Performance of the GBT at 43 GHz (7mm).

Tests done November 22, 2003, including estimates of efficiency, and test observation of an SiO maser source.
December 1, 2003; revised Dec 11.
F.Ghigo

Tests of the GBT performance in the 7mm band (Q-band) were done 22 November 2003. Because the L1 receiver was not working, we used only beam number two, receptors L2 and R2, and did observing in total power mode -- no beam switching. The gain versus elevation was measured by observing several strong calibration sources over a wide range of elevations. The subreflector focus was adjusted before each peak procedure. The observations were done at a frequency of 43.124 GHz with a bandwidth of 0.320 GHz. The frequency of 43.124 was chosen to match the VLBA 7mm band.

System Temperature Measurements

The Tsys values as calculated by "GOpoint" from each "peak" procedure, are plotted in Figures 1&2 for all data. We can see that the data are consistent with no changes in the atmospheric characteristics over the whole 16-hour observing session.

A simple model for the atmosphere was fit to the data:
TSYS = Trcvr + Tatm ( 1 - exp( -A*TAU) )
in which A = 1/sin(Elev)

The fitted parameters are as follows:
Parameter Fit to LCP data Fit to RCP data
Trcvr 28.4 K 39.2 K
Tatm 203 K 251 K
TAU 0.098 0.091
Thus we can see that the atmospheric transparency is characterized by a TAU of about 0.1, which is pretty respectable for Green Bank!
We also note that the RCP temperatures are consistently about 10 degrees higher than LCP.

Note that Tatm and TAU are highly correlated, so that we can determine the product Tatm*TAU, but either of these terms by itself is not well determined. An analysis by Ron Maddalena suggests the LCP TCals are probably too low and that the slope of Tsys vs A for LCP should be close to that for RCP.

Aperture Efficiency Measurements

Estimates of the efficiency (E') are plotted in Figure 3 for the source 3C286, assuming a flux density of 1.45 Jy. References differ on the flux density of 3C286 at 43 GHz, ranging from 1.35 to 1.8 Jy. I have used 1.45, the value from the VLA calibrator list.
Figure 3 shows the efficiency (E') for RCP (diamonds) and LCP (triangles). The solid curve is the expected atmospheric absorption for TAU=0.1.

E' = (2k/A) (Tant/S) i.e., the "raw" efficiency uncorrected for atmospheric absorption.

The efficiency has a maximum at an elevation of about 50 degrees, then drops and becomes double valued. This can be understood by referring to Figure 4, which shows the data as a function of scan number, which really means time, since scan number increases throughout the observing run. The elevation is plotted as the open stars using the right-hand scale on the plot. The left hand scale is used both for the ambient temperature in Centigrade and the Ap. Eff in percent.

The local time runs from 02:50 AM (scan # 330) to 11:35 AM (scan 802).

The green symbols near near the bottom are the ambient temperature in Centigrade. One may note that the temperature stays fairly constant at about -4C until scan number 600, which corresponds to sunrise. After this, the temperature rises rapidly. As the temperature rises, the efficiency drops.

Corrected Aperture Efficiency

In the above plots, the "Efficiency" has not been corrected for atmospheric absorption. The following plot shows the values corrected for an atmosphere with tau=0.08.
i.e, E' * exp( 0.08*secZ)

In this plot, the RCP and LCP data have been averaged, and we have plotted only the night-time data.

The numbers that are plotted here are listed in the Table of Data.

Subreflector Y-focus corrections.

The focus corrections that were used for the 3C286 observations are plotted in Figure 5, as a function of elevation. The effect of rising temperature in the latter part of the observing run probably explains the hook in the upper right part of the plot.

Temp curve for 3C84

We display the antenna temp vs elevation for 3C84. This was not converted to efficiency because of the uncertainty in the flux density. The ambient temperature is shown at the bottom of the plot -- the observations were all at night, and the temperature only varied by less than 3 degrees Centigrade. The fall-off at high elevation is smooth.

Spectrum of R Aqr

A spectrum was made of the star R Aqr, a source of SiO maser emission. The Spectrometer was used in 50 MHz mode with 16384 channels, i.e., 3 kHz channel spacing. A series of 7 one-minute off/on observations were done. The on-source spectra were corrected by doing: Tsys*(on-off)/off.

Plot of the average of 7 off-corrected spectra.

An expansion of the vertical scale shows that ripples remain in the baseline after correction for the off-spectrum. Possibly doing beam switching and/or nodding will improve this.

The rms noise in the spectrum was calculated for a flat part of the spectrum (43.111 to 43.122 GHz), and is reasonably close to the predicted rms from the radiometer equation.

RMS noise in flat part of spectrum
as measured, and as calculated from the radiometer equation.
t(ONsource) rms(LCP) mK rms(RCP) mK rms(Calc)
60 s 162 160 169
120 s 109 115 119
240 s 79 84 84
420 s 68 60 64
The following plot shows the rms as function of on-source integration time. The solid line is the prediction from the radiometer equation. A system temperature of 70 K was used for the calculation.