Overdriven vs underdriven lines
What do "overdriven" and "underdriven" lines look like on a scope, and how do you fix each?
| Level | ●● onsite technical |
| Track | SI & high-speed |
| Really testing | Diagnosing termination from the waveform. |
Going deeper
Extra explanation that is not in the book: worked numbers, where people go wrong, and how to say it in the room.
Overdriven, step by step. A 25 Ω driver on a 50 Ω line into an open receiver: Γs = −1/3. The launched step is 2/3 of the source voltage, doubled at the open end to 1.33. Then 0.89, 1.04, 0.99 … — overshoot, then ringing that settles.
Underdriven, step by step. A 100 Ω driver on the same line: Γs = +1/3. Launched 1/3, so the receiver sees 0.67, then 0.89, then 0.96 … — a staircase.
What the scope will tell you. The time between steps or ring peaks is the round trip (2·TD). Measure it and you know the line length — a fast way to check which trace you are looking at.
Where people go wrong
- Probing at the driver: with series termination the driver-end waveform shows a half-step shelf that looks like a fault but is correct.
- Assuming the driver is a fixed resistor — its output resistance changes with voltage and process; check IBIS models across corners.
Errata
No corrections known for this question. All errata
Ask about this question
Stuck on a step, or think something is wrong? Send it with SI-03 in the subject. Good questions get answered on this page, and corrections go into the next printing.
Related
← SI-02 Reflections and the reflection coefficientSI-04 Termination choices →