Measurements in 1149.4 environments - correcting the infrastructure switches influence

Abstract

Measuring the values of discrete components frequently takes place during the test or debug phase of Printed Circuit Boards (PCB). This operation requires tools that are based on some access type. The shrinking geometries constrain the straightforward use of tools based on physical access. One of the aims of the IEEE1149.4 Std. is to facilitate those on-board measurements. This infrastructure relies on electronic access that includes high quality analog buses and a set of electronic switches, which enable to completely isolate a component under characterization, e.g. by injecting a known current and measuring the voltage across it. During this process, the infrastructure switches have a negative impact in the measurement accuracy. This paper analyses the measurement of one resistor in two situations: connected between a pin and ground and between two pins. The infrastructure switches that affect the measurement quality are identified and the upper limit of its systematic error is characterized. When the systematic error is completely defined then it is possible to remove its negative effect from the final result.