Any way to sample/log the 50Mhz ISM signal in gnuradio with USRP2?

Hi guy,

Thanks in advance.

Currently I am implementing a project based on ofdm_bentchmark example
with USRP2. I remember that the sample rate of ADC of USRP 2 with RF2450
can be up to 100MS/s. That means we can use USRP2 to capture at most
50Mhz
bandwidth signal.

Initially, I plan to log the signal on my computer (save as dat.
format)

and analyze it offline. Everything is fine when the sample
rate(bandwidth)
is set to 20MS/s in terminal. However, when we set the sample rate up to
50MS/s, nothing can be log in my computer. I guess there some software
limitations in Gnuradio.

My question is how to remove such limitation to achieve the 50MS/s
sample
rate. Any help will be greatly appreciated.

Best,
Kay

On 07/11/2014 10:20 PM, Yingjie C. wrote:

format)/ and analyze it offline. Everything is fine when the sample


Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page
To achieve 50Msps, you’ll need to use 8-bit wire format, and set the
peak= parameter appropriately. There’s just not enough bits/second on a
1GiGe link to support 50Msps with 16-bit, full-width samples.

But you also need to consider the format you’re writing samples to disk
in, and how much disk-write bandwidth you’ll require to support that.
If you just use the same format (2 x 8bits) into the file, that’s
100Mbyte/second into the disk. That may or may not be supportable on
your
particular system.

Just a side note: the 100MHz master sample clock rate of the USRP2
platform is complex sampling, so internally the USRP2 processes 100MHz
of bandwidth (not 50!), but that’s far too much to get it out via
gigabit ethernet, and so it decimates these 100MHz down to the sampling
rate you request as a user.

On 07/11/2014 11:04 PM, Yingjie C. wrote:

Thanks for your reply, Marcus.

8bits is enough for my case. But can you tell how to modify the adc
from 16bits to 8bits? I have no idea where to start it on gnuradio
example. /What to mean by setting the peak/?? what is the notation
of peak?

The UHD source in GRC has a parameter for setting the wire format.

http://files.ettus.com/uhd_docs/doxygen/html/structuhd_1_1stream__args__t.html

That’s for the raw UHD C++ interface, but see the notes about the
“peak=” stream argument, which you can specify in the UHD source
block in GRC.