A logic probe

// A logic probe, and the fault a voltmeter cannot see.
//
// A 3 Hz clock into one gate of a 74HC04, and an LED on the output so you can
// watch it. Two probes:
//
//   the left one   on the inverter's output. It flips between H and L three
//                  times a second and PULSE stays lit, because the probe
//                  stretches every edge to a fifth of a second so an eye can
//                  see it.
//   the right one  on pin 3, the input of the next inverter along, which is
//                  wired to nothing. Every lamp is dark. Not low: dark.
//
// That is the difference the probe exists for. Put a voltmeter on a floating
// CMOS input and you get a number, and the number means nothing: nothing is
// driving the pin, so what is on it is whatever leaked there. The probe says
// so by saying nothing, and no other instrument on this bench does.
//
// The thresholds are not the probe's. What each lamp is showing is the level
// the *driver* declared, so a 3.3 V board's pin is a clean H here for the same
// reason it is one on the board. There is no TTL/CMOS switch to get wrong,
// because there is no number inside the probe to be wrong. The voltage under
// the lamps is there for the case with no declared level at all: a divider, a
// capacitor charging, an analog node on its way somewhere.
//
// Things to try. Wire pin 3 to the ground rail and the right-hand probe goes
// from dark to L. Wire it to the 5 V rail instead and it goes to H. Wire it to
// both at once and it goes to ?: two drivers disagreeing, which is
// contention, and which is a real fault a real probe shows the same way.