On 14/07/2023 10:58, [email protected] wrote:
I am running an experiment aimed at assessing the accuracy with which a computer
can timestamp a file recorded from an X310, i.e. using the stat information on
the file saved when recording a known number of samples (GNU Radio head block)
at a
known sampling rate. For this I connect an X310 fitted with a Basic RX to a PTP
or,
here, an NTP synchronized computer, and feed the 1-PPS output of a White Rabbit
switch (PTP server) as input to the Basic RX. This same White Rabbit Switch
1-PPS
output feeds the Ext 1 PPS in of the X310 and its 10 MHz output feeds the Ext
Clock of
the X310. Attached is the result: on the bottom chart is the time at which the
first
rising edge of the record is detected. Each record is 3-s long (so two
consecutive PPS edges
at least) and the measurement is restarted every 16 seconds (X-axis of the
chart). The
Y-axis is graduated with the rising edge index divided by the sample rate of
5e6 samples/s.
While the duration from first to second PPS in each file is exactly 5e6+/-1 (oe
1s+/-200 ns),
the first PPS is detected with a 50 ms fluctuation from one measurement to
another (bottom chart).
What could this fluctuation be due to? I would have expected the X310 to wait
for the next
PPS edge and start recording from there. I could understand some lag between
the PPS detection
and first sample, but I would expect this lag to be constant and the
fluctuation to be 200 ns
(one sample) at most.
FYI the cm4_x310.py script used for these recordings, executed on a Compute
Module 4, for
configuring and collecting the measurements from the X310, with Time Source set
to External.
Thanks, JM
Looking at the script, it looks like you've set the *time* source to
"external", but are still relying on the internal
clock for the actual reference clock that is used to drive everything
internally--including the ADCs, FPGA
DSP bits, etc.
The "1PPS" on USRPs is used ONLY as a 1-time synchronization point for
resetting the timestamp clocks to a known value
*at a known time* (the 1PPS pulse). That timestamp register "ticks
over" based on either the internal reference clock,
or an external one, but you haven't requested external reference clock.
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]
