Logic probe

The pen with a needle on the end. Touch it on a pin and it tells you one of four things, and the fourth is one no other instrument here can tell you.

Pins

Pin	What it does
TIP	The needle. 10 Mohm to the ground clip and nothing else.
GND	The ground clip, out of the back of the pen.

What the lamps mean

lamp	what it means
H	Something is driving the tip high.
L	Something is driving the tip low.
?	Two drivers disagree: the net is in contention.
(none lit)	Nothing is driving the tip at all.

and across all of them, P (pulsing), which lights when the tip changed level inside the last fifth of a second, with the frequency in the tooltip. A real probe stretches a pulse so an eye can catch it; so does this one.

The dark face is the reading worth owning a probe 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 last. The probe says so by saying nothing.

The thresholds are not the probe's

This is the one decision in the part, and it is why there is no TTL/CMOS switch on it. A real logic probe has that switch and reads wrong when it is on the other one. This part is not deciding: what it shows is the level the driver declared. A part that knows its own supply says which level it is driving, so an ESP32-C3 pin at 3.3 V is a clean H here for exactly the reason it is one on the board, and a 5 V AVR pin is a clean H for its own reason. There is no number inside the probe, so there is no number to be wrong.

Where nothing declared a level (a resistive divider, an analog node, a capacitor on its way up), the kernel falls back to half of five volts and the probe shows what that says. The voltage printed under the lamps is there for exactly that case, so a node sitting at 1.8 V is visibly not a clean anything.

What it does not model

Ten megohms to the ground clip is the whole of its input: no capacitance, so touching it on a high-impedance node does not disturb the node the way a real probe's few picofarads do. No minimum pulse width, so a nanosecond glitch lights P where a real probe's stretcher would have had to be quick enough to catch it. No audible tone, because the lamp says the same thing; the one noise on this bench is the multimeter's continuity beeper. No memory or single-shot latch, and no pulse-train/level switch.

See it in action

A logic probe has two of them: one on a pin an inverter is driving and one on a pin wired to nothing, which is the whole point side by side. See also the bench instruments.