PlatformsFaststream SiliconFaststream RadioFaststream VisionConnected EdgeFaststream SecureMobility & Rail
ProductsSemiconductor IPWireless & RANEdge & GatewaysTracking & IdentificationSoftware & FrameworksConnected Systems
Technology5G protocol stackWireless and RF architectureBaseband and low PHYForward error correctionControl and data planeHigh-speed interfacesFirmware and bootSilicon root of trustSoftware-defined vehicleAutomotive OTAFunctional safety
AIAI Engineering ServicesEdge AI & Embedded MLComputer Vision EngineeringSensor Fusion & PerceptionAI Silicon & AccelerationMLOps for DevicesAI Visual InspectionPredictive MaintenanceDriver MonitoringVideo Analytics & Safety
SolutionsSemiconductorIndustrial AIConnected ProductsAsset TrackingBluetooth AoA RTLSWearable TrackingAutomotive & MobilitySmart InfrastructureSecure IdentityWireless & SatellitePrivate 5GSmart WashroomsFuel ManagementSmart BuildingsWorker SafetyEnergy MonitoringSmart AgricultureSmart CityAutonomous PlatformsAssembly AutomationLiDAR Rail SafetyHardware Wallet
IndustriesSemiconductorTelecommunicationsIndustrial & ManufacturingAutomotive & MobilityTransportation & RailAerospace & DefenceHealthcare & MedicalEnergy & UtilitiesOil & GasRetailConsumer ElectronicsMedia & EntertainmentSmart Infrastructure & IoT
ServicesSystem Integration overviewASIC & SoC DesignRTL to GDSIIVerification methodologyDFT and silicon testLow-power designMixed-signal integrationDesign enablementFPGA DesignFPGA-to-ASIC ConversionAnalog, Mixed-Signal & RFHardware & High-Speed PCBEmbedded SoftwareCloud, OTA & Device ManagementManufacturing TransitionHow we engage
CompanyAbout FaststreamEngineering ExcellenceLeadership & OrganisationHow We EngageQuality & ComplianceStandards & EcosystemPartners & EcosystemTrust CentreLocations & DeliveryNewsroom & MediaCareersCase StudiesKnowledge CenterWhite PapersGlossaryNewsletterResources & Support
ContactStart a projectHow we engage
Talk to us
ENGINEERING INSIGHT

Signal integrity, and why the trouble starts at the pins

Inside the die the data is clean; the trouble begins where it leaves the chip. A multi-gigabit link has to cross a package, board and connector that attenuate, reflect and add jitter until the received eye is almost shut — and reopening it over that real channel, not an ideal one, is what signal integrity is about.

ShareLinkedInXEmail
The channel closes the eye; equalisation reopens it.
AN IDEAL CHANNELA REAL CHANNELThe signal pathShort, controlled, on-diePackage, board, connector, cableHigh frequenciesPass cleanlyAttenuated, the eye closesDiscontinuitiesNone to speak ofReflect at every junctionJitterNegligibleAccumulates across the pathThe received eyeWide openNearly shut without helpWhat recovers itNothing neededEqualisation, tuned to the channelThe datasheet eye is measured on an ideal channel. The link has to close over the board it actually meets.
THE LEVERS

What closes a high-speed link.

Signal-integrity levers
LeverWhat it addressesWhere it acts
Transmit pre-emphasisChannel loss, anticipatedAt the transmitter
Receive equalisationLoss and reflection, undoneAt the receiver
Clock and data recoveryTiming recovered from dataAt the receiver
Jitter budgetingTotal jitter across the pathWhole link, not per stage
Channel designLoss, impedance, discontinuitiesPackage, board, connector
Link signoffBit-error-rate marginOver the real channel, all conditions
THE DISTINCTION

The eye is closed by the channel, not the chip.

A serialiser sends data at multi-gigabit rates, and inside the die that data is clean because the interconnect is short and controlled. The moment the signal reaches the pins it enters a channel the chip designer often does not own — a package, a board, a connector, sometimes a cable — and that channel attenuates the high-frequency content the fast edges depend on, reflects at every impedance discontinuity, and adds jitter. At these rates the received signal is barely open; the receiver has to reopen it before it can recover a bit at all.

Reopening it is the work of equalisation. The transmitter pre-distorts the signal, boosting the high frequencies the channel will attenuate, and the receiver applies its own equalisation to undo the remaining loss and reflection, restoring an eye wide enough to sample reliably. Tuning both to a specific channel — enough boost to open the eye, not so much that it amplifies noise — is an iterative, channel-specific task, which is why a link that closes over an ideal channel can still fail over the real board it actually meets.

Two things make it unforgiving. Jitter is contributed by the transmitter, the channel and the recovery loop together, so the link only closes if the whole budget is accounted as one — a link that looks fine stage by stage can fail as a whole. And when the channel is outside the silicon team's control, the margin has to be robust to a board and connector that vary, which raises the bar over a fixed channel. Signoff, therefore, is bit-error-rate margin proven over the real channel across conditions, not a datasheet eye measured on an ideal one.

IN PRACTICE

Closing a link over a real channel.

COMMON QUESTIONS

What engineers ask before they call.

01

Why does a high-speed link degrade outside the chip?

Because inside the die the interconnect is short and controlled, so the signal stays clean, but the moment it leaves the pins it crosses a package, board, connector and possibly a cable. Those attenuate the high-frequency content that fast edges depend on, reflect at every impedance discontinuity, and add jitter. At multi-gigabit rates this closes the received eye until it is nearly shut, so the receiver has to equalise the channel back out before it can recover a bit — which is where the real engineering lives.

02

What is equalisation doing?

It compensates for what the channel did to the signal. The transmitter pre-distorts the data, boosting the high frequencies the channel will attenuate, and the receiver applies equalisation to undo the remaining loss and reflection, reopening the data eye enough to sample reliably. Tuning both to a specific channel — enough correction to open the eye without amplifying noise — is the core, iterative work of bringing up a high-speed link, and it is why an ideal-channel result does not guarantee a real-channel one.

03

Why budget jitter across the whole link?

Because jitter accumulates from every stage — the transmitter, the channel and the clock-recovery loop — and the link only closes if the total stays within budget. Accounting for each contributor in isolation misses how they combine, so a link that looks acceptable stage by stage can fail as a whole. Treating jitter as one shared budget across the path is what makes the signoff honest and the link reliable in the field.

FOUND THIS USEFUL?

Pass it on.

Written for engineers. Share it with one.

ShareLinkedInXEmail
KEEP READING

Related work.

BUILD WITH FASTSTREAM

Bring us the difficult part.

Tell us the specification, the constraint and the deadline. Programmes that cross silicon, radio, embedded and AI are where Faststream is strongest.