Project: TBASErdn021211
Roger Norrod

Continue spectrometer tests from previous day (TBASErdn021210). Injected Noise
Source II into SF 3,4,7,8. [Note: Found in midst of tests that I failed to
connect the Cal TTL control, so cal was not firing in scans 1 through 54.
Later scans it was connected.]  Spectrometer configured as CalOn/CalOff, 1
second switch rate, blanking 0.002 seconds.

Note:
SF1 --> HSS0
SF2 --> HSS2
SF3 --> HSS4
SF4 --> HSS6

SF5 --> HSS1
SF6 --> HSS3
SF7 --> HSS5
SF8 --> HSS7
*********
Configured Spectrometer to A1 (Bank A, Quadrant 1), 8192 lags, 3-level,
200 MHz mode: 1W2-045-200.  SF3/HSS4 & SF7/HSS5 used. 20 second integrations,
120 second scan time.

Scan 1 st@ 13:44:57 -> Scan 10 fi @ 14:06:02.

Discussion: Figure 1 shows the spectra from scan 1. Red is from HSS4; green
from HSS5. Figure 2 plots (Sig - Ref)/Ref for HSS4 with alternate scans taken as
Ref and Sig. The standard deviation is consistant at 10.5X10^-4, and DC
offsets fluctuate a few parts in 10^4.  Figure 3 is a similar plot for HSS5.
Figures 4 (HSS4) and 5 (HSS5) show the ratioed spectra when scans 1-2 and 
scans 1-10 are paired. Note that scans 1 and 10 were taken 20 minutes apart.

*********
Reconfigured spectrometer to mode 1W2-067-200. SF4/HSS6 and SF8-HSS7 used.

Scan 11 st @ 14:07:00 -> Scan 20 fi @ 14:28:14.

Discussion: Figure 6 shows the spectra from scan 11. Figures 7 and 8 correspond
to figures 2 and 3 in the previous section. Figures 9 and 10 correspond to 
Figures 4 and 5. The results for this mode are very similar to the previous
mode. Only thing of note is that the SF4/HSS6 combination shows somewhat higher
slope in the spectra than the other three checked today.


*********
Reconfigured spectrometer to a two bank, two quadrant mode, four sampler mode.
Bank A, Quadrant 1 - Mode 1W2-045-200 - 8192 lags
Bank B, Quadrant 2 - Mode 1W2-167-200 - 8192 lags

Scan 21 st @ 14:34:21 -> scan 30 fi @ 14:55:45

Discussion:
Figure 11 shows the spectra from scan 21, bank A. Compare to Figure 1. Figure 12
shows the spectra from scan 21, bank B. Compare to Figure 6. Analysis plots
of ratioed spectra were viewed, but not saved as the results were equivalent to
those given above.


*********
Moved NS II to SF 1, 2, 5, and 6 inputs. Reconfigured spectrometer to record
corresponding samplers.
Bank A, Quadrant 1 - Mode 1W2-001-200
Bank B, Quadrant 2 - Mode 1W2-123-200

Scan 31 st @ 15:00:27 -> scan 40 fi @ 15:21:52

Discussion:
Figure 13 shows the two spectra collected from Bank A (HSS0, HSS1), and may
be compared to Figure 7 in the TBASErdn021210 report.  Figure 14 shows the
spectra from Bank B (HSS2, HSS3), and may be compared to Figure 10 in the
TBASErdn021210 report. The four sampler outputs were analyzed, but again the
results were so unremarkable the plots were not saved. There is essentially
no difference in ratioed spectra standard deviation for adjacent scans and for
scans taken 20 minutes apart.


********
In the same configuration, did two 10 minute scans with 30 second integrations.

Scan 41 st @ 15:24:40 -> scan 42 fi @ 15:44:54

Discussion: The four samplers' ratioed spectra are flat with standard
deviation of 4.6+/-0.1X10^-4. Figure 15 shows an example, for HSS3.


********
Changed to try various 800MHz modes. Added 6dB to NSII output attenuator to 
get near optimum sampler levels. Stayed in two bank, two quadrant, four
sampler configuration:

Bank A, Quadrant 1 - Mode 1W2-001-800 - 2048 lags
Bank B, Quadrant 2 - Mode 1W2-123-800 - 2048 lags
20 second integration time, 120 second scan time.

Scans 43 and 44 accidently setup with mismatched integration times in two
banks. Ignored.

Scan 45 st @ 15:57:42 -> scan 54 fi @ 16:19:06

Discussion:
Figures 16 and 17 show the four spectra from scan 45. Note the spike at
340 MHz in all four. This also dominates the ratioed spectra, Figures 18-21.
Figures 22-25 show ratioed spectra for paired scans 45-46 and 45-54. The
SF5/HSS1 data shows noticable structure after 20 minutes. It seemed to
gradually detoriate as shown in Figure 26, scans 46-54 all ratioed to scan
45.

********
Realized that I forgot to hookup the cal control cable, so previous scans
had no cal firing. Connected it now, and decided to repeat above test.

Scan 55 st @ 16:25:23 -> scan 64 fi @ 16:46:48

Discussion:
Figure 27 shows the calOn and calOff phases of scan 55 for HSS0.
The only notable difference in the analysis of this sequence of scans compared to
the previous ten, is that all the 55-64 scan ratios show some structure similar
to that seen in Figure 26.

********
Reconfigured to other 4 samplers.
Bank A, Quadrant 1 - Mode 1W2-045-800 - 2048 lags
Bank B, Quadrant 2 - Mode 1W2-167-800 - 2048 lags

Ran for 40 minutes over lunch.

Scan 65 st @ 16:54:24 -> scan 84 fi @ 17:37:17

Discussion:
Figures 28 and 29 show the four spectra from scan 65. Figures 30 and 31 show
the ratios of alternate scans for HSS4 and HSS6; plots for HSS5 and HSS7 are
similar.  Note there are noticable DC offsets (the traces should be uniformally
spaced by 0.002 units). However, all four samplers showed the same pattern of
offsets, so it is probably related to the know periodic cycling of the 
Equipment Room temperature.

The ratioed spectra develop more structure as the Signal scan moves further
in time away from a Reference scan. Figures 32-35 show the results of using
scan 65 as a Reference, and Signal scans at multiples of 6 minutes later.
Note the feature at 525MHz that grows with time.