Micrologic 1960 Verify Circuit Function Wafer Truth Table

# Verify Circuit Function on the Wafer, Not Just One Transistor's Parameters: Testing a Truth Table, Not a Number

## 1. Why Passing Every Parameter Test Still Does Not Mean the Gate Works

This step probes each finished circuit while it is still on the wafer, exactly as the 1957 series documented at its own Step 16, except that the probe now exercises every combination of logic inputs this gate can receive and checks the output against the truth table the circuit was designed to implement, rather than measuring any single transistor's electrical parameters. A transistor measured individually can meet every specification — correct current gain, correct breakdown voltage, correct leakage — while the circuit built from several such transistors still fails to perform its logic function, because a wiring error, a resistor shifted by the contact effects Step 4 characterized, or an isolation island merged by the depletion-region overlap Step 3 warned against, would never show up in a single-transistor measurement at all. For a gate with $n$ logic inputs, this test can, in principle, be made exhaustive:

where $N_{\text{vectors}}$ is the number of distinct input combinations a complete test applies, and $C$ the fraction of possible single-node stuck-at-high or stuck-at-low faults this test set actually catches. For the small gates this series produces — two or three inputs — $N_{\text{vectors}}$ is small enough that exhaustive testing is entirely practical, and $C$ can approach completeness in a way this project's later, more complex circuits will not be able to afford.

## 2. Real Diagram: A Probe Card Applying Logic, Not Reading a Number

Where the 1957 series' wafer probe read a single electrical parameter from each die, this probe applies a sequence of input voltage patterns to the gate's input pads and compares the measured output pattern against the truth table — a fundamentally different kind of measurement, exercising the circuit's behavior rather than reading a static value.

## 3. Why Every Earlier Wafer Probe in This Project Checked a Number, and This One Checks a Behavior

The 1957 series' own wafer probe at Step 16 read a parameter — cutoff frequency, current gain, leakage — and compared it against a single acceptance band, exactly the kind of static measurement every transistor-level test this project has performed since 1954. This step cannot be satisfied with a static reading, because the thing it is testing is not a number a transistor produces but a *behavior* a circuit exhibits across a sequence of different conditions. That distinction — testing a function rather than a parameter — is the direct and necessary consequence of this series' first four steps building something with more than one component whose correctness depends on how those components cooperate, not merely on how any one of them measures in isolation.

Step 6 does not measure the chip; it interrogates it, across every input condition this test set can afford, to confirm that what several components built together is actually doing the one thing the circuit was designed to do.