Setting the Observing Frequency and the Bandwidth

Once the apparent frequency $\nu_{o}$ of the transition of interest is known, the Local Oscillator (LO) frequencies can be tuned to select this frequency for observations. In general, there can be more than one LO that need to be tuned. Consider the situation at the GMRT. The First LO ($\nu_{ILO}$) can be chosen such that $\nu_{ILO}$ = $\nu_{o} \pm \nu_{IF}$, where, $\nu_{IF}$ is the Intermediate Frequency (IF). The First LO can be tuned in steps of 5 MHz. The IF is 70 MHz. The IF bandwidth ( $\delta\nu_{IF}$) can be chosen from one of 6, 16, and 32 MHz. Thus, the output of the first mixer will be over a frequency range of $\nu_{IF} \pm \delta\nu_{IF}$/2. The baseband LO ($\nu_{BBLO}$) can be tuned in the range of 50 to 90 MHz in steps of 100 Hz to bring the IF down to the baseband. The bandwidth of the baseband filter ( $\delta\nu_{BB}$) can be chosen from 62.5 KHz to 16 MHz in steps of 2. The bands from $-\delta\nu_{BB}$/2 to 0, and from 0 to $\delta\nu_{BB}$/2, which are the lower, and the upper side bands respectively, will be processed separately. The FX Correlator at the GMRT will produce 128 spectral channels (0 - 127) covering each of these bands. The 0th channel corresponds to a frequency of $\nu_{o}+\nu_{BBLO}-\nu_{IF}$ and the frequency increases with channel number in the USB spectrum and decreases with channel number in the LSB spectrum.

While setting the LO frequencies one needs to make sure that (a) the desired LO frequency is in the allowed range and that the oscillator is $'$locked$'$ to a stable reference, and, (b) that the required power output is available from the oscillator. The choice of the baseband filter bandwidth depends on the velocity resolution and the velocity coverage required for a given observation. In addition, it is preferable to have as many line-free channels in the band as there are channels with the line in order to be able to obtain a good estimate of the observed baseline (or reference spectrum). One would also like to center the spectral feature within the observed band so that line-free channels on either side can be used to estimate the baseline. The velocity resolution should be at least a factor of two better than the full width at half maximum of the narrowest feature one is expecting to detect.

At present, the FX Correlator at the GMRT produces 128 channels per side band for each of the two polarizations. The two polarizations are identified as the 130 MHz and the 175 MHz channels. In principle it should be possible to drop one of the polarizations to obtain 256 channels for one polarization. This will improve the spectral resolution by a factor of 2 keeping the velocity coverage (the bandwidth) the same. This can be very useful in observing narrow lines over a wider range of velocities.