International Business Machines Corporation
Detecting damaged TMR sensors using bias currents and outliers

Abstract:

A computer-implemented method to detect a damaged tunneling magnetoresistance (TMR) sensor includes applying current at at least two different current values to the TMR sensor and measuring a resistance, R.sub.TMR, at each current value. The method also includes measuring a slope in resistance vs. bias current, RD.sub.SLP, using the measured resistances R.sub.TMR and the at least two different current values. The method includes calculating a .DELTA.RD.sub.SLP value as a difference between the RD.sub.SLP value and an expected value, RD.sub.SLP-expected, for the TMR sensor. The method includes determining whether the .DELTA.RD.sub.SLP value is within a predefined range. In response to determining that the .DELTA.RD.sub.SLP value is outside the predefined range, the method includes outputting an indication that the TMR sensor fails. In response to determining that the .DELTA.RD.sub.SLP value is within the predefined range, the method includes outputting an indication that the TMR sensor passes.