Test Everything

Your Crystal-EM Hybrid is built. Now we characterize it. Open the Test Lab and we'll run five tests in order. After each, I'll tell you what the numbers mean and what 'good' looks like.

What it does: applies a small voltage across source–drain with the gate ON, measures resistance.

What good looks like: low resistance (a few kΩ or less for a small device). High resistance = broken channel or bad contacts.

What it does: applies mechanical stress to the piezo gate and measures the voltage it generates.

What good looks like: strong d₃₃-proportional output (PZT should give hundreds of mV per modest stress). Zero output = wrong gate material or no piezo layer.

What it does: sweeps drain voltage at several gate voltages and plots the family of curves.

What good looks like: clear cutoff, linear, and saturation regions; saturation current that grows steadily with V_GS. This is the device's full performance fingerprint.

What it does: sweeps gate voltage at fixed drain voltage. Tells you the threshold voltage (where the device turns on) and the subthreshold swing SS in mV/decade (how sharply it switches).

What good looks like: SS close to the 60 mV/decade thermionic limit. GAA + crystal devices can hit 65–75 mV/dec.

What it does: runs the device at typical operating conditions and reports total power (dynamic + static).

What good looks like: as low as you can make it. Crystal-EM with resonant EM gating should beat silicon by a wide margin because the gate signal arrives 'for free' near resonance.

Open a notes app. Write down three numbers from the tests above:

1. Maximum switching frequency (from I-V or Transfer Curve) 2. Subthreshold swing in mV/decade (from Transfer Curve) 3. Total power in µW or mW (from Power test)

You'll use these in Lesson 5 when you ask the AI Copilot to optimize them.