Comments, with reply, on 'Modeling of the 1/f noise overshoot in short-channel MOSFETs locally degraded by hot-carrier injection' by B. Boukriss et al

It is pointed out that a letter by Boukriss et al. (see ibid., vol.10, p.433-6, 1989) contains no reference to the actual value of the characteristic exponent of the noise spectrum ( gamma ). This omission may result from the assumption that gamma is equal to one. Similar measurements by the present author indicate that gamma varies strongly following hot-carrier injection near the drain. The change in the characteristic component is caused by the injected carriers present in the oxide near the interface, rather than the increased number of traps. This injected charge has an effect on noise provided the channel is not pinched off. It is argued that to study only the amplitude of the noise and not gamma misses critical information, the scaling properties, and that if one is to measure noise amplitude, one should measure the characteristic exponent as well. In a reply, the original authors state that the 1/f noise was studied for a nonhomogeneous MOSFET. The 1/f noise level was investigated from weak to strong inversion.<
>