Standard Insight｜From IEC 61851-23:2023 to CRA: How DC Charger Residual-Current Protection Evolves and How to Build Compliance Evidence Chains

In October 2026, TÜV Rheinland East China Advanced-Technology Conference for New Energy & Novel Power Systems was held in Yancheng, Jiangsu Province. Centered on the theme "Pursuing Green Sustainability · Forging a New Energy Future", the event gathered more than one thousand representatives from government agencies, research institutes, and industrial chains covering photovoltaics, energy storage, and charging-and-swapping infrastructure. Questions raised by charging-and-swapping enterprises converged sharply on three points: how standards evolve, how to ensure equipment safety, and how to achieve compliance for overseas market access.

Among all sessions, the Advanced Charging-Swapping Technology Seminar drew intensive industry attention. Morning sessions focused on standard iteration and ultra-fast-charging technologies, while afternoon sessions covered electric energy metering, cybersecurity and global market-access requirements. Most attendees were technical and certification managers from OEMs of charging equipment. The policy context for this conference can be condensed into a timeline:

Within two years, standards have advanced from "text-based recommendations" to "design-binding constraints".

The opening session kicked off with an interpretation of IEC 61851-23:2023. Instead of reciting clauses one by one, the speaker directly addressed design constraints: structural modifications, retesting requirements, and limitations on component selection. What the standard specifies is no longer the priority; what products need to adjust is the core concern.

Across the full-day agenda, one consistent thread runs through all topics: the burden of compliance-related evidence-gathering is shifting from complete equipment OEMs onto component suppliers. At the end of this chain lie residual-current sensors.

I. Standard Interpretation Shifts from Clauses toward Design Constraints

IEC 61851-23 is the international standard for DC charging equipment. Its 2023 revision introduces multiple amendments for safety architecture, fault protection and hardware testing. The core change concerns residual-current protection: it has been upgraded from an optional configuration recommendation to a mandatory clause. Beyond requiring protection-action capability, the standard mandates verifiability of such protective responses.

Directly-quotable text comes from the scope clause of IEC 62955:2018: an RDC-DD shall disconnect or initiate disconnection of power supply to an electric vehicle upon detecting smooth DC residual current equal to or greater than 6 mA. Note 2 within this standard explains the origin of the 6 mA threshold: this value is selected to avoid impairing correct operation of upstream Type A / Type F Residual-Current Devices (RCDs). Namely, 6 mA represents the lower detection limit for preventing upstream AC-side RCDs from being "blinded" by DC components, rather than a human-body-protection threshold (human-body protection remains at the 30 mA level). Regarding operating time, per IEC 62955 and national implementation practices: protection shall activate within 10 s for 6 mA DC residual current, and within 0.3 s for 60 mA DC residual current.

For IEC 61851-23 itself, the 2014 edition imposed no mandatory DC residual-current-protection requirements for DC chargers. For non-isolated DC chargers, it only required at minimum a Type A RCD fitted on the AC side and compatibility with that RCD. Following structural reorganization in the 2023 edition (electrical-safety requirements relocated to Clauses 8 and 12), mandatory requirements apply to DC chargers: 6 mA smooth DC shall be covered on the AC side (via RDC-DD or equivalent functionality), and the DC output side shall be safeguarded by Insulation-Monitoring-Device (IMD). It should be noted that the original scope of IEC 62955 Ed.1 covers Mode 3 AC charging stations; DC chargers are treated as equivalent applications implemented by reference. (Clause statements shall be subject to the officially-published standard text.)

Comparing legacy and new-generation chargers against old and new standards yields one key takeaway: DC charging stations must be equipped with DC residual-current protection.

This revision directly reshapes the BOM (Bill-of-Materials) for DC chargers: residual-current sensors have changed from optional parts to mandatory components.

Two categories of performance indicators apply. First, discrimination capability for different residual-current waveforms: power-frequency AC, pulsating DC and smooth DC call for distinct detection mechanisms. Second, self-test capability of sensors themselves, including power-on self-test and periodic runtime self-test.

Where shall residual-current-protection devices be installed? Two distinct locations apply, not a single point. The 6 mA smooth-DC requirement targets the AC-side incoming circuit. Clause 11.3.2 of GB/T 18487.1-2023 mandates residual-current-protection functionality (covering smooth-DC current of 6 mA and above) on the AC-side main circuit of DC power-supply equipment, to be fulfilled by RDC-DD. The underlying principle: AC-side Type A RCDs can only handle power-frequency AC and pulsating DC. Smooth-DC components can desensitize Type A RCDs; this independent sensing component fills that functional gap. Protection for the DC output side is realized via IMD, with insulation-resistance criteria (e.g., 100 Ω/V) instead of the 6 mA threshold. These two protection layers serve separate purposes and cannot substitute for one another.

The 6 mA value originates from IEC 62955 (RDC-DD; corresponding domestic standard GB/T 40820). It defines the lower detection limit for smooth-DC residual current and the threshold for preventing upstream RCD blinding — it is not a personal-safety threshold. The major technical challenge for this metric lies in magnetic-circuit design.

Conventional Type A residual-current protection relies on current transformers that sense magnetic-flux variation. When DC components exist within circuits, DC magnetic flux biases the magnetic core, shifting its operating point toward saturation and lowering permeability. Induced signals for AC components become distorted. Higher residual-current aggravates distortion, ultimately resulting in missed faults: protection fails to trip while faults persist.

The performance limitation of Type A RCDs lies not in sensitivity but magnetic-circuit architecture. To detect 6 mA smooth DC, alternative detection principles must be adopted: Type B devices, or dedicated independent DC-detection channels added onto Type A hardware.

A frequently-raised question at the conference concerned not solution selection, but the timing for introducing components into BOMs. The relatively clear conclusion: as early as feasible. If residual-current-detection hardware is finalized late in prototype development, PCB revisions, new mold-opening and recertification will be triggered, directly compressing time-to-market windows.

II. Power Upscaling & Unattended Connection Reshape Residual-Current-Detection Operating Conditions

The combination of megawatt-level MCS (Megawatt-Charging-System) and automatic charging represents the most demanding scenario for residual-current detection, subject to overlapping constraints of power rating, unattended operation and certification requirements, deserving separate analysis.

For context: currently mainstream single-gun power ratings for public DC chargers in China range from 120 kW to 240 kW; liquid-cooled ultra-fast chargers commonly reach 480-600 kW, while megawatt-class MCS pushes the upper limit up to 3.75 MW. Implementing 6 mA smooth-DC detection on AC incoming lines carrying hundreds of amperes requires sensors to maintain resolution across a dynamic range on the order of 1 : 10⁵ — this creates substantial technical challenges.

MCS supports maximum power of 3.75 MW at 1250 V and 3000 A. Power-rating increases drive changes in cable design: larger conductor cross-sections and longer wiring lengths raise stray-to-ground capacitance. Capacitive leakage follows I = C × dU/dt, determined by capacitance magnitude and switching speed. SiC power devices push dU/dt to 10-20 kV/μs, an order-of-magnitude improvement compared with legacy IGBT devices. With both factors elevated, steady-state capacitive leakage can reach ampere levels, containing abundant frequency components above 10 kHz.

Automatic charging coupling modifies conditions on the other side of the system. When charging-gun mating is completed via robotic arms or auto-docking mechanisms, human visual judgment during plug-in / unplug-out is eliminated. Contact integrity, foreign-object intrusion at gun terminals and insulation degradation must all be autonomously verified by equipment.

Within unattended connection scenarios, self-test evolves from an optional feature to a prerequisite condition. This explains why standards explicitly list self-test as an independent requirement: safety-component failures occur silently without alarms and manifest only as loss of protective responsiveness.

Parallel agenda topics covered lightweight design for high-power chargers and the 2PfG standard for liquid-cooled cables. Although seemingly structural subjects, together they reflect unified constraints: charger weight reduction, smaller cable cross-sections and higher power ratings. Available installation space for components shrinks and clearances for insulation are reduced. Sensors must withstand higher dielectric strength within more compact form factors.

III. Overseas-Market-Access Barriers Shift from Product-Level Performance toward System-Level Evidence

Afternoon sessions spanned five threads: metrology, cyber resilience, market access, low-carbon requirements and grid interconnection. Diverse as they appear, they converge on one core insight: competing for overseas markets no longer depends merely on discrete product performance, but on complete sets of systematic compliance evidence.

Metrology presentations were delivered by experts from Zhejiang Institute of Quality Sciences, covering technical requirements for electric-energy metering and progress on domestic type-evaluation programs. Metrology and residual-current protection belong to separate technical domains yet share identical regulatory logic: both form part of type evaluation; failed testing prevents OEMs from obtaining certification.

The EU CRA (Cyber-Resilience Act) entered into force in December 2024; key obligations become applicable starting December 2027. Combined with AFIR (Alternative-Fuels-Infrastructure-Regulation, applicable April 2024, governing communication protocols, plug-and-charge functionality and data security), these two EU regulations shift market-access evaluation from standalone product performance toward systemic compliance.

While requirements under both regulations appear software-focused and seem implementable through code modification, certification demands hardware-level evidence: whether protective functions remain effective under cyber-attacks, interference or mis-operation. Software defects can be patched remotely; hardware capability can only be upgraded via component-model replacement.

IV. Compliance Requirements Flow Downstream: Components Become Starting Points of Evidence Chains

Synthesizing morning and afternoon discussions reveals one clear trend: compliance obligations are moving downstream. Responsibility is no longer borne solely by complete-equipment OEMs but shared across the whole supply chain.

For OEMs to obtain certification, deliverables include test reports and type-evaluation conclusions. These documents are built upon measured raw data from individual components. OEMs cannot substitute component suppliers to validate component performance; pressure inevitably propagates upstream.

Translated to residual-current-sensor requirements, concrete specifications emerge: discriminate DC offset signals at mA-level superimposed on circuits carrying hundreds-ampere input current; bandwidth shall cover frequency components above 10 kHz introduced by SiC devices; zero-point stability shall be maintained across -40 ℃ to 105 ℃. A 20 mV zero-point drift will produce systematic offset at mA measurement ranges: no alarm is triggered; protection performance simply degrades persistently.

These requirements impose mutual trade-offs: wider bandwidth raises noise-suppression difficulty; larger measurement ranges degrade resolution; miniaturized packaging reduces thermal-dissipation and insulation margins. Component-selection complexity stems from these conflicts rather than completeness of datasheet parameters.

V. Opportunities & Risks: Expanding Market Demand Coexists with Low-Quality Competition

For upstream component manufacturers, this trend creates clear market opportunities. Growing demand does not equate to lower technical thresholds.

Three major opportunities stand out. First, rising component consumption driven by higher power ratings: residual-current-detection points are distributed at input sides, DC output sides and gun terminals within megawatt-scale ultra-fast chargers; quantities further multiply for multi-gun simultaneous-charging deployments. Second, earlier design-in windows: components incorporated into BOM at early development stages face higher barriers for replacement. Third, substantial substitution potential: conventional closed-loop solutions rely on imported chips and permalloy magnetic cores, suffering disadvantages in cost and lead time.

Risks are equally tangible and easily underestimated.

Standards remain under continuous revision. Subsequent amendments to IEC 61851-23 and divergent national-adoption timelines mean certifications obtained today may require supplementary testing two years later. Components parameterized exclusively for one standard revision will face repeated modification requirements.

Low-quality solutions amplify information asymmetry. Claiming "DC-measurement capability" has low entry barriers. Reliable 6 mA detection across full operating-temperature ranges represents a far higher technical hurdle. If purchasing teams treat certification pass status as the sole acceptance criterion, solutions that pass prototype testing yet become unstable in mass production may be adopted.

Passing certification does not guarantee long-term field reliability. Zero-point drift and acoustic whistling rarely manifest during prototype validation. Drift accumulates over operating runtime; whistling arises predominantly after mass assembly. Neither phenomenon appears on datasheets yet translate into after-sales costs borne by charger OEMs.

VI. MAGTRON's Technical Approach & Market Positioning

To recap the full technical chain: identify mA-level smooth-DC residual current superimposed within high-current DC circuits — this constitutes MAGTRON's core strength. MAGTRON's residual-current-sensor product line is the RCMU101SM9 series, adopting switched flux-gate architecture. Its technical design addresses the three risk categories outlined above.

At chip level: in-house-developed SoCs replace imported alternatives. Excitation drive, synchronous sampling, signal amplification and switch-level decision-making outputs are integrated on-die, enabling supply-chain autonomy and optimizing cost structures.

At magnetic-material level: novel low-loss magnetic materials replace permalloy. Magnetic-circuit topologies are re-engineered to reduce BOM costs while preserving excitation performance and anti-interference capability.

At packaging level: mechanical-structure optimization shrinks form factors while maintaining dielectric-strength ratings and long-term reliability.

For parameters: the RCMU101SM9 series operates from -40 ℃ to 85 ℃ ambient temperature; maximum storage and surrounding-air temperature reaches 105 ℃. It runs on single +5 V supply with push-pull switch-output signals. Self-test (CHK test pin) and power-on auto-zeroing (≤ 1 s) are supported. Operating thresholds cover AC 30 mA and 6 mA smooth DC, complying with Type A plus RDC-DD (IEC 62955) residual-current-protection requirements. Components are designed for compatibility against IEC 62955, IEC 62752, IEC 61851, GB/T 18487.1 and other standard systems from the R&D phase onward, minimizing PCB-revision cycles for host equipment.

Excitation frequency represents a key differentiator among residual-current-detection schemes. Open-loop architectures commonly use excitation frequencies below 10 kHz, falling within human-audible frequency bands. Vibration of magnetic cores and mechanical structures is amplified into audible acoustic noise. This issue originates from fundamental principle-level frequency conflicts; potting and mechanical adjustments can only attenuate but cannot eliminate noise. When excitation frequency is raised to 300-400 kHz, this problem is fundamentally resolved. This accounts for the phenomenon where certain projects perform well at prototype stage yet suffer high repair rates post mass-production launch.

VII. Conclusion

The industry catchphrase "standards are getting stricter" is imprecise. The core tightening lies not in the number of standard clauses, but in the burden-of-proof framework. Previously, charger OEMs only needed to declare protection-feature availability. Today they must demonstrate that protective functions remain valid under diverse operating conditions, and such demonstration must trace down to component-level evidence.

This represents cost pressure for charger OEMs and market opportunity for component suppliers; schedule planning matters for both parties. To maintain initiative for projects launching in 2027, residual-current-sensor component selection should be finalized at the current development phase.

MAGTRON continuously tracks revisions to domestic and international charging-and-swapping standards. If you are developing high-power DC chargers, megawatt-level ultra-fast-charging equipment or charging-gun terminal modules and require residual-current-sensing solutions, please contact us to obtain selection guides and product samples.

References

1. TÜV Rheinland: Charging-Swapping Sub-Session: Driven by Dual Engines of Technical Breakthroughs and Compliance Foundation-Building to Boost High-Quality Industrial Development, 24 Oct 2026

2. IEC 61851-23:2023, Electric vehicle conductive charging system – Part 23: DC electric vehicle supply equipment

3. IEC 62955:2018, Residual direct current detecting device (RDC-DD) to be used for mode 3 charging of electric vehicles

4. Regulation (EU) 2023/1804, Alternative Fuels Infrastructure Regulation (AFIR), applicable 13 Apr 2024

5. Regulation (EU) 2024/2847, Cyber Resilience Act (CRA), entered into force 10 Dec 2024; major obligations apply from 11 Dec 2027

6. CharIn: Megawatt Charging System (MCS) technical definition: max. 1250 V / 3000 A, power ceiling 3.75 MW

7. MAGTRON (Zhejiang Magtron Intelligent Technology Co., Ltd.), Current-Sensor Selection Handbook, Version V1.0 2026; MAGTRON datasheet RCMU101SM9-EI-KG-D2, V1.0.0