Chip Reverse Engineering

From Tmplab
Revision as of 11:22, 24 February 2009 by Iv (talk | contribs) (traduction en cours...)

Microscope

Thanks to Karsten Nohl presentation at Hacker Space Fest, there's an interesting DIY approach for reverse engineering.

Now we have to find a microscope...

  • Friend of Phil will lend a microscope with camera
    • not perfect but should enable us to do the first etching/abrasive steps
  • We'll try to select and acquire a perfect microscope for this job.

Features

These are the features we're looking for:

  • 500x magnification
  • 2 MP to 8 MP camera
  • Objective table to move the chip slowly (i.e. not by hand, too much shaking and imprecision)
  • Reflected (incident) lighting

Feature we DON'T NEED:

  • Confocal microscope (it seems, because we'll abrase/grind away the upper layers anyway and the bottom layers are opaque anyway too)

Links

Perfect Microscope

Still selecting the right model.

Leads

Poor man microscope setup

  • Microscope: 640x optical Bresser Biolux
    • lent by Phil's friend
  • Objective table:
    • micro-screws controlled moving plate
    • possibility to buy one: [1]
  • Camera: 3.2 MP Canon Ixus V3
    • Mounting: Attached directly to the eye-piece
    • Drawback: produces round (not square) pictures (with no noticeable deformation) of smaller size/coverage than regular microscope-camera.
    • Advantage: working and free
  • Camera: 0.3 MP Ulead IC Chip camera
    • Drawback: Low res.
    • Advantage: square images, adapted for the scope.


Results so far

  • Req: 500x magnification or higher
    • This one goes up to 640x
  • Req: 2 MP to 8 MP camera
    • The Canon IXUS is 3.2 MP
    • Will need to take more photo as just 2/3rd of the picture is usable (round pictures)
  • Req: Objective table to move the chip slowly (i.e. not by hand, too much shaking and imprecision)
    • Poor man setup with micro-screws, TO DO
  • Req: Reflected (incident) lighting
    • Reflected light works at lower magnification with daylight
    • Even for microscope which is designed for see-through
    • TODO: Need to buy adequate lighting. Got to test with external lighting (HB LEDs?) at higher magnification (640x)

Abrasion

  • What kind of glue to use to attach the chip to solid/manoeuvrable support?
  • Abrasive paste

Software

Documentation

Microscope

La présentation de Karsten Nohl (en) au Hacker Space Fest ouvre la voie à des bricolages interessants de reverse engineering.

Maintenant il nous faut un microscope.

  • Un ami de Phil prêtera un microscope avec camera
    • Pas l'idéal mais devrait suffire pour les premières étapes d'abrasion et de grattage.
  • On va essayer de choisir et de se procurer le microscope idéal à cet emploi.

Fonctionnalités

Voila ce qu'on cherche:

  • grossissement 500x
  • camera de 2 à 8 megapixels
  • Table d'objectif qui puisse faire bouger lentement la puce (c'est à dire pas à la main, ça tremble et c'est trop imprécis)
  • Lumière réfléchie (incidente)

Ce dont on n'a PAS besoin :

  • Microscope confocal (en tous cas on dirait pas, puisque on va gratter et retirer les couches supérieures et que les couches inférieures sont opaques de toutes façons)


Liens (en)