Signal to noise ratio comparison of SAL versus dual-stripe magnetoresistive heads under comparable reliability conditions

Abstract

At high current densities the reliability of a magnetoresistive head is predominantly determined by the temperature rise of the sensor. Current dual-stripe magnetoresistive head (DSMR) designs operate two sensors, each at 2×107 A/cm2, resulting in twice the temperature rise of a typical SAL head. The proper design scaling for reliability would require a similar temperature rise for both heads with each sensor operating at 2×107 A/cm2. In this article the implications of this assumption are investigated with respect to the signal to noise ratio. The scaled DSMR design requires thinner sensors and lower bias current resulting in a higher output. The outputs of the SAL and DSMR heads are baselined via experimental data. A micromagnetic model is used to scale the results for comparison. The higher sensor output of the DSMR is compensated by the increase in preamp noise resulting from a differential versus single ended preamplifier, and from the increased Johnson noise due to higher resistance DSMR elements. Several limits are examined relative to the media noise and the bandwidth of the system. Practical resistances of the leads are included in the analysis as well as preamp noise figures. The results show a 0.2 dB gain of the DSMR head relative to the SAL head, not including linear density differences. This is at constant current density and constant joule heating. The comparison is expected to worsen at constant linear density. For a constant 25 μV noise floor the higher output of the DSMR design gives a 3 dB improvement over SAL. Thinning the SAL sensor to the same thickness as the DSMR sensor largely negates this advantage. The DSMR design does not show a definitive advantage over SAL despite claims often referred to in the literature. The role of DSMR heads in future areal density improvements will be discussed based upon these results.