Read claim 1 before you read the headline. Enphase's grant US12456870B2, "Microinverters for PV applications" (issued October 28, 2025; inventor Raghuveer R. Belur), arrives wrapped in a category everyone recognizes — the per-panel microinverter that made Enphase a household name in residential solar. But the patent does not fence off "the microinverter." It fences off a method.
The limitation that matters sits in the control path. The CPC classifications — H02J 3/46 (controlling the sharing of output between parallel sources feeding a network) and G05F 1/67 (regulation to maximize power from a photovoltaic source) — tell you where the inventive weight lands: not on the existence of a small inverter per panel, but on how that inverter decides what to push onto the grid moment to moment. A claim drawn to a control method is narrower, and more defensible, than a claim drawn to a device.
Why does the distinction matter commercially? Because a competitor can build a microinverter without infringing this patent, so long as it does not practice the specific control loop the claims describe. The patent is a fence around a behavior, and behaviors are easier to design around than they look from the press release. Enphase's moat in microinverters is real, but it is built from a thicket of such method claims, not from one patent owning the whole product category.
This is the recurring trap in energy-IP coverage: "Enphase patents its microinverter" is the kind of sentence that conflates a claim with a product. The product is a system Enphase has shipped for over a decade. The patent is a 2025 grant on one method inside it. Both can be true; only one is what the claim language actually covers.
For readers tracking the residential-solar IP race, the takeaway is to read the CPC tags as a map of intent. H02J 3/46 plus G05F 1/67 says: this is about grid-interaction control and maximum-power-point regulation. That is where Enphase is staking new ground in 2025 — the intelligence in the conversion, not the conversion itself.