It appears that the alltime channels read actual counts at night; however,
THE 21 SHADOWBAND CHANNELS HAVE THE NIGHTIME VALUES SET TO 1.

Lee cosine corrects the direct, not the diffuse, and adds the corrected 
direct to the original diffuse that has only been corrected using the sideband
measurements; he reports this as total.

L -41 39.99 105.26 for TSR548 on roof of DSRC

"O" command arg -->  {90 - [elv angle of motor axis (suggest 30 deg) - 39.99]}*65536/360

O 18203 for DSRC

I $20 $11FF 0 20 1

584.SOL.yes.org is YES original cosine file
584.SOL.yes is my modification of YES 584.SOL because of using different channels to
take data. Only header modified.
584.sol100819_unsmoothed.txt is modified by Charles Wilson to include YES thermopile values plus 
his filter values from unsmoothed cosine run in AUG 2010.

Commands to download data and clear memory:
M 1	# moves ram data to mem card
T 2	# dumps mem card
M $7FFF	# clears mem card (raw data taken during dump is transferred to mem card)

Subtract 0.6 mb from pressure obtained at http://gpsmet.noaa.gov/cgi-bin/gnuplots/rti.cgi.
Time is given at that site in UTC.

*** The file tsr584.120130.xmd had only one channel activated. ***

***************
Run rsrsplit -g 720 to get .mtm files.
Once you have .mtm files, 'ls' them to a file, e.g., 'ls E6DE.*.mtm > foo'; go in and create a 
vi macro using 'q', followed by some character for the macro name, say 's', then what you want
to do to the first line, for example 'tu -o 584.021106 B76A.021106.mtm', if the
first line contained the mtm file name 'B76A.021106.mtm'; then position yourself at the
beginning of the next line and type another 'q' to end the macro creation; to execute
the macro type '@s' or to do it 100 times '100@s'

After changing 'foo' to an executable, execute it to get files with names like '584.021106'

'la_stuff.s 584.yy*' will create a file with the la outputs concatenated and called la.out

This removes unsuccessful langleys from file
grep -v points la_out > fn1

This grabs the time line and the 5 lines of data and removes lines with only '--' in them
grep -B1 -A4 'Channel   1' fn1 | grep -v '^--' > fn2

Or strung together:
grep -v points la.out | grep -B1 -A4 'Channel   1' | grep -v '^--' > la.clean
Note that at times not all 5 filters return good data, and that messes this up, so check by 
editing to remove sets that do not have all five filters if needed.

RUN 'strip_data.s la.clean' after editing any remaining messed up data in la.clean
Then run jd -f  foo.temp > foo.day
In 'R' run get_vos.f

***************
Use 'beav' to edit binary (xmodem) files; the three bytes that are next to the last header
byte are the bytes remaining in the file starting just after the header. There is an email
from Jim Schlemmer from Dec 2011 that explains the header.

'beav' has bytes in hexadecimal in correct order (no switched bytes).


0D E6 DE 01 C7 4A D9 1C 6F 20 00 14 00 14 9F AE
vr  ID    ID    LON   LAT  FL Samp   Avg   JD

00 C7 10 07 00 00 00 00 00 00 11 FF 00 00 00 1A
Secs_UT  Ch | Day chn | | All tim | |cntrs | bytes

52 50 00
left  err

Apparently "00" is a placeholder in a file with ERROR CODE 01 to keep the time correct so
removing these removes 20 secs of time.

For dsrc get pressure from gps met site, but subtract 0.6 mb since it is measured at a lower level
of about 20 feet. gps met site time in UTC.