Back to Astronomical Checkout summary.
Report on the Spectrometer Tests of September 10/11, 2002
Configuration: 4N2-6A-50-3; Rcvr 1-2 GHz
D. Hogg and F. Ghigo
Summary
The GBT spectrometer was used in five configurations during the
night of September 10/11, 2002. This report summarizes what was
observed in the fourth configuration.
The results are promising,and some of the scans give data that
is in fair agreement with expected values. The properties of
the test HI profile are in good agreement with the values from
the Fisher survey, but the observed rms is 18% smaller than the
expected value. Moreover, two scans show an unexpected problem in
channel one, and it is recommended that this correlator configuration
not be released for general use until the problems with the rms
and with channel one have been solved.
Analysis
The data are in the GBT data base as /home/gbtdata/TSPECTEST_05.
The data were filled into an area on EUCLID:
/export/home/euclid/scratch/dhogg/ and are in file:
TSPECTEST_05_SPECTROMETER_A
Configuration 4. 4N2-6A-50-3
L Band, at 1429 MHz; 50 MHz bandwidth; 3-level sampling.
Observations were made of the hydrogen line in the galaxy
UA93o
chosen from the Fisher
GBT HI Galaxy Survey,
scans #73 to #81.
Scan #73 is spoiled, so that the first good OffOn pair is
#74/#75.
The spectra in scans #75 and # 79 are reasonable, though of
course noisy. For 3-level sampling, a channel width of 381.5 Hz,
and a system temperature of 19 K we expect an rms on a single
channel of an OFF/ON pair of total duration 10 minutes to be
76.4 mK. Figure 1 (ttday5#75pla.ps) shows the spectrum of each
channel, averaged for the two scans, with intensity as a function
of pixel number. Figure 2 (ttday5#75plb.ps) shows the spectrum with
intensity as a function of frequency.
The observed results are tabulated below.
Effective RMS, in mK Ratio
Int. Time Expected RX1 RX2 Both O/E
150 76.4 63.3 62.8 0.83
300 54.0 44.7 44.5 44.5 0.83
600 38.2 31.5 0.82
The receivers are of comparable sensitivity. The observed rms is
about 18% less than the expected value. The difference is significant
in view of the agreement found in other correlator configurations,
and the cause of the difference should be examined. The rms decreases
with increasing integration time exactly as expected, but the test is
with only a limited amount of data.
The final spectrum obtained by integrating all data is given in
Figure 3 (ttday5#75ava.ps), which shows intensity as a function of
barycentric radial velocity. The figure has been cropped to give
a display similar to that in the Fisher catalog.
Figure 4 shows the spectrom from the Fisher catalog.
The properties of the profile are compared in the following table.
We use the average of the two channels for each IF, and we convert
from antenna temperature to flux density using the conversion factor
found in early August by R. Maddalena, 1 Jy = 1.91 K.
Quantity Fisher This Test
Maximum in profile (Jy) 0.42
Peak Intensity (Jy) 0.38 0.41
Line Width @ 20% (km/s) 194.7 194.8
Systemic Velocity (km/s) 717.7 718.3
Flux Integral (Jy km/s) 57.30 58.17
The velocity parameters are in excellent agreement. The peak
intensity and the flux integral are only slightly higher in the
test than in the Fisher profile, by 8% and 1.5% respectively.
It should be noted that the peak intensity in the test is an
estimate, and may not be at the same point in the profile
as the Fisher value.
There is a serious problem in this group of spectra which will
have to be solved before this correlator configuration is released
for use. Scan# 77, and to a lesser extent scan #81, shows a dramatic
increase in the noise on CH1:XX that does not appear in scans
#75 and #79, nor is it seen in CH2:YY on scan # 77. The spectrum
of scan #77 is given in figure 5 (ttday5#77pla.ps).
A quantitative evaluation of the effect is given in the following table.
Frequency Range RMS (mk)
MHz RX1 RX2
1395-1399 466 69.5
1405-1409 325 62.1
1425-1429 213 59.3
1435-1439 138 56.2
Back to Astronomical Checkout summary.