Source use of the time.sleep function causes USRP 'U'

[????????]

Hi,
 I'm using USRP to send and receive data, with two host 
PCs and two USRPs as senders and receivers respectively. I am using the ofdm 
example in GNURadio. The source of the file on the sending side is an 
out-of-tree module I wrote myself, which will output N sets of data, and I want 
to output the next set of data after a 10-second interval after each set of 
data, so I use the "time.sleep(10)" function, at this time the problem occurs 
as follows.
 1??Except for the first group of data which was 
received normally, all other groups of data were lost??Terminal print 
"U"??underrun??PC not providing data quickly??
 2??When I remove the "time.sleep" statement, all data 
is received normally
 Why does this happen? How can I fix it?

 Thanks in advance??
Best,
Regards.

Cross-Compiling UHD Cmake version error

[GNU Radio, the Free & Open-Source Toolkit for Software Radio]

Hi, i'm trying to Cross-Compiling UHD a USRP E312 following this guide

but
i get a cmake version error message, the problem is that i have the 3.22.1
version.

Please can someone help me with this problem?

Re: Maximum sample rate for two channels

[isaac mario tupac davila]

Thanks Marcus

https://files.ettus.com/manual/page_converters.html#:~:text=The%20format%20used%20by%20the,%2Dthe%2Dwire%20format
).
https://files.ettus.com/manual/structuhd_1_1stream__args__t.html

I hope the increase in the quantization noise doesn't affect my graphics so
much.

Regards
Isaac T.

El lun, 3 oct 2022 a las 12:59, Marcus D. Leech ()
escribió:

> On 2022-10-03 12:10, isaac mario tupac davila wrote:
> > Hello community
> >
> > I'm Isaac. I'm trying to receive a signal with two channels using an
> > USRP N200 and a Basic RX.
> > I'm setting the sample rate of each channel to 20MSp/s, but what I
> > obtain is: The total sum of rates (40.00 MSps on 2 channels)
> > exceeds the maximum capacity of the connection, causing overflow (O)
> > in the acquisition.
> >
> > So, How do I know which is the maximum capacity for two channels? Why
> > is it not possible to acquire two channels with a sample rate of
> > 20MSp/s in each one if I can acquire one channel with 20MSp/s?
> >
> > Thanks in advance
> >
> > Isaac T.
> The maximum possible bit-rate for a 1GiGe connection is 1Gbit/sec.
>
> The absolute most you can push over such a connection with the N200 is
> 25Msample/second--with 16-bit I and Q.
>
> If you move to 8-bit samples, using "otw_format=sc8" in your device
> arguments, you can move up to 50Msps over
>that same connection.
>
> This is basically just the maximum delivered rate of a 1Gbit
> connection.  So, one channel at 25Msps or two channels at 12.5Msps,
>or if you're using sc8 "wire" format, 50Msps for one channel and
> 25Msps for two channels.
>
>
>
>

Re: Maximum sample rate for two channels

[Marcus D. Leech]

On 2022-10-05 17:13, isaac mario tupac davila wrote:

Thanks Marcus

https://files.ettus.com/manual/page_converters.html#:~:text=The%20format%20used%20by%20the,%2Dthe%2Dwire%20format).
https://files.ettus.com/manual/structuhd_1_1stream__args__t.html

I hope the increase in the quantization noise doesn't affect my 
graphics so much.


Regards
Isaac T.



Well sure, but a 1GiGe network interface carries what it carries.

If you want dual-channels at 20Msps or more while preserving 16-bit 
sample values, consider the X310, which supports 10GiGe

  interfaces.

The N200 series is now over 10 years old, and 1GiGe at the time was a 
"sensible" design choice.