where the effective collecting area is assumed (as in CCIR 224-4, which is reproduced in Appendix A) to be that of an isotropic antenna, and is the incident flux density at the antenna. We therefore arrive at a tolerable noise flux density (in dB) of EFFECTS OF HARMONIC RADIATION Two principal effects of harmonic radiation must be considered: the possibility of saturating the front end with a harmonic signal, and the perturbing effects of a non-Gaussian noise on the VLBI signal sampling system. Saturation is a potential concern if the input noise power of the receiver doubles. A typical Mark II VLBI system has a 2-MHz bandwidth, so that the total input power for a 10 K system would be -167 dB W. A flux of harmonic power at the receiver of (-156 -20 log X)dB W/m£ would be harmful, under the assumptions of the preceding section. The second, less commonly considered, effect is that of a strong nonGaussian signal such as a spurious harmonic or side-band emission on a 1-bit sampling system. This in effect acts like a local oscillator in the receiving band, scrambling the received noise signal and making it impossible to recover the original data. The 1-bit sampling process is nonlinear, and so the requirement that the non-Gaussian signal be harmless is that its power be very small compared to the total noise power. A safe criterion appears to be that the amplitude of the discrete signal be less than 0.01 of the noise signal amplitude, i.e., its power should be 40 dB below the input noise power. Again, under the assumption of the first section, we obtain a harmful flux level Sng from this source of
RkJQdWJsaXNoZXIy MTU5NjU0Mg==