Enphase Envoy No Internet vs No Device Comm: Fault Tree

Why this matters

An Enphase Envoy / IQ Gateway sits between two completely separate communication links: power-line communication (PLC) down to the microinverters, and an internet uplink (Wi-Fi, Ethernet, or cellular) up to the Enphase cloud. A "no data" alarm can mean either link is down, and the two have opposite fixes. If the uplink is down but PLC is fine, the system is still producing and only reporting is interrupted, so the remedy is networking and the array needs no field work. If PLC is down, the gateway cannot see the microinverters and production data is genuinely missing locally, which is an on-site comms or wiring problem. Confusing the two sends a tech onto a roof to chase a router problem, or sends a customer to reboot a router when a phase coupler died. The gateway's own LEDs and local interface tell you which link failed in under a minute.

Symptom presentation

The Enphase app or installer portal shows the site offline or reporting gapped data. In a no-internet case, the array's energy still accumulates and back-fills to the cloud once the uplink returns, the gateway's communication-to-Enphase LED is off or red, and the local gateway interface (accessed on the LAN) shows all microinverters reporting. In a no-device-comm case, the network/cloud LED is fine but the device-communication LED is off or red, the local interface shows missing or partial microinverter counts, and no new production data accumulates because the gateway cannot hear the units. Both surface as "offline" in the cloud, which is why the LEDs and local view matter.

Quick checks

Read the gateway LEDs: the network/cloud (uplink) indicator and the device-communication (PLC) indicator are separate, and on current IQ Gateways each has a distinct solid-green, blinking, or red state documented in the manual. Access the gateway locally over the LAN or its AP to see the live microinverter count and last-report times without depending on the cloud; this single step splits the two failure modes faster than any other check. Confirm the uplink medium: if Wi-Fi, check signal strength (RSSI), whether the SSID or password recently changed, and whether the router was replaced; if Ethernet, check the cable, the switch port, and DHCP; if cellular, check the modem signal bars and that the data plan is active. For PLC, note how many microinverters report versus the installed count and whether the missing ones cluster on one branch circuit or one phase leg, since clustering versus scattering is the key that separates a phase-coupling or branch problem from individual-unit faults. Verify the gateway is powered, has completed boot, and is not mid-reboot or mid-firmware-update, any of which briefly shows offline normally.

Isolation tree

Branch one, split the two links at the gateway. Does the local interface (on the LAN, bypassing the cloud) show all microinverters reporting? If yes, PLC is healthy and the fault is the internet uplink only. Go to the uplink branch. If the local interface shows missing units, PLC is the problem regardless of internet state. Go to the PLC branch. Branch two, uplink. Is it Wi-Fi (check RSSI, SSID change, password change, router reboot), Ethernet (cable, port, switch, DHCP), or cellular (signal, data plan/SIM)? Restore the uplink and confirm the cloud back-fills the buffered data. Branch three, PLC. Do the missing microinverters share a branch circuit or phase leg? Clustered loss points to a dead phase coupler, a tripped branch breaker, a noisy load swamping the carrier, or distance/attenuation. Scattered single losses point to individual unit faults or marginal MC4 connections. Branch four, gateway placement. A gateway electrically far from the array or on the wrong phase leg without a coupler degrades PLC for whole groups.

Confirming diagnosis

Confirm a no-internet fault when the local gateway interface shows full microinverter reporting and current production while the cloud shows offline; restoring the uplink back-fills the gap with no missing energy, proving the array was producing all along and that the only thing lost was visibility. The back-fill behavior is the decisive proof: a true PLC loss leaves a permanent hole in the local energy record because the gateway never heard the units, whereas an uplink loss leaves a hole only in the cloud that fills in once connectivity returns. Confirm a no-device-comm fault when the local interface itself shows missing microinverters and no new production accumulates; the uplink LED being healthy rules out the network. Clustered PLC loss on one branch or phase, recovered by restoring the branch circuit or adding/relocating a phase coupler, confirms a PLC-path cause rather than a unit defect. A scattered loss of single units across unrelated branches, by contrast, points away from the PLC path and toward individual unit or connector faults that each need their own check.

Remediation

For an uplink fault, restore Wi-Fi credentials/signal, the Ethernet path, or the cellular modem, then confirm the gateway reconnects and back-fills buffered production. Relocate the gateway or add an Enphase-approved network bridge if Wi-Fi is chronically weak. For a PLC fault, restore the tripped branch circuit, replace a failed phase coupler, add a coupler on a multi-leg service, filter a noisy load, or relocate the gateway closer to the array in the load center. Re-terminate any high-impedance MC4 on a scattered single-unit loss, and RMA only a microinverter that under-produces during the day with a logged power-stage fault. Confirm the full installed microinverter count reports locally and to the cloud before closing.

References

• Enphase IQ Gateway (Envoy) installation and operation manual, LED status, local interface, and PLC guidance
• Enphase network connectivity guide, Wi-Fi, Ethernet, and cellular uplink configuration
• Enphase phase coupler installation guide for split-phase and three-phase services
• UL 1741, microinverter and gateway requirements
• NEC 2023 Article 690.4 and 690.6, AC module and microinverter circuit requirements