• DocumentCode
    2774642
  • Title

    Generation and verification of boundary independent compact thermal models for active components according to the DELPHI/SEED methods

  • Author

    Pape, Heinz ; Noebauer, Gerhard

  • Author_Institution
    Siemens Semicond. Group, Muenchen, Germany
  • fYear
    1999
  • fDate
    9-11 March 1999
  • Firstpage
    201
  • Lastpage
    211
  • Abstract
    In the European project SEED (Supplier Evaluation and Exploitation of DELPHI), methods for thermal characterization of active components developed in DELPHI were evaluated by component suppliers. The methods were improved for practical application and extended to all IC package types. Test parts were chosen to cover as wide as possible a range of different plastic packages currently in use, including comparison of Alloy 42 (FeNi42) and Cu-based leadframe material, as well as a large and a small chip inside the same package, standard and thermally enhanced power devices, DSO, QFP and BGA packages. For all device types investigated, it was possible to generate simple resistor networks which reproduce junction temperatures and fluxes of a detailed finite element model for all 38 boundary conditions suggested in DELPHI with an average accuracy of 1-2%. Maximum errors are in general about 10% or less. As result of the SEED project, methods for thermal characterization of active components are available, which not only work in practice, but are also robust and low cost, because no test PCB is needed for measurements. Physically, they are much better suited to model validations than existing thermal resistance measurements. The resulting thermal resistor network allows end users for the first time to routinely predict the junction temperatures of components in their specific applications with an acceptable accuracy. The final goal of a simple and universal thermal characterization of electronic components is achieved. Standardization and adoption is in progress.
  • Keywords
    ball grid arrays; error analysis; finite element analysis; integrated circuit modelling; integrated circuit packaging; integrated circuit testing; plastic packaging; standardisation; thermal analysis; thermal management (packaging); thermal resistance; Alloy 42 leadframe material; BGA packages; Cu; Cu-based leadframe material; DELPHI; DELPHI/SEED methods; DSO packages; FeNi; IC package types; QFP packages; SEED European project; SEED project; Supplier Evaluation and Exploitation of DELPHI project; active components; boundary conditions; boundary independent compact thermal models; chip size; component junction temperature prediction; device types; electronic components; finite element model; junction fluxes; junction temperatures; maximum errors; model validation; plastic packages; resistor networks; standardization; test parts; thermal characterization; thermal model generation; thermal model verification; thermal resistance measurements; thermal resistor network; thermally enhanced power devices; Application specific integrated circuits; Electrical resistance measurement; Electronic packaging thermal management; Integrated circuit packaging; Lead; Materials testing; Plastic packaging; Resistors; Temperature; Thermal resistance;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Semiconductor Thermal Measurement and Management Symposium, 1999. Fifteenth Annual IEEE
  • Conference_Location
    San Diego, CA, USA
  • ISSN
    1065-2221
  • Print_ISBN
    0-7803-5264-5
  • Type

    conf

  • DOI
    10.1109/STHERM.1999.762449
  • Filename
    762449