The compliance test suite behind magnetic wireless charging has just changed versions. Authorized third-party labs have opened Qi 2.3.1 test environments covering both the transmitter and receiver sides of MPP25, in other words the 25W magnetic chargers and phones shipping today. This release breaks from the pattern. From raising power to 15W in 2015, to adding magnetic alignment with Qi2 in 2023, to 25W and the Qi2 25W badge in 2025, the story has always been faster. Qi 2.3.1 adds no power. It changes how products are tested and how clearly the protocol states what is going on.
The most immediate change is that a product has to declare itself first. Qi 2.3.1 introduces an Electronic Self Declaration File (ESDF), a JSON-based list of the features and power modes a device supports. Test software reads that file and selects the applicable tests on its own. That step used to mean an engineer ticking boxes on the test bench, and misconfiguration was routine; leave out one item and the whole test scope drifts. The declaration file makes the process checkable and reusable, and it lines different labs up on the same interpretation.
What follows from the declaration is per-mode testing. Under the specification, the magnetic profile stack has five modes: DPM for backward compatibility with older Qi devices, LPM for low-power charging below roughly 5W, NPM covering the everyday 4.5W to 20W range, HPM for fast charging up to 25W, and CPM for continuous operation across the full range without dropping out, which matters most in automotive and fixed-power installations. Why split it that finely? One hardware design cannot sit at its best operating point at both 4W and 25W. Separate modes are how efficiency, electromagnetic interference, zero voltage switching and control-loop stability each stay in range, so the tests cover only the modes a product actually claims.
The protocol layer gained two pieces. The first is SGC, or System Gain Change: when a transmitter switches its resonant capacitor configuration mid-charge, it now signals the receiver over FSK, and the receiver re-tunes its operating point instead of pressing on with stale parameters. The second is MSN, or Mode Selection Notification: when conditions change and the current power mode is no longer supported, the transmitter says so, and the receiver asks for a mode that still works. Both used to sit in a gray zone of assuming the other side would figure it out. Now they are explicit message exchanges.
One more revision sounds technical but lands somewhere very concrete: under MPP, the time before the next digital ping after an End Power Transfer is now capped at 400 milliseconds. In the older BPP/EPP rules that wait was generous. After a charge-complete EPT a transmitter could go a good fifteen minutes before pinging again, an over-temperature EPT required a minimum of five minutes, and several codes suppressed pinging altogether until the user physically lifted the phone. The release notes point to ASK decoding, where some packets could be read as an EPT code and produce a longer pause than intended. The new timing bounds that gap. In plain terms, those odd moments where the phone sits right there and stops charging for a quarter of an hour should get rarer.
The test suite itself grew in two directions. Negative-case coverage carries more weight: beyond confirming that a product behaves when messages and conditions are valid, the tests check whether it can reject, ignore or recover from an unsupported protocol version, unexpected content, or a transition at a boundary. Interoperability failures almost always surface in those corners, and since transmitter and receiver are developed independently by different companies, neither side can cover for the other. The other direction trims work for MPP15 by dropping the dPLoss calibration. MPP25 keeps it: above 15W, foreign object detection leans on an accurate power-loss figure.
One easily overlooked detail. The Wireless Power Consortium's certified product database lists brand, model, power profile (BPP, EPP, MPP, MPP25) and the certified load power. As of early October 2026 it holds no Xiaomi, Honor or Huawei phone certified since 2023, even though all three advertise wireless charging; a good number of Samsung Galaxy models are registered at just 4.4W to 5W under the baseline profile; Apple's iPhone 15 family is listed at 15W. The 25W on a marketing page and the load power in the database were never the same number. A missing entry does not prove a phone lacks wireless charging, but checking for the Qi or Qi2 mark and a traceable certification ID remains the least painful way to judge a purchase.
For anyone building chargers, the practical meaning is straightforward. Testing has shifted from one exam for everyone to an exam matched to what a product claims, which strips a pile of irrelevant items out of a project and saves some development time and lab cost. The trade-off is that the capability declaration has to hold up: claim HPM and the full 25W test set applies, and any gap between the declared tier and measured behavior shows up at interoperability testing. On the magnetic side, built-in magnet arrays (Pixel 10 and 11 except the 10a, and the Motorola Edge 70 Max) and Qi2 Ready through a magnetic case (Galaxy S25 and later) are two different paths, and tighter certification language makes it easier to say exactly which phones an accessory covers. Power figures will keep climbing, but getting the rules straight first means fewer arguments later.
This article is originally published by Shanwu. Please credit the source when reposting.
Copyright © 2026 Shanwu. All Rights Reserved.