Search

A rate of 0.3 failures per billion hours or less is desirable for memory components in large integrated arrays. This unusually stringent requirement complicates the determination of lifetime from accelerated aging studies. The value and limitations of step stress aging techniques are discussed in terms of experimental results obtained using plated wire memory arrays designed to withstand the high ambient temperatures required for accelerated aging. Step stress aging measurements alone are insufficient for confident lifetime prediction. Therefore, longer term measurements at lower temperatures must also be made to establish the validity of the lifetime extrapolations. It is essential to protect the plated wires against corrosion. Given proper protection a shelf life in the hundreds of years is forecast. The importance of duty cycle on lifetime in exercising the memory is discussed and the results of aging under extreme pulsed magnetic field stress conditions are reported. Criteria for wire selection, with long term stability in mind, are discussed.

This paper discusses the design and characteristics of ruby traveling-wave masers operating at 4 gc. These masers, characterized by an average gain of ≈35 db over a bandwidth of 25 mc, are equipped with waveguide input transmission lines, rather than the previously employed coaxial cables. This change results in an over-all noise temperature of 3.5°K for these devices, rather than the 10°K exhibited by earlier masers. The maser noise temperature now closely approximates sky temperatures, which set the ultimate limit on earthbound receiver sensitivity. The improvements to be had by further reduction in amplifier noise are therefore almost negligible. A less well known maser property, i.e., its freedom from distortion, even when driven well into gain saturation, is discussed