Hi,
I am trying to use the USRP and gnuradio as a data acquisition system
for
feedback control of my physics experiments on atomic force microscopy
(AFM).
I am using the LFRX and LFTX daughterboards for my system. The bandwidth
that I am interested is DC ~ 1-2 MHz. I’ll be getting in 4 analog
inputs
which correspond to different physical parameters in the experiment and
based on the measurements I send 4 analog output signals for feedback
control. I have a few questions and it would be great if somebody can
answer
them.
-
After reading the archives I did these calculations on what I can
expect
in terms of the maximum signal frequencies that I can record from the
system. The bottleneck is the USB, hence Max data rate of USB = 32 MB/s
=
16 MS/s = 2MS/s per I and Q channel = 2 MS/s per input channel assuming
that
I set all Qs to zero and route ADC0 to I0, ADC1 to I1 etc. This means
that
the maximum signal frequency that I can get is 1 MHz. Is this
calculation
correct? By setting all Qs to zero I am wasting half the bandwidth. Is
there
a way by which I can increase the max signal bandwidth? For example
instead
of setting Q to zero if I could alternate samples between I and Q
channels,
I can make use of the full USB bandwidth. -
What is the number of samples that is stored on the on board FIFO?
After
reading the archives it seems that the number is 32k but I am not sure
of
this. I would be great is somebody can clarify this. -
Does USRP use DMA for transferring data or is the data written to
disk
through the USB every time?
Thanks!
Karthik
http://www.stanford.edu/~karthikv