Signal integrity concerns whether the receiver sees the required waveform at the required time. Use a review process that links a symptom to a physical mechanism and then to a design decision. This open workbook is intended for on-page reading and project review; it does not require a gated download.
Define the receiver decision
Identify the voltage and timing conditions that determine success at the receiver. Record applicable eye, jitter, noise-margin or settling requirements. A visually clean oscilloscope trace is not enough when its reference plane, bandwidth or timing relationship does not match the acceptance criterion.
Choose models with a stated purpose
Driver and receiver models, transmission-line models and S-parameters answer different parts of the problem. Check model versions, port definitions, operating corners and frequency range. Keep a simple model that explains the dominant behavior alongside the more detailed model used for acceptance.
Separate the main failure mechanisms
Evaluate reflections from discontinuities, attenuation along the route, coupling from nearby nets and noise from power or reference structures. Change one design variable at a time when investigating a failure. Otherwise an apparent improvement may be difficult to reproduce or may conceal a different trade-off.
Make bench measurements comparable
Record probes, fixtures, calibration, reference planes and instrument settings. Include the effect of probe loading and any de-embedding. Compare the same operating mode and channel boundaries in simulation and measurement before deciding that a model or board is wrong.
A review worksheet for one channel
Write down the expected receiver waveform, the worst operating corner, the dominant suspected impairment and one controlled experiment. For example, compare termination options while keeping the driver and route model fixed. Record which margin changed, then verify that the chosen option still satisfies power and timing constraints.
Design review checklist
- Define receiver acceptance criteria.
- Check model provenance and valid operating range.
- Review reflections, loss, coupling and supply effects separately.
- Archive reproducible simulation and measurement setups.