As far as I know the maximum sampling rate for USRP N210 is 25M sps. To
receive signal with high frequency, such as 300 MHz to 2.2 GH, the RF
front
end converts the RF into IF, so the ADC can sample the IF signal. My
question is what the IF value is for daughter-board WBX? Where can I
reset the immediate frequency of the daughter-board, such as WBX and
TVRX2?
It may be a simple question, but I cannot find the answer. Any
explanation
will be appreciated.
WBX is a direct-conversion receiver, meaning there is no IF: it is a
quadrature receiver where the LO frequency is equal to the RF frequency.
TVRX2 is a low-IF receiver using a 12.5MHz IF.
Both daughterboards can be treated as zero-IF direct conversion
receivers
for tuning purposes, as the digital downconverter in the USRP’s FPGA
will
downconvert any residual IF to baseband before returning samples to the
host. In other words, when you tune the USRP to 2.2GHz, the resulting
samples will have a center frequency of 2.2GHz.
But I am not sure whether I understand what you said. My understanding
is
that the WBX converts RF signal into baseband signal directly, while for
TVRX2, the RF signal is converted into low IF signal firstly by TVRX2
and
then converted into basebad signal by USRP’s FPGA. Am I right?
Then, what is the sampling rate should I set to sample the signal of
interest. For example, FM radio signal at 106.3MHz.
I think I can clarify things a bit. The WBX and SBX default to what is
called direct conversion or homodyne mixing (see Direct-conversion receiver - Wikipedia). I’m not sure
about this, but also if you offset the LO as per Nick’s link, you
effectively make those boards do superheterodyne mixing or IF
conversion.
The TVRX2 is set to a fixed IF or superheterodyne mixing and you
cannot adjust the LO offset (need confirmation?). It also has AGC
always on.
At the end these variations don’t matter much because when the signals
pass through the DDC stage in the FPGA, they all end up being moved to
baseband and so look the same.
Broadcast FM has a channel bandwidth of 150 kHz, so a sampling rate of
200 or 256 ksps should work fine. Don’t confuse the sampling rate of the
zero-IF data with the sampling rate of the front-end, which is 64 msps
(USRP1) or 100 msps (USRP2).
The UHD API tends to hide this. In the old classic interface, you
directly set the decimation factor, so you had to know the hardware
sample rate in order to know what sample rate you would end up with.