from:	 Young, Ken 
to:	 Mark Gurwell ,
         Rurik Primiani ,
         Jonathan Weintroub ,
         Chunhua Qi ,
         Alice Argon ,
         Jun-Hui Zhao 
date:	 Thu, Nov 21, 2013 at 3:33 PM
subject:	 Real SWARM data available
mailed-by:	 cfa.harvard.edu

Dear SWARM Data Enthusiast,

   We got some genuine SWARM data last night, during this week's 
Testing Night. The data resides in two new-format data files:

engineering/mir_data/131121_04:59:22

and

engineering/mir_data/131121_10:51:27

Although both sidebands of both SWARM chunks were recorded, 
both sidebands
off both SWARM chunks actually contain exactly the same data: 
The USB of the 8-9 GHz portion of the IF.  The upper half of the 
Legacy Correlator (LC) was displaced higher than normal by 1 GHz, 
so instead of covering the usual 6-8 GHz of the IF, it covered 7-9 GHz, 
and thus overlapped with SWARM.   The hydrogen recombination line 
at 231 GHz was centered in USB chunk s42 of the LC, and that also 
put it near the center of SWARM.   The hydrogen recombination line 
is very clearly detected by both correlators, even in single 
integrations. The weather was poor, and was forecast to be very 
very poor, so we chose a short scan length of 10.3 seconds. Both 
correlators operated with exactly the same integration time, but 
the starting times of the SWARM and LC integrations were not 
synchronized. Normally dataCatcher will refuse to store data for 
which all crates did not start the scan at the same time, but that 
test was disabled for SWARM data. Since the phase stability was poor, 
you'll see significantly different phase noise between the two 
correlators in individual scans, because of the time offset, 
but of course that should disappear when the data is summed over 
longer time intervals.   There should not be any gaps in time were 
one or the other correlator was not integrating - there will just 
be an offset of a few seconds in one data stream relative to the other.

   At some point during the test SWARM stopped doing fringe rotation, 
and the data will be bad after that.   You'll see that fringe rotation 
was regained near the end of the track (on Mars and Jupiter), but the 
delays could not be fixed up (there was no good-enough planet), so the 
data during the last hour or so is probably useless even though fringes 
are seen on some SWARM baselines.   We started getting good data at 
about 5:00 UT, and I think the data is definitely good until about 
12:25 UT, and might be good until about 17:20 UT. Because the phase 
stability was really poor,  it'll take fairly careful processing, 
I think, to determine exactly when SWARM lost fringes.

   So - have at the data!   Let's see how well SWARM is working.

Taco