In this paper, low frequency noise has been investigated in SWIR HgCdTe photodiodes from 4Hz to 3.2 KHz at different reverse bias. At low frequency the noise mainly consists of flicker noise and generation recombination (g-r) noise while at high frequency thermal noise is the dominant component. The flicker noise current is proportional to the detector current at small reverse bias, and the Hooge parameter αH of the device is evaluated. In addition, the low frequency noise at 100mV reverse bias measured from 250-300K are reported. The fluctuation time constant of g-r noise is extracted by fitting the curve of the low frequency noise, and the trap thermal activation energy of the deep level is obtained from the relation of τ and temperature.
The current-voltage (I-V) characteristics of HgCdTe photodiode under up to 33 dB of irradiance level have been investigated with a continuous-wave solid state laser. Short-circuit photocurrent and open-circuit photovoltage are observed to be saturated under high-level photon injection. It is found that the increasing of reverse bias can dramatically improve the response linearity of photodiode under high irradiance. The I-V characteristics and dynamic resistance versus bias relationship of photodiode under irradiance cannot be explained satisfactorily with present theory. Considering the effect of irradiance on photodiode series resistance and parallel resistance, an interpretation based abrupt junction model is presented to account for the effect of bias on I-V characteristics of HgCdTe photodiode.
This study is concerned with electrical shock effect on performance of the n-type HgCdTe photoconductive detector with a composition of x≈0.225 in order to improve the reliability under this certain situation. Ideally, responsivity of HgCdTe photoconductive detector is proportional to its bias current. Since there are a number of thermal interfaces between the detector and the cooling reservoir, the temperature of HgCdTe chip will rise and the response will decrease with the increase of bias current.
The detectors we used in the experiment were divided into two groups, one for room temperature experiment and the other for liquid nitrogen temperature experiment. A large pulse bias current was used to give the detector "an electrical shock". After the shock the resistance, responsivity and detectivity of detectors were measured at normal conditions. For the group of room temperature, the experiment results show that the detector responsivity will decrease under a 200mW pulse electric-power and the resistance will change under 400mW. When the pulse electric-power reaches 2000mW the detector will be burned out. But for the group of liquid nitrogen temperature experiment the detector responsivity will change under a 90mW pulse electric-power and the resistance will change under 290mW. A 550mW pulse electric-power will burn the detector out. Analysis with one-dimension approximation model was given, which showed that epoxy used between the interfaces was essential for increasing the reliability of HgCdTe detector.