Here's something I've always wanted to try. It occurred to me that
someone on this list might have access to the equipment.
I've heard that low-voltage signals on wires are visible as dark/light
changes when viewed by a scanning electron microscope. I suspect that
the various potentials at the surface of the IC might be visible in the
SEM image as dark and light changes. The SiO2 oxide layer might
interfere with this, but if it does not, it means that it is possible
to indirectly SEE the changing potentials on a functioning IC.
To perform a simple test, put together an oscillator circuit based on
an LM741 or LM555 chip. Use an old TO-5 can rather than the usual DIP
package. Carefully cut the top away from the can so the silicon is
exposed, taking care not to break the tiny lead-bond wires. Connect
a battery, get the circuit working, then use the SEM to view the
IC surface while the circuit runs. With luck, the metal conductors
on the surface of the IC will flash dark grey and light as the
oscillator slowly runs.
If this works, then here's my REAL goal:
Wire up a single-chip microprocessor which has a UV-erase window.
Program it, remove the window, set up the clock speed to be something
around 3Hz, attach a battery, then view it under the SEM. The
thousands of conductors on the surface of the operating computer will
flash with black/white voltage signals.
So, what does Windows 95 look like when the hardware runs 50 million
times slower than normal? ;)
If this doesn't work, all may not be lost. Germanium and GaAs
technologies do not require an oxide layer, so it still might be possible
to find (or have made) a processor chip which has uninsulated
metallization which will give the desired interaction with the SEM beam.
((((((((((((((((((((( ( ( ( ( (O) ) ) ) ) )))))))))))))))))))))
William J. Beaty DESIGN ENGINEER
beatywj@ch.etn.com INDUSTRIAL PHOTOCONTROLS
EATON/CUTLER-HAMMER Everett, WA 206-353-0900