Wall coding for field access bubble device

Abstract

We report results of wall state coding in a field access device. The wall states chosen to be logical ones and zeros are S=1 and S= ½ states, both of which have unsaturated capping layers. The 2.7 μm epi garnet (YSm Gd CaGe) is grown on top of a thin (.2 μm) epi garnet (Y Nd CaGe) layer with in-plane magnetization, called a boot. With the capping in-plane layer placed below the bubble film, the stability of the wall states is considerably enhanced against an upper cap switch ( S=1 \rightarrow S=0 ) caused by the fringe field of the permalloy. The device has a nucleator and chevron shift register which feeds into a current-access deflectometer where bubble wall states are differentiated by their deflection angle. After differentiation, S=1 bubbles go to a guard rail and S= ½ bubbles are sent to a chevron expander and detector. The bubbles are coded by applying current pulses to conductors which are placed adjacent to the shift register. The coder geometries will be described. Generally, a ½ state is made by supplying an in-plane field which adds to the external in-plane field and causes an upper cap switch, making an S=0 bubble. The lower cap switch is ∼80 Oe and the S=0 bubble decays to the S= ½ (logical zero) in the 50 Oe drive field. An S=1 state is formed by applying current such that an in-plane field subtracts from the external drive field causing the S= ½ \rightarrow S=1 transition by Bloch line annihilation. Coding margins of current, phase and bias have been taken for both states. We find that the margins are almost constant over the bias range of the deflectometer, which is 17 Oe. For ½ coding, the phase margin is 30° with a current margin of about 40%. The margins for coding the S=1 state are somewhat narrower with phase \sim25\deg and current 20%.