Debugging the xband hub amps (20db) showed
that they were causing a problem. on 23jun24 they were brought
down to the lab for testing. The xband chain was left
connected with a bullet rather than the amps. This reduced the
gain in the chain by 20db and increased Tsys by about 40 K (more info).
Cal on,offs were then repeated on and off
the sun, as well as a few driven scans across the sun. This
was to see whether the system was still saturating while on
the sun. This would give us an idea of how large an attenuator
we'd need when looking at the sun.
The cal on off setup was:
- A jpl horizons ephemeris was used to track on and then off
- 30 second cal on followed by 30 second cal off was
performed on and then off the sun.
- the time was increased to 30 secs so we could average
over the total power variations while on the sun.
- The sequence was; off sun calon/off, on sun calonoff, off
sun calonoff, on sun calonoff.
- The cal diode is probably around 45K (need to measure in
- When off the sun.
- 2 deg (great circle ) offset in azimuth was used to get
off the sun.
- xband receiver using 8050 to 9200 MHz
- Mock spectrometer used to record the data.
- 7 172 MHz bands centered at:
- Each band spaced by 164 MHz
- 2 pols, 4096 channels/pol, 42 KHz channel width
- spectra were dumped once a second
- the power levels were adjusted off source so 1 sigma
voltage rms was 20 samples (rather than the normal 30).
- This was to try and give us a little more dynamic
range on,off sun.
- The total power was computed for each 1 second sample of
each 172 MHz band. no rfi excision was attempted.
- For each frequency band, the total power was normalized to
the average of the cal Off data when off the sun.
Plotting the calOn,off results:
The plots show the results of the calon,offs
- Page 1: calOn,offs all freq bands
- The x axis is the total power samples (1 per second). 30
calon, 30 caloff,, repeated for the 2 sets
- the y axis units are Tsys (caloff, off sun avg).
- 1st,2nd frames: polA
- 1st : on Sun: calOn,off on sun
- On the sun the signal increased by 13 to 22
TsysOffSun (depending on freq band)
- When the hub amps were in the cal deflection on the
sun was negative (we were well into compression).
- 2nd: Off sun, calonoff
- the cal off is .18 to .34 of Tsys (depending of freq
- removing the hub amps should have increased Tsys
from 50 to 90 K
- The cal deflection off sun with the amps in
was about .4 to .6 Tsys, so these numbers are
- 3rd,4th frames: PolB
- 3rd: on sun
- the levels are similar to those measured with polA
- 4th: off the sun. we see a cal deflection similar to
- Page 2: same plots blown up showing a single freq band:
- top: cal deflection on sun polA,B
- there is lots of total power variation from the sun.
- Black: polA cal deflection
- red: polB cal deflection
- green line, average calon value
- blue line: average caloff value.
- Using the average values:
- polA Caldeflection .20 [Tsys off sun]
- polB calDeflection .15 [Tsys off sun]
- Bottom: cal deflection off sun polA,B
- black: polA
- red: polB
- both deflections are now about .26 TsysOfSun
- With the hub amps in the calDeflections off the sun
were .5,.3 TsysOfSun
- So it looks like the system temps for polA ,B are
now a lot closer (assuming the cal is the same in
- The ratio calDeflection on sun to cal deflection
- polA .20OnSun/.27OffSun = .74
- polB .15OnSun/.25OffSun = .6
- So the system looks to have compressed going on the
- This compression is either in the fiber optics or
maybe the mocks.
- This is a lot less compression that what we saw with
the 20db hub amps installed (but remember the Tsys
also probably doubled when the amps were removed.
- So a solar attenuator will need to be more that 23
db (may 30db?).
Scanning across the sun
Processing the data:
The azimuth offset was computed using the
azimuth position in the header and the azimuth value coming
from the ephemeris file (that tracked the center of the sun);.
Plotting the strips
The plots show 3 scans across the sun (.ps) (.pdf)
- Page 1: scans using the 8624MHz band
- black is polA, red is polB
- Sampling rate: 10Hz.
- The scanning was started prior to the datataking.
- The first strip went -1.5 to +5 deg across the sun
(great circle) and got well off the sun..
- The scaling for all 3 strips used the average value of
offset 3 to 5 degrees of the first strip.
- Top: scanning at .1 deg/sec +/- 2deg (great circle)
- the sun is about 30 amin across, the beam is 10amin,
so we had about 3 beams across the sun.
- scanning at .1 deg/sec= 6amin/sec
- sampling at 10 Hz gave .6 amin/sample.
- with a beam of 10 amin this gave 16 samples across
- there were about 50 samples across the sun.
- Middle: scanning at .05 deg/sec +/- 1deg
- bottom: scanning at .05 deg/sec +/- .05 deg
- Page 2: over plot all 7 freq bands for the +/= 1 deg .05
- top: polA
- middle : polB
- bottom: tpPolA/tpPolB
- each strip was first normalized to the off sun value
- Looks like the ratio is always less than one.
- if Tsys polB was < tsys polA you would see this
- there may also be a bit of beam squint so they may not
be pointing at exactly the same spot.
- With the hub amplifiers removed, the cal deflection
(normalized to Tsys) is now withing 12% for polA and
- We thing the cal values to polA,B are close so this
means that TsysA is close to TsysB
- When the hub Amps were in, the deflections were .5 and
.3 Tsys. So the hub amps were affecting Tsys.
- With the hub amps out the gain went down by 20db and Tsys
went up from 50 to 90 (computed, not measured).
- The cal deflection on the sun is now .74(a),.6(b) times
that off the sun.
- So the system is still compressed, but no where near the
compression with the amps in (where cal Off was greater
- a solar attenuator for observing the sun needs to be at
least 23 db (probably closer to 30db).