2008/2009 Remarkable interest has recently arisen about the search for Weakly Inter- acting Sub-eV Particles (WISPs), such as axions, Axion Like Particles (ALPs), Minicharged and chameleon particles, all of which are not included in the Stan- dard Model. Precision experiments searching for WISPs probe energy scales as high as 10^6 TeV and are complementary to accelerator experiments, where the energy scale is a few TeV. The axion, in particular, is the oldest studied and has the strongest theoretical motivation, having its origin in Quantum Chromodynamics. It was introduced for the first time in 1973 by Peccei and Quinn to solve the strong CP problem, while later on the cosmological implications of its postulated existence also became clear: it is a good candidate for the cold dark matter, and it is necessary to fully explain the evolution of galaxies. Among the different interactions of axions, the most promising for its detection, from an experimental point of view, is the coupling to two photons (Primakoff effect). Using this coupling, several bounds on the axion mass and energy scale have been set by astrophysical observations, by laboratory experiments and by the direct observation of celestial bodies, such as the Sun. Most of these considerations, as was recently recognized, not only constrain the mass and coupling of the axion, but are more generally applicable to all ALPs. The current best limits on the coupling, over a wide range of ALP masses, come from the the CAST (Cern Axion Solar Telescope) experiment at Cern, which looks for ALPs produced in the solar core. The experiment is based on the Primakoff effect in a high magnetic field, where solar ALPs can be reconverted in photons. The CAST magnet, a 10 T, 10 m long LHC superconducting dipole, is placed on a mobile platform in order to follow the Sun twice a day, during sunrise and sunset, and has two straight bores instrumented with X-ray detectors at each end. The re- generated photon flux is, in fact, expected to be peaked at a few keV. On the other hand, there are suggestions that the problem of the anomalous temperature profile of the solar corona could be solved by a mechanism which could enhance the low energy tail of the regenerated photon spectrum. A low energy photon counter has, for this reason, been designed and built to cover one of the CAST ports, at least temporarily. Low energy, low background photon counters such as the one just mentioned, are also crucial for most experiments searching for WISPs. The low energy photon counting system initially developed to be coupled to CAST will be applicable, with proper upgrades, to other WISPs search experiments. It consists of a Galilean telescope to match the CAST magnet bore cross section to an optical fiber leading photons to the sensors, passing first through an optical switch. This last device allows one to share input photons between two different detectors, and to acquire light and background data simultaneously. The sensors at the end of this chain are a photomultiplier tube and an avalanche photodiode operated in Geiger mode. Each detector was preliminary characterized on a test bench, then it was coupled to the optical system. The final integrated setup was subsequently mounted on one of the CAST magnet bores. A set of measurements, including live sun tracking, was carried out at Cern during 2007-2008. The background ob- tained there was the same measured in the test bench measurements, around 0.4 Hz, but it is clear that to progress from these preliminary measurements a lower background sensor is needed. Different types of detectors were considered and the final choice fell on a Geiger mode avalanche photodiode (G-APD) cooled at liquid nitrogen temperature. The aim is to drastically reduce the dark count rate, al- though an increase in the afterpulsing phenomenon is expected. Since the detector is designed to be operated in a scenario where a very low rate of signal photons is predicted, the afterpulsing effect can be accepted and corrected by an increase in the detector dead time. First results show that a reduction in background of a factor better than 10^4 is obtained, with no loss in quantum e ciency. In addition, an optical system based on a semitransparent mirror (transparent to X-rays and re ective for 1-2 eV photons) has been built. This setup, covering the low energy spectrum of solar ALPs, will be installed permanently on the CAST beamline. Current work is centered on further tests on the liquid nitrogen cooled G-APD concept involving different types of sensors and different layouts of the front-end read-out electronics, with a particular attention to the quenching cir- cuit, whether active or passive. Once these detector studies are completed, the final low background sensor will be installed on the CAST experiment. It is important to note that the use of a single photon counter for low energy photons having a good enough background (