This directory contains a number of environments that are used, in
conjunction with ../models.m8cbscan, to try to identify various
components attached to a test target.

The script "all" executes all these models, and reports whether
they are congruous with the measured behaviour or not.

There are the results obtained with a CY8C29466, operating at 5V:

Single pin:

                             NC              PULL_UP
                             |   GND         |   DIODE_R_DOWN
                             |   |   VDD     |   |   DIODE_R_UP
                             |   |   |   PULL_DOWN   |
                             |   |   |   |   |   |   |
Not connected                Y   .   .   Y   Y   Y   Y
Shorted to GND               .   Y   .   .   .   .   .
Shorted to Vdd               .   .   Y   .   .   .   .
1k Ohm pull-down to GND      .   Y   .   Y   .   Y   .
10k Ohm pull-down to GND     Y   .   .   Y   .   Y   Y
1k Ohm pull-up to Vdd        .   .   Y   .   Y   .   Y
10k Ohm pull-up to Vdd       .   .   Y   .   Y   .   Y
Diode and 330 Ohm, to GND    Y   .   .   Y   .   Y   Y
Diode and 330 Ohm, to Vdd    .   .   Y   .   Y   .   Y


Pair of pins:

                               NC          DIODE
                               |   WIRE    |   DIODE_R
                               |   |   RESISTOR|
                               |   |   |   |   |
Not connected                  Y   .   .   .   .
Wire                           .   Y   .   .   .
1 kOhm                         .   Y   Y   .   .
10 kOhm                        .   .   Y   .   .
Diode, forward                 .   .   Y   Y   Y
Diode, reversed                .   .   .   .   .
Diode and 1 kOhm, forward      .   .   .   Y   Y
Diode and 1 kOhm, reversed     .   .   .   .   .
Diode and 10 kOhm, forward     .   .   Y   .   Y
Diode and 10 kOhm, reversed    .   .   .   .   .

Note that some of the false positives could be avoided by forcing
m8cbscan to also try configurations where strong drives are expected
to collide (e.g., "1" into GND). Also note that such collisions can
happen during regular testing, if the model and the actual circuit
differ.

Also note that some results are essentially non-deterministic and may
even depend on parasitic effects, e.g., anything based on interpreting
the signal on a pin driven to Z from both sides (i.e., because the
circuit attached to the pin does not correspond to the model), will
basically only capture interferences and parasitic capacitances, and
may thus yield different results when repeated.
