What is Jitter?

Think of a pulse train as a rhythmic beat. If every pulse arrives exactly when expected, jitter is zero. If some arrive slightly early and others slightly late, the signal has jitter.

Synchronous vs. Asynchronous Timing

The approach to jitter depends on how the transmitter and receiver agree on time.

• Synchronous: A shared clock signal is sent alongside the data. Common in local PCB buses (e.g., DDR Memory). If the clock signal and data signal wander apart (Skew), the system fails.
• Asynchronous: No shared clock. The receiver must 'guess' the clock by looking at the data transitions. This is used in almost all long-haul networking (Ethernet, Fiber).

Clock and Data Recovery (CDR)

In asynchronous systems, we use a Phase-Locked Loop (PLL) to perform Clock Recovery. The PLL 'locks' onto the incoming data transitions and generates a stable internal clock to sample the data.

Conclusion

As we move to 100G and 400G networking, we are fighting physics in the picosecond domain. Understanding the sources of jitter—from noisy power supplies to poor cable shielding—is the hallmark of a master network engineer.