[Top][All Lists]

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

Re: [Discuss-gnuradio] B210 RX bandwidth setting from gnuradio-companion

From: Marcus D. Leech
Subject: Re: [Discuss-gnuradio] B210 RX bandwidth setting from gnuradio-companion ?
Date: Thu, 18 Apr 2019 15:36:47 -0400
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.7.0

On 04/18/2019 02:14 PM, jmfriedt wrote:
I am characterizing a weak (-80..-60 dBm) signal next (5 MHz away) to a
strong (-30..-40 dBm) carrier interfering with my measurement. I am
looking at power and phase (these are CW signals), and hit the issue
that the interfering carrier affects the sideband power measurement.

To demonstrate the effect, I mix a 295 MHz +7 dBm LO input (-45 dBm
output from the mixer isolation) with a -64 dBm RF intput at 5 MHz,
creating a strong carrier at 295 MHz and a weak sideband at 300 MHz
with -70 dBm power (due to mixer conversion efficiency of -6 dB).

As mentioned at https://files.ettus.com/manual/page_usrp_b200.html

The analog frontend has a seamlessly adjustable bandwidth of 200 kHz to
56 MHz.

Generally, when requesting any possible master clock rate, UHD will
automatically configure the analog filters to avoid any aliasing (RX)
or out-of-band emissions whilst letting through the cleanest possible

If you, however, happen to have a very strong interferer within half
the master clock rate of your RX LO frequency, you might want to reduce
this analog bandwidth. You can do so by calling

so obviously I'd like to reduce RX bandwidth to reject the interfering
carrier. The question is: how do I do that from the UHD USRP block in
gnuradio-companion. Since the RX bandwidth in the RF settings seem to
have no impact on the B210 RX characteristics (I tried a RX bandwidth
of 200 kHz with no effect on the impact of the carrier 5-MHz away), I
did follow https://files.ettus.com/manual/page_configuration.html
and set master_clock_rate=5.12 MHz (my sampling rate is 512 kS/s) which
indeed allows me to reject the interfering carrier (further away than
5.12/2=2.56 MHz) and get a clean phase and power measurement on the
sideband whatever the carrier power.

1/ what is the impact of master_clock_rate, i.e. what parameter am I
modifying by doing this ? The experiment is a physics experiment (ie
digital lock-in amplifier) and not digital communication, so I do not
care about bandwidth but I'd like to understand the impact of this
It's the rate that data are clocked between the AD9361 RFFE chip, and
  the FPGA.

I'm not sure that the chip can actually be clocked at that low a rate, and I'd choose a larger decimation factor, for example, x40, to give a master-clock rate


2/ I am not too happy about changing a clock where I try to set a
filter. What could be the proper argument to the UHD USRP block to set
the receiver bandwidth to a sub-5 MHz value ?

Just set the RX bandwidth in the UHD source block to a sub-5MHz value. If you're only looking at 512ksps, there may be no discernible difference between what UHD chooses by default, and when you set it manually.

You might also want to experiment with different RF Gain settings, to prevent mixing products from forming early in the RF chain, before the
  IF filter has a chance to do its thing.

reply via email to

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