Noise in photon detectors

Noise arises in semiconductor detectors from several mechanisms. Johnson noise is found in all resistive elements. It has already been discussed in coimection with thermal detectors [see Subsection 5.1 l.b and Eq. (5.11.20)]. If the load resistance in the circuit is larger than the detector resistance, the Johnson noise of the detector element dominates because load and detector act electrically in parallel as far as the noise properties are concerned. 

Another type of noise, often dominating at low modulation frequency, is called 1// noise because the noise power is inversely proportional to frequency. Its physical cause is not well-understood. In PV detectors it is proportional to the direct current flowing through the detector, and can be greatly reduced by operating the detector with a small d.c. bias. Because PC detectors must operate with a d.c. current, 1// noise is always present, particularly at low modulation frequencies. With advanced fabrication techniques, the level of this noise can be reduced below other, more dominant noise sources. 

Generation-recombination (G-R) noise is found in PC detectors, and is caused by random fluctuations of charge carriers. These fluctuations can be due to thermal excitation within the semiconductor. Sometimes G-R noise is also defined to include random arrival of photons at the detector. PC detectors are normally operated at temperatures low enough to reduce the thermal generation of carriers well below that of all other noise sources. The residual G-R noise is then the photon noise already discussed in Subsection 5.1 l.b [see Eq. (5.11.33)]. Photon noise can also be understood in terms of the Poisson probability of n photons arriving during a given time interval. 

The photon noise is simply due to the uncertainty, a, in the number of photons per second, i.e., the detected signal level. In the photon noise limit the signal-to-noise ratio is proportional to the square root of the photon arrival rate. When the measured noise is the photon noise in the background radiation, the detection is said to have reached background limited performance (BLIP). 

The total noise voltage can be found from the individual noise voltages by adding the squares of the individual rms noise voltages [see Eq. (5.11.21)]. Whenever possible, detectors are operated with sufficient incident radiation to achieve BLIP performance, that is, where the background noise is larger than the noise from all other sources combined. For ground-based observations the noise is then due only to the radiation from the source, sky, and warm optical elements in the field of view of the detector. 

---

In this section, we give a brief phenomenological description of noise in photonic detectors. Only the most basic practical details are included necessary for us to consider the possibilities to decrease it. An attempt has been made to give a theory valid simultaneously for photoconductors and photovoltaics. 

---