T. Zak et al., An experimental procedure to derive reliable IBIS models, PROCEEDINGS OF 3RD ELECTRONICS PACKAGING TECHNOLOGY CONFERENCE, 2000, pp. 339-344
One of the greater parts of PCBs research and development is to manage the
electromagnetic compatibility and signal integrity (EMC/SI) constraints. Ho
wever, the performances of these EMC/SI tools are based on the quality of u
sed library models: PCB interconnections and IBIS models (I/O Buffer Inform
ation Specification). To get an IBIS model, three ways are possible: the co
mponent maker using the web, Spice conversion if the corresponding Spice mo
del is available and measurements. This paper describes an experimental pro
cedure that allows creating From measurements reliable IBIS model version 2
.1. This procedure allows building models using classical measurement devic
es, controlling all steps of the procedure and achieving the required model
accuracy with regard to its application field. To check the accuracy of th
e developed procedure, the paper presents the characterization of a TQFP co
mponent. And through a classical SI phenomenon (reflection), the comparison
between measured and simulated results is performed, using the available I
BIS web model and the measured IBIS model.