Observations made at CBAND October 29,2002
An observation of a continuum calibrator showed that the gain of the system varied from 2.4 K/Jy at 4850 MHz to 2.1 K/Jy at 5350 MHz. This is higher by 20% than the value observed in configuration 4W2-6XY-200 later in the same night. Birdies are seen near 4775, 4875, and 4975 MHz. There is an apparent reduction in the signal in the 15 MHz region between 5240 and 5255 MHz.
In this configuration the noise on a wide-band spectrum is dominated by the systematic effects of baseline ripple. In an observation of TMC-1 in which seven OFF/ON pairs were observed, the noise for a narrow signal reaches the value expected. However, over a wider bandwidth only three of the scans were usable. For the three scans the rms (3.4 mK) is close to the theoretical value, but was higher by 55% if all seven scans are used.
The velocity observed for the formaldehyde absorption line, 4829.557 MHz or +6.3 km/s, is consistent with the radial velocity expected (+5.8 km/s) for molecular lines in the source TMC-1.
In an observation of blank sky two of six scans were poor. Using the good scans only the system achieved the expected sensitivity, and the noise on the spectrum decreased with integration time in the expected manner. The effect of the baseline ripples in the poor scans is to increase the rms by 50%.
This configuration of the Spectrometer is ready for general use at CBAND in studies of spectral features which are narrower than 10-20 MHz.
Scans #18 to #27 were made with a pilot tone superimposed on the signal from NGC 7027. The power level for the tone was 70 dbm in all cases. The observations consisted of 15 seconds OFF, followed by 15 seconds ON. The observations were made without Doppler tracking, so it is expected that the pilot tone will appear at exactly the frequency of injection. In analyzing the pilot tone data we used only the OFF scan, since it provided a high signal-to-noise. The frequency observed each case was derived by averaging the frequency values of two points (one on either side of the peak) near the half-power position. These values are given in Table 1.
Table 1.
SCAN TONE OBSERVED LAB-OBS
NUMBER MHz MHz kHz
18 4870.000 4869.99561 +4.39
20 4795.000 4794.99613 +3.97
22 5035.000 5035.00157 -1.57
24 5034.680 5034.68323 -3.23
26 5034.968 5034.97026 -2.26
The differences are of order 3 kHz for a channel width of 97.656 kHz,
or about 4%. This is a reasonable measurement error. A subsequent test
using Gaussian fits gave residuals of about 1 kHz.
We conclude that there is satisfactory agreement between the value of the frequency at which the pilot tone is injected, and that at which it is observed. A more precise measurement of this must be made in a correlator configuration which has higher frequency resolution.
II. The continuum calibrator 0502+2516.
Scans #57 and #58 comprise a 5 minute Of/On sequence observing the
calibrator NVSS0502+2512=VLA0502+252=3C133, with a flux density
at 6cm of 2.37 Jy.
Figure 1 (tday11#58pla) shows the spectrum.
In addition to the baseline ripple there are "birdies" in the
intervals 4774.8-4776.0; 4874.2-4875.7; and 4973.2-4976.9 MHz.
The most significant feature is a region of reduced signal, between
5242 and 5255 MHz; the signal is reduced from 5.25 K to 3.50 K,
a loss of 1/3.
The following table gives average values for the observed antenna temperature of this calibrator.
Frequency Range Average Sensitivity
MHz Ta* K K/Jy
4800-4900 5.72 2.41
4900-5000 5.51 2.32
5000-5100 5.40 2.28
5100-5200 5.37 2.27
5300-5400 4.97 2.10
A pointing run was made on the source immediately before the
spectrometer observation, but the active surface was not in
operation. The elevation at the time of the observation was
68 degrees.
III. Observations of TMC-1
An observation of OMC-1 was judged to have too much unbalance between off source and on source to make it useful as a calibrator. The observations therefore concentrated on the molecular source TMC-1, and the formaldehyde absorption at a rest frequency of 4829.6639 MHz.
The first data base is comprised of a set of seven Off/On pairs,
scans #39-#51. It was apparent that two of the scans (#41, #49)
show considerable baseline structure with large rms, and two
more (#45, #51) are poor. For the three good scans (#39, #43,
and #47) the individual values of the rms are 5.0,5.0, and 6.0
mK; the rms of the average of the three is 3.4 mK.
The spectrum is shown in Figure 2 (tday11#39ava.ps).
For a system temperature of
23 K, and a channel width of 97.656 kHz, the expected rms in one
five-minute pair is 5.3 mK, giving 3.0 mK in the average of three
scans.
It is interesting to examine why the rms is higher in two of the
scans. Figure tday11#47pla.ps compares the three scans #47,#49,
and #51.
Figure 3:
Although the pattern of baseline ripple is quite similar
in all three scans, it is evident that for some reason the amplitude
of the ripple has increased dramatically in scan #49.
Figure 4 (tday11#49plb.ps) shows that during the first several of the
constituent 10-second integrations in scan #49 the baseline ripple
was small, but it increased in the later integrations.
 is greater than that using only the best data.
However, if the calculation of the rms is restricted to a small region
near the formaldehyde absorption feature,
then the value is improved by the inclusion of all data.
See Figure 5 (tday11#39avb.ps).
<img src=){kind=spacer}
The average
value of the rms in an individual scan is 4.9 mK, slightly smaller
than the expectation value, and the average for the seven scans is
1.8 mK compared with an expected value of 2.0.
We do not know the expected strength of this absorption line. The apparent intensity is -0.095 K. We can however compare the observed velocity of the feature with the expected value. The negative peak is estimated to occur at a frequency of 4829.5565 MHz (channel 1582.66). This differs from the expected frequency (rest frequency 4829.6639 MHz, radial velocity +5.8 km/s) by 107 kHz = 6.3 km/s. In view of the fact that the line is asymmetric because of the blending of multiple components, and given that the width of one channel is 97 kHz, or 5.8 km/s, this agreement is judged to be satisfactory.
IV. Observations of Blank Sky Near TMC-1
Because of our concern that the baseline ripple might be a strong function of source strength, or of the difference in power level between the Off and On positions, we made a sequence of observations in which two "OFF" positions were compared. The new "source" position was the OFF position from the earlier observation of TMC-1. The scans are #61-#71. Fours scans are of comparable quality, while two scans (#63, #69) exhibit rms values which are higher by about 50 % (8 mK instead of 5.3 mK).
To assess the rms we have averaged only the four good scans. The resultant
spectrum is shown in Figure 6 (tday11#61avb.ps).
Figure 6: 
The baseline structure is
present, as is the anomalous feature at 5250 MHz. The average rms of an
individual scan is 5.3 mK, and the rms of the average of four scans is
2.7 mK. The expected rms values are 5.0 mK and 2.5 mK, respectively,
assuming a system temperature of 22 K and 3-level sampling with Hanning
smoothing. Interestingly, the average spectrum using all six scans is
quite similar to this, and the rms is 2.6 mK. Using the additional two scans
offset the higher rms (6.2 mK) that inclusion of the poorer scans
introduced.
We conclude that the system achieved the expected sensitivity on this blank sky test, if the obviously poor scans with poor baselines are excluded. The noise in the spectrum decreased with integration time in the expected manner.