ER vs LR

LR and ER are adjacent reach classes in the IEEE single-mode Ethernet hierarchy. LR covers 10 kilometers. ER covers 30 to 40 kilometers. The fourfold distance gap is the smallest between any two adjacent single-mode reach classes, and that proximity makes the choice between them less obvious than the gap between DR and ER or SR and LR.

Both interfaces use duplex LC connectors, both multiplex multiple wavelengths onto a single fiber pair, and both serve the metro and regional segments of the network. The differences lie in the optical power budget, the wavelength window, the transmitter and receiver design, and whether the link can be amplified. LR is engineered for a 10-kilometer span with modest launch power and a receiver sensitivity that matches the loss of that span. ER extends the reach to 30 or 40 kilometers by launching more power, using a more sensitive receiver, and in the case of the 1550 nm variants, positioning the link for erbium-doped fiber amplifier insertion.

The choice between LR and ER is not a matter of one being better than the other. It is a matter of how much margin the link needs, what the fiber plant looks like, and whether the operator expects the reach requirement to grow over the life of the installation.

1. What LR Covers

LR is defined for 10 kilometers over standard single-mode fiber. The 100GBASE-LR1 interface uses a single 1310 nm wavelength with 100G PAM4 modulation over a duplex LC pair. The 100GBASE-LR4 and 400GBASE-LR4 interfaces use four LAN-WDM wavelengths around 1310 nm, each carrying 25G NRZ or 100G PAM4, multiplexed onto a single fiber pair. The 800GBASE-LR4 interface uses four wavelengths of 200G PAM4.

LR operates in the 1310 nm window, where chromatic dispersion in standard single-mode fiber is near zero. This eliminates the need for dispersion compensation and simplifies link engineering. The trade-off is higher fiber attenuation than the 1550 nm window, but over 10 kilometers the additional loss is manageable.

1.1 LR Characteristics

• Reach: 10 kilometers.

• Wavelength: 1310 nm (LR1) or LAN-WDM around 1310 nm (LR4).

• Connector: Duplex LC.

• Fiber count per link: 2 fibers.

• Dispersion: Near zero at 1310 nm.

• Amplification: Not applicable at 1310 nm.

• Typical power budget: 12 to 14 dB.

2. What ER Covers

ER is defined for 30 to 40 kilometers over standard single-mode fiber. The 100GBASE-ER4 interface uses four LAN-WDM wavelengths around 1310 nm, each carrying 25G NRZ, multiplexed onto a single fiber pair. The 400GBASE-ER8 interface uses eight wavelengths, each carrying 50G PAM4. The 800GBASE-ER8 interface uses eight wavelengths of 100G PAM4.

Some ER variants use wavelengths in the 1550 nm window. The 1550 nm window has the lowest fiber attenuation of any wavelength band used in optical communications, and it falls within the gain band of erbium-doped fiber amplifiers. When an ER link exceeds the reach of a single span, an EDFA can be inserted to boost the signal without electrical regeneration. This amplification capability is the structural difference between ER and LR.

2.1 ER Characteristics

• Reach: 30 to 40 kilometers.

• Wavelength: LAN-WDM around 1310 nm or 1550 nm.

• Connector: Duplex LC.

• Fiber count per link: 2 fibers.

• Dispersion: Low at 1310 nm, moderate at 1550 nm.

• Amplification: EDFA possible on 1550 nm variants.

• Typical power budget: 18 to 25 dB.

3. The Power Budget Gap

The most consequential difference between LR and ER is the optical power budget. A 10-kilometer LR link at 1310 nm consumes approximately 3.5 dB of fiber loss. A 40-kilometer ER link at 1310 nm consumes approximately 14 dB. If the ER variant uses 1550 nm, the fiber loss drops to approximately 8 dB over the same distance, but the dispersion penalty increases.

The additional budget in ER comes from three sources: higher transmitter output power, a more sensitive receiver, and a larger margin allocation for connector loss, splice loss, and aging. The transmitter in an ER module launches 2 to 4 dB more power than an LR module. The receiver is typically 3 to 6 dB more sensitive. Together, these differences account for the 6 to 11 dB of additional budget that ER provides.

4. Wavelength Window and Dispersion

LR operates at 1310 nm, where chromatic dispersion in standard single-mode fiber is near zero. This is a structural advantage for link engineering: the signal does not spread as it travels, and no dispersion compensation is needed. The trade-off is higher fiber attenuation, approximately 0.35 dB/km, compared to 0.2 dB/km at 1550 nm.

ER is defined at 1310 nm in most IEEE specifications, but 1550 nm variants exist for longer spans and amplifier compatibility. At 1550 nm, the dispersion is approximately 17 ps/nm/km, which causes pulse spreading over long distances. Dispersion compensation modules or careful link design are required to manage this. The benefit is the lower fiber attenuation and the ability to insert an EDFA when the span exceeds the reach of a single transmitter.

The choice between a 1310 nm ER and a 1550 nm ER depends on the link length and the fiber plant. A 30-kilometer link at 1310 nm is feasible with a standard ER module. A 40-kilometer link at 1310 nm pushes the power budget harder. At 1550 nm, the same 40-kilometer link has more margin, but the dispersion penalty must be managed.

5. Transmitter and Receiver Design

LR and ER use the same type of laser—EML or directly modulated laser—but the ER laser is driven harder to produce more output power. The higher drive current increases power consumption and thermal load. The ER receiver uses a transimpedance amplifier with higher gain and a photodiode with better responsivity to detect the weaker signal that arrives after 40 kilometers of fiber.

The multiplexer in multi-lane LR and ER modules is a LAN-WDM device in both cases. The difference is the insertion loss and the isolation requirements. ER modules often have tighter specifications because the signal-to-noise ratio at the receiver is lower and the margin for crosstalk between wavelengths is smaller.

6. Power Consumption and Thermal

ER modules consume 2 to 5 watts more than LR modules of the same data rate. The additional power comes from the higher laser drive current, the more sensitive receiver, and the tighter wavelength control. In a switch chassis with 32 or 64 ports, this power difference affects the thermal design and the cooling capacity required.

7. Fiber Plant and Connector

LR and ER both use duplex LC connectors and a single fiber pair per link. The fiber plant requirements are identical: single-mode fiber, LC patch panels, LC patch cords. There is no difference in the physical cabling infrastructure between the two.

The difference is in the optical engineering. An ER link at 40 kilometers requires careful calculation of the optical budget, including every connector, splice, and patch panel in the path. An LR link at 10 kilometers has enough margin that minor variations in connector loss do not affect link performance. In a campus or metro environment where the fiber plant is already installed, the decision between LR and ER depends on the measured loss of the actual path.

8. Application Scenarios

LR is used for metro DCI within a city, campus backbone connections between buildings, and carrier access links up to 10 kilometers. It is the standard interface for leaf-to-spine connections that exceed the reach of FR but stay within a single metropolitan area.

ER is used for regional DCI between cities, long-haul metro links that exceed 10 kilometers, and carrier transport between a central office and a remote access point. It is also used when the operator wants additional margin on a link that is nominally under 10 kilometers, to accommodate future fiber re-routing or aging.

9. Cost Profile

ER modules cost more than LR modules. The higher laser power, the more sensitive receiver, and the tighter wavelength specifications add cost. The premium is typically 20 to 40 percent depending on the data rate and the vendor.

The fiber plant cost is the same for both, because both use a duplex LC pair. The cost difference is entirely in the module. For a network with many links, the module premium accumulates. For a network where the reach requirement is fixed and known, LR is the lower-cost option if it can cover the distance. ER is the lower-cost option if the alternative is deploying an amplifier or regenerator on an LR link that cannot reach.

10. Standards and Ecosystem

LR and ER are both defined by IEEE 802.3. LR4 is defined in 802.3ba for 100G and 802.3bs for 400G. LR1 is defined in 802.3cu. ER4 is defined in 802.3ba for 100G. ER8 is defined in 802.3bs for 400G and 802.3df for 800G. The two interfaces share the same connector, the same fiber type, and the same LAN-WDM wavelength plan, but they differ in reach and power budget.

11. Comparison Summary

12. Selection Framework

The choice between LR and ER is determined by the measured link distance and the margin required. The following framework organizes the decision:

• Measured distance under 8 kilometers: LR. The 10-kilometer rating provides 20 percent margin, which is sufficient for most installed fiber plants.

• Measured distance between 8 and 10 kilometers: LR or ER. If the fiber plant has low loss and few connectors, LR is sufficient. If the path includes multiple patch panels or the fiber is older, ER provides additional margin.

• Measured distance between 10 and 30 kilometers: ER. LR cannot reach without regeneration or amplification, and neither is practical at 1310 nm.

• Measured distance between 30 and 40 kilometers: ER. The 1310 nm variant may be tight; the 1550 nm variant provides more margin and amplification options.

• Distance beyond 40 kilometers: Coherent optics. ER is not rated for these distances.

Where the link distance is uncertain or expected to grow, ER provides additional margin that can accommodate future fiber re-routing or additional patch panels. Where the distance is fixed and well within 10 kilometers, LR is the lower-cost choice.

13. Common Misconceptions

• "ER is just a higher-power LR." Not entirely. ER uses a higher-power transmitter and a more sensitive receiver, but in the 1550 nm variants it also uses a different wavelength window that supports amplification. The 1550 nm ER is architecturally different from LR, not just a power upgrade.

• "LR and ER use different connectors." False. Both use duplex LC connectors. The physical fiber plant is identical for both interfaces.

• "ER always costs more per gigabit." Not necessarily. ER modules cost more per port, but if the alternative is deploying an amplifier or regenerator on an LR link that cannot reach, ER is the lower-cost option. The comparison depends on the link distance and the availability of regeneration equipment.

• "LR can be amplified to reach ER distances." False. LR operates at 1310 nm, outside the gain band of erbium-doped fiber amplifiers. There is no practical amplification option for LR at 1310 nm. Reaching beyond 10 kilometers with LR requires electrical regeneration, which adds cost, latency, and complexity.

• "ER replaces LR." False. LR is the standard interface for 10-kilometer links and is deployed in far greater volume than ER. ER is reserved for links that exceed 10 kilometers or require additional margin. The two interfaces coexist in metro and regional networks.

14. Summary

LR and ER are adjacent reach classes in the single-mode Ethernet hierarchy. LR reaches 10 kilometers with a moderate power budget and 1310 nm wavelengths. ER reaches 30 to 40 kilometers with a higher power budget, a more sensitive receiver, and optionally 1550 nm wavelengths that support erbium-doped fiber amplification.

Both use duplex LC connectors and a single fiber pair per link. The fiber plant is identical. The difference is in the module: ER launches more power, detects weaker signals, and in the 1550 nm variants, positions the link for amplification when the span exceeds what a single transmitter can reach.

The choice between LR and ER is determined by the measured link distance and the margin required. LR covers the 10-kilometer segment that includes most metro DCI and campus backbone links. ER covers the 10 to 40-kilometer segment that includes regional DCI and long-haul metro transport. Where the distance is uncertain or expected to grow, ER provides the margin. Where the distance is fixed and within 10 kilometers, LR is the lower-cost and lower-power choice.

LR and ER do not compete for the same link. They occupy adjacent segments of the metro and regional transport hierarchy, and the choice between them is determined by the physical distance between the endpoints.

15. Q&A