[USRP-users] N210 WBX annoying spurs while switching
Ralph A. Schmid, dk5ras
ralph at schmid.xxx
Thu Oct 23 02:56:39 EDT 2014
The signal theory behind is that a perfect steep rise (100% vertical on the scope screen) produces a signal peak of infinite bandwidth. Google for Dirac pulse, that makes it clear. Thus it is essential to do some ramping, to avoid interference to the neighbor channels.
A real world example where exactly this (almost) infinite bandwidth is the dedicated goal would be UWB, wireless USB and related stuff.
From: USRP-users [mailto:usrp-users-bounces at lists.ettus.com] On Behalf Of bob wole via USRP-users
Sent: Thursday, October 23, 2014 8:07 AM
To: Matt Ettus
Cc: usrp-users at lists.ettus.com
Subject: Re: [USRP-users] N210 WBX annoying spurs while switching
Okay, I'll make a custom block in gnruadio to achieve ramp up/down for FSK waveform to avoid "keyclicks" .
On Thu, Oct 23, 2014 at 10:52 AM, Matt Ettus <matt at ettus.com <mailto:matt at ettus.com> > wrote:
No problem. The RF filter at the output really isn't the proper way to solve this problem, it will just keep the noise within a band. What you really want is to avoid putting out energy outside of your channel, and that can only be done with ramping.
On Wed, Oct 22, 2014 at 10:46 PM, bob wole <bnwole at gmail.com> wrote:
Thanks Matt for your response, that is the exact answer I was looking for. Actually I am using USRPs for TDMA and I do not want my time slot to increase very much. I wanted to know the ramping duration because this will affect the time slot duration. Second option could be that I can add an RF filter at the output of USRPs.
On Thu, Oct 23, 2014 at 10:25 AM, Matt Ettus <matt at ettus.com <mailto:matt at ettus.com> > wrote:
Yes, but it is even better to ramp up at the beginning and down at the end, and then pad a little bit with zeros. If you ramp up and down smoothly for 1us, you'll confine the emissions to about 1 MHz on either side. All bursty communication systems will do this, like Bluetooth, Wifi, even morse code, and most newer standards even specify the ramping rate.
On Wed, Oct 22, 2014 at 10:08 PM, bob wole <bnwole at gmail.com <mailto:bnwole at gmail.com> > wrote:
So you are suggesting that I should zero pad samples at start and end of my burst to avoid this ?
On Wed, Oct 22, 2014 at 9:16 PM, Matt Ettus <matt at ettus.com <mailto:matt at ettus.com> > wrote:
When you start and stop transmission without smoothly ramping the signal to zero at the beginning and end, you will create wideband spurs. Just think about it as amplitude modulation. Even if you are transmitting a pure sine wave, if you turn it on and off quickly, it will create out of band spurs. Hams call this keyclicks.
On Wed, Oct 22, 2014 at 1:25 AM, bob wole via USRP-users <usrp-users at lists.ettus.com <mailto:usrp-users at lists.ettus.com> > wrote:
I noticed that there are a lot of significant spurs generated in the output of WBX, on different frequencies randomly, when I use tx_time, tx_sob_tx_eob tags. I do not see these spurs if I transmit continuously. These spurs appears only when I start and stop transmission using stream tags.
How can I get rid of these spurs and what is the cause?
USRP-users mailing list
USRP-users at lists.ettus.com <mailto:USRP-users at lists.ettus.com>
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the USRP-users