Search Network

Polysilicon Fuses

30 November 2011
print version share on facebook share on linkedin

Executive summary

Developing a mathematical model of the physics of blowing polysilicon fuses.

Challenge overview

Polysilicon fuses are components of silicon chips which are blown as an irreversible programming step. Currently the ideal blowing voltage must be redetermined experimentally for every new chip design. These experiments are time consuming and delay the introduction of new products. Analog Devices contacted MACSI to ask if a mathematical model could be developed to model the fuse blowing process. This was a very challenging problem since the physical basis of fuse blowing was very poorly understood and only limited experimental evidence was available: namely the current-time transient of the blowing fuse and electron micrographs of blown fuses.

Implementation of the initiative

After initial analysis of the problem MACSI decided that the problem would benefit from an intensive brainstorming session involving using experts from a wide range of mathematical disciplines. Therefore the problem was included in the 62nd European Study Group with Industry: the first ever industrial study group held in Ireland, organised by MACSI at the University of Limerick. The study group format also simplified the consideration of intellectual property and confidentiality issues since results obtained by study groups are released into the public domain.

The problem

There were two challenges to this problem. The first was to determine the physics behind the blowing of the fuse; the second was to implement that understanding in a mathematical model. The first objective was achieved by considering a great number of fuse blowing scenarios and by order of magnitude estimates determine whether they were consistent with the observed phenomena. The final conclusion was that fuse blowing was the result of a complex combination of electronic, thermodynamic and fluid mechanical phenomena. To fulfil the second objective a 'multiphysics compartment model' was constructed, reducing the complexity of the blowing process to a handful of ordinary differential equations.

Results and achievements

Remarkably, given the extreme simplicity of the model its results show good agreement with experimental observations. This strongly suggests that the model has captured the underlying physics of the fuse blowing process. MACSI continue to collaborate with Analog Devices on fuse simulations with the aim of turning the mathematical model into a tool that can replace the experimental characterisation of fuses. Analog Devices also submitted a problem to the 70th European Study Group with Industry which was also organised by MACSI at the University of Limerick.


Sandra Healy,
Raheen Business Park
Raheen Limerick Tel : +353 61 229011
Web :


Attached Documents

PDF icon Irish Mathematics and Industry (Irish_Mathematics_and_Industry.pdf | 1.42 MB)