Intro
Daily maps of the sun are made with the
12meter telescope at xband.
- The maps are 2deg x 2deg center on the sun
- Each strip drives in RA and then steps in dec by 1/2 a
beam(5amin)
- At the start of each map a 10 second calon, 10sec cal off
is done at -1deg RA(greatcircle), -1deg dec
- At the end of each map a 10 second calon, 10 sec cal off
is data at +1deg RA(great circle), +1deg dec.
- a map takes about 30 minutes.
- When converting the data from spectrometer counts to degK
the average of the 2 cal scale factors are used.
- The electronic gain of the telescope is know to vary with
temperature (more
info).
- A test was done to see how much the calDeflection changed
between the first calon,off and the last calOn,off.
- This would give an idea for the conversion error from
spectrometer counts to degK caused by electronic gain
variation.
- We could use this to decide if we need to take a cal
on,off on each strip rather than only at the start/end of
the map.
Processing:
- All sun maps taken 19APR22 thru 02aug22 where used for the
test.
- there were 211 complete maps (with both cals)
- The two calon,offs were input for each map
- The ratio of the calDeflection for the Lastcal/Firstcal
was computed
- this was done for polA, B in the 7 frequency bands.
- Changes in the ratio would occur if:
- the electronic gain of the system changed during the
30min of the map
- the temperature of the cal diode is stabilized
with a peltier cooler so its output power should not
vary.
- There was rfi that was not removed completely from the
total power computation.
- If rfi was a problem, it should not show up in all 7
freq bands (unless the system saturated)
Plotting the data:
The plots show the
variation of the caldeflection for the 211 sunmaps (.ps) (.pdf):
- Page 1 Ratio of CalDeflections vs date (all
freq bands over plotted)
- Top frame: polA
- bottom frame:polB
- each color is a different frequency band
- the pink band(8789)and dark red band(8931) have large
excursions.
- During these excursions the other bands are
relatively stable. t
- this says that the jumps are rfi related and not
from the electronic gain variation.
- + have been plotted at each measurement
- gaps in the data end up looking like straight lines.
- Vertical axis
(calDeflectionEndOfMap/calDeflectionStartOfMap)
- The vertical scale in clipped to .8 to 1.2 . some
points go offscale.
- horizontal axis: date at yymmdd
- Page 2-3: Ratio of calDeflection vs date with each freq in
a separate frame.
- Black is polA
- red is polB
- splitting it out makes it a little easier to see the
variations.
- Page 4: Histogram of the Ratio.
- the data bins are .01 in length. The max,min values were
set to .9 to 1.1 (+/- 10%
- each color is a different frequency band.
- top frame: polA
- bottom frame: polB
- the center of the polA and polb histograms are less than
1.
- this means that the cal deflection at the end is
smaller than the cal deflection at the start
- the maps are normally taken around 10am when the
temperature is still increasing
- As the temperature increases, the electronic gain
decreases.. so this shift is to be expected.
- The center of polB is a little less than the center
of polA
- the gain vs temperature plots (see more info above)
shows that the gain of polB changes faster than polA
with temperature.
- so we would expect polB to shift a little more.
- Page 5: ratio of cal deflections vs elevation
- top polA
- bottom polB
- there is not obvious elevation dependence
- you wouldn't expect there to be it the change is from
the electronic gain.
Summary
- 211 sun maps were examined covering 19apr22 thru 02aug22
- the calDeflection for the first and last cal was computed
and then the ratio CalDefltLast/calDeflFirst was formed.
- changes in the ratio of the cal deflections will show the
electronic gain variations during the 30 min sun map.
- looking at histogram of the calratio:
- the center of the histogram are offset below 1.
- maps typically made in the morning when the
temperature is increasing.
- the electronic gain decreased with temperature.
- PolA center is around .98
- polB center is around .97 (polB electronic gain changes
faster than polA with temp)
- the histogram widths are about .04 or a 4% variation
- There is an electronic gain variation of about 2-3%
- We use the average of the two cals when converting to
kelvins
- if the variation is linear then that reduces the error
by 1/2.
- If a 1% error in the cal caused by electronic gain
variation is acceptable, then you don't need cals every
strip.
- Note: this does not measure how accurate the cals
values are, only how much they vary.
processing:x101/mapsun/test/suncals_inp.pro,
plotratio.pro