# cat /proc/cpuinfo

to show current cpu Mhz, then

# kill -s SIGUSR1 `pgrep cpuspeed`

to force full speed (see /usr/sbin/cpuspeed --help), and to resume
normal self-adjusting

# kill -s SIGHUP `pgrep cpuspeed`

Anyway, manually kicking in high gear stopped the few lines of underruns
(UUUUUUUU) I was getting when starting gnuradio scripts.


reply via email to

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