Back to Astronomical Checkout summary.
Spectrometer Test of Configuration 1N4-0A-50-9 on October 4,2002
D. Hogg and F Ghigo
Summary
Tests of the spectrometer in the 1N4-0A-50-9 configuration at
L Band show two problems which should be addressed before this
configuration is released for general use. First, the bandpass
as shown by a spectrum taken of a moderately strong (6 K)
continuum source contains structure of amplitude up to 6%
over a frequency interval of 5-10 MHz. Second, the rms on
a trial spectrum of a galaxy integrates down more slowly than
expected. At an effective integration time of 2400 seconds,
that is, eight pairs of Off/On scans, each pair taking 10 'clock'
minutes, the rms is higher by factors of 1.31 and 1.22 in IF#1 and
IF#2, respectively. The parameters of the galaxy HI profile
measured in this configuration agree closely with those listed
in the Fisher catalog.
See comparison of 1N4 and 1N2 modes.
Introduction
Observations to test the correlator mode 1N4-0A-50-9 were made
on October 4,2002. The data are saved in the file TSPECTEST_09.
The observations in IF#2 were offset by 4 MHz from the
observations in IF#1. Calibration was made by a spectral line
observation of the continuum source NVSS0106+013. Eight pairs
of Off/On scans, comprised of 5 minutes Off and 5 minutes On,
were made of the galaxy
UGC668o which is included in the Fisher
GBT HI Galaxy Survey.
Calibration using NVSS0106+013 = VLA0108+015
The spectra for each IF and each channel show considerable
structure. The structure is similar at each of the IF's for
a given channel, even though the LO has been shifted by
4 MHz. This is seen in Figure 1 (tday9#20f.ps), which shows the
four spectra plotted in velocity coordinates. The two IF's
can be distinguished by looking at the ends; IF#1 is plotted
in red and green, and extends beyond -4600 km/s to -5400 km/s,
while IF#2 is plotted in dark and light blue, and extends
from 5150 km/s to 6000 km/s. It is clear that all of the structure
seen in the spectra has remained at a fixed frequency on the sky.
The structure differs somewhat between the two channels, with
channel two showing more variation. The maxima and minima
for IF#1 are shown below; the values for IF#2 are comparable.
Channel Velocity Frequency Pixel Ta*
One -4222 1440.412 500 6.33
One + 320 1418.891 2263 6.07
Two -4052 1439.606 566 6.24
Two + 299 1418.988 2255 5.98
For purposes of calibrating the galaxy profile, we need a temperature
in the velocity range of the galaxy. We will measure the mean amplitude
in the velocity range -500 km/s to -150 km/s in Figure 2 (tday9#20ple.ps).
We adopt as the flux density for this source the value 3.30 Jy, which is
the value measured at the VLA in August 1998 and again in August 2001.
IF Number Channel Number K/Jy
One One 6.109 1.851
One Two 5.991 1.816
Two One 6.111 1.852
Two Two 6.011 1.822
The gains in the two IF's are similar. Since we will measure the
properties of the galactic HI profile using the average of both
channels, we will adopt the mean K/Jy = 1.835.
As is discussed below, we expect that the rms for an observation
of five minutes On, five minutes Off, is 9.4 mK for a Tsys of 19 K,
or 12.4 mK when the system temperature is increased to 25 K as it
is on-source. In an area where the bandpass structure is small, for
example, near +3000 km/s (Figure 1: tday9#20plf.ps), the rms observed on
this calibrator is only slightly higher than expected, with values
17.2,14.1,14.2, and 14.5 mK for the four IF/channel combinations. In
regions where the structure changes rapidly, for example in channel
two between 150 km/s and 1700 km/s the rms values are much larger,
26.9,76.9,30.0 and 79.9 mK, and observations of features of width a
few hundred km/s would be limited in accuracy by the bandpass structure,
rather than by receiver noise.
Observations of the galaxy UGC668o
A total of eight pairs of scans were made for this galaxy. The On
scans are 18,22,24,26,28,30,32,and 34. All scans are usable, although
scan #22 shows three transient spikes between 1430 MHz and 1434 MHz
(see figure 3: tday9#22pla.ps).
Scan #30 shows a higher level of ripple
in the baselines than is seen in the other scans, but it has been
retained in the analysis. A linear baseline was removed from each scan
at each IF, and the rms values of the data in the region of the
baseline fit were compiled. The spectra derived by averaging all
scans are shown in Figure 4 (tday9#18pla.ps) and Figure 5 (tday9#18plb.ps)
for IF#1 and IF#2, respectively. In the figures the individual channels
are plotted in different colors.
The rms values computed for differing
integration times are tabulated in the following table. The expected
value of the rms is based on a receiver temperature of 19 K, a
frequency pixel width of 12.207 kHz, and Hanning smoothing.
Effective RMS, in mK
Integration IF#1 IF#2
Time (secs) Expected RX1 RX2 Both RX1 RX2 Both
150 9.93 10.39 10.34 9.85 10.14
300 7.02 7.77 7.59 7.48 7.09 7.37 7.15
600 4.96 5.58 5.56 5.64 5.31 5.47 5.24
1200 3.51 4.36 4.28 4.11 4.01 4.14 3.96
2400 2.48 3.26 3.02
The rms observed in one pair of Off/On scans is close to the expected
value. However, the rms does not integrate down with time as quickly
as is expected, suggesting that there is structure in the baseline
at low level which begins to dominate the receiver noise for integration
times longer than five minutes. The deviation from the expected values of
rms is shown graphically in Figure 6 (rmsif#1.ps) and Figure 7 (rmsif#2.ps).
We have compared the derived values from the HI profile of the galaxy
UGC668o with those listed by Fisher in his catalog. 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 above, K/Jy = 1.835.
Quantity Fisher IF#1 IF#2
Maximum in profile (Jy) 6.56 6.63 6.66
Line Width @ 20% (km/s) 32.5 33.5 33.2
Systemic Velocity (km/s) -231.6 -232.2 -231.4
Flux Integral (Jy km/s) 155.0 156.4 156.3
The spectra, expanded to be consistent with the display in Fisher's
catalog, are given in Figure 8 (tday9ava.ps) and Figure 9 (tday9avb.ps).
The agreement between the two IF's, and with the values from Fisher,
is excellent. The displacement of 4 MHz in the second IF has been
accounted for correctly.
For comparison, the line profile from the Fisher survey is shown
as Figure 10.
Back to Astronomical Checkout summary.