On 09/16/2020 09:00 PM, Minutolo, Lorenzo wrote:
The GPU does the signal demodulation (specific for each detector). The
phase shift has the strongest response usually, a representation of
the resonator in phase magnitude is available here
<http://www.its.caltech.edu/%7Eminutolo/server_docs/images/resonator_model.html>.
In this
<http://www.its.caltech.edu/%7Eminutolo/server_docs/images/screenshot_calib_diag.html>
other plot you can see a VNA overlapped to a noise acquisition in
which we make the phonon flux oscillate.
Tone tracking comes into play when the resonator does not shift few
kHz from the original resonant frequency but ~2MHz: using a fixed tone
on resonance, say 300MHz, we will detect very small variation in the
S21 if the resonator moves from 301MHz to 302MHz.
Is the resonator drift due to photon flux, or some physical non-ideality
in the physical implementation of the resonators? They're supercooled,
so I wouldn't expect temperature effects--but I'm not a physicist. I
just play one on the Internet....
An FPGA firmware has the downside of not being programmable in simple
languages like python or cuda and would fail the original flexibility
of the project.
I’m curious anyway: suppose I want to implement a tone tracking
firmware: would I still have space on a N321 to keep the stock
firmware? In other words: can I add functionalities without discarding
all the UHD api already in place?
I'm not sure what the "headroom" is on the N321. It has generally been
the case that really-exotic "stuff" in the FPGA tends to cause one
to "throw stuff out", but there may be enough room on the N321 to
make that not a thing.
Please copy the list on this dialog, since there are others on the list
better situated to answer questions about FPGA "headroom", etc. I've done
so on this reply.
Lorenzo
_______________________________________________
USRP-users mailing list
[email protected]
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
