Hi,
I am testing a transceiver link in ISM band using two USRP2+RFX2400, one
configured as TX and the other one as RX. I would like to test my own
FPGA program, so for now I have your FPGA codes in which I only changed
the signal driving the DAC in TX to be an square wave (+8191 or -8192)
with frequency about 400Hz. At the RX side, I am looking (using an
oscilloscope) at the input to the ADC and what I see is quite like my
square wave modulated with a carrier of couple of MHz. I looked at the
schematic of RFX2400 and I didn’t find any PLL or clock recovery
circuitry! so I wonder how the frequency offset at RX could be avoided?
Regards,
Malihe
On Tue, 2011-01-04 at 19:25 -0700, Malihe A. wrote:
circuitry! so I wonder how the frequency offset at RX could be avoided?
Regards,
Malihe
In a software-defined radio, frequency estimation and clock recovery are
typically done in software, to keep the hardware general-purpose enough
for many different uses. You can either synchronize the two USRP2s by
using a common 10MHz reference input to both devices (and enabling the
reference lock), or deal with the inevitable frequency mismatch in
software.
–n
Hi,
I have included this line in my python flowgraph:
|||sink.config_mimo(usrp2.MC_WE_LOCK_TO_SMA)
where sink is: ||sink ||=| |usrp2.sink_32fc()|
but I think USRP2 is not locked to the external reference clock since
when I turn this ref clock generator off, there is no change in the
spectrum of the output of USRP2. My understanding is that if USRP2 is
locked to the ref clock, when I turn the clock off I should see no
signal at the output of USRP2. can somebody please help me with this?
How do I get the USRP2 to lock to the ref clock?
Thanks,
Malihe
Yeah, I mean ref clock. I actually connect a 10MHz sin wave with 15dBm
amplitude to two USRP2, one acting as TX and the other one acting as RX.
the reason I am using this ref clock is to avoid the frequency offset
between the TX and RX (I have been observing about 1MHz offset between
the two). But yet there is about 1MHz frequency offset between TX and RX
and that is the reason I thought their are not locked to the ref clock.
Why would I still observe about 1MHz frequency offset then?
On Wed, 2011-02-23 at 19:08 -0700, Malihe A. wrote:
Yeah, I mean ref clock. I actually connect a 10MHz sin wave with 15dBm
amplitude to two USRP2, one acting as TX and the other one acting as RX.
the reason I am using this ref clock is to avoid the frequency offset
between the TX and RX (I have been observing about 1MHz offset between
the two). But yet there is about 1MHz frequency offset between TX and RX
and that is the reason I thought their are not locked to the ref clock.
Why would I still observe about 1MHz frequency offset then?
Dumb question, but are you syncing the source USRP2 (using config_mimo)
as well? Incidentally, 1MHz of offset at 2.4GHz is >400ppm, which is a
lot more frequency offset than I’d expect from two USRP2s tuned to the
same frequency. Can you post your source and sink flowgraphs?
–n