EMC Emission Reports: GB 4824 Sits under a Higher Layer

Why emission compliance keeps bouncing between several documents

When an active medical device goes through EMC testing, the immunity half usually moves along smoothly: the items are discrete, each one corresponds to a form of disturbance, you tick a box as each is completed, and if something fails you know which port to investigate. The emission side is where jobs stall, and what stalls them is usually not the data. It is the basis of the report.

Start by clearing up the layer that most often gets inverted. An EMC test report used for domestic registration is issued under YY 9706.102. An EMC report following the CE route, submitted to overseas customers and notified bodies, is issued under IEC 60601-1-2. Those two are the EMC collateral standards for medical electrical equipment. They do not define a fresh set of emission limits of their own; on the emission side they refer respectively to GB 4824 and CISPR 11 for grouping, classification, limits and measurement methods. In other words, GB 4824 and CISPR 11 are not an alternative basis for submission sitting alongside YY 9706.102. They are the source of the limits and the methods, pulled in by reference from the collateral standard.

Reading that relationship as "GB 4824 for China, CISPR 11 for overseas" leads directly to the wrong kind of test request: commissioning a single stand-alone GB 4824 emission report and taking it to registration, where the formal review stage may well rule that the test basis does not match and require the report to be reissued under the EMC collateral standard for medical electrical equipment, with the immunity portion completed as well. That is not something a new report cover can fix; in most cases it means rescheduling and going back into the chamber.

So this article aims to help you settle three things before samples are shipped: which layer the report is issued under, how group and class are determined, and whether one set of measurement data can be cited separately by a report for domestic submission and a report for export customers. It does not restate standard clauses. It deals with how this lands in a real project, and how far your preparation has to go if the trip is not to be wasted.

Layering the basis: which document issues the report, which one supplies the limits

The layer occupied by YY 9706.102 and IEC 60601-1-2 governs the overall electromagnetic compatibility requirements for medical electrical equipment in its intended use environment. What they provide is framework material: applicable requirements divided by intended use environment, the EMC information that has to be given in the accompanying documents, and where the emission and immunity requirements are each to be taken from. When it comes to emission, they point to the emission standard, and the emission standard specifies how equipment is grouped, how it is classified, what the limit curves look like and which detector is used.

GB 4824 and CISPR 11 are that emission layer, covering industrial, scientific and medical radio-frequency equipment. GB 4824 is the Chinese national standard corresponding to CISPR 11; the two share a technical lineage, and the framework for grouping and classification and the way emission items are organised follow the same logic. That is what makes "measure once, issue two sets of conclusions" technically possible.

Common lineage is not the same as being interchangeable in writing, however. How a particular national standard adopts the corresponding international standard, the degree of correspondence, and whether nationally specific supplementary requirements have been added, must be taken from the foreword and adoption statement of the current valid version. Do not write it from memory or from experience on a previous project. This is a trap we have walked into: a technical file contained an assumed correspondence statement, the reviewer asked for the supporting source, and the schedule stretched out instead. Writing "corresponds to the international standard" is safe; nailing down a specific correspondence statement without a source is digging a hole for yourself.

In the submission dossier the layering has to show up as two statements: the test basis of the report is the EMC collateral standard for medical electrical equipment, and the source of the emission limits and measurement methods is the emission standard. Invert those roles and the classic symptom appears: the basis on the report cover, the citation on the conclusion page and the basis entered on the submission form disagree in three places, and the file comes back for correction at the formal review stage.

Two further confusions are worth flagging. First, the immunity test methods come from a different family of basic electromagnetic compatibility standards, the family commonly associated with immunity test methods, which has nothing to do with emission grouping and classification. Some companies enter an immunity test-method number as the basis for an emission item, and that stands out plainly in formal review. Second, items such as harmonic current and voltage fluctuation and flicker, which relate to the public supply network, sit on the emission side but are based on other parts of that same basic standard family; they do not share the limit system covered here for conducted and radiated emission. Whether a given device needs them, which part specifies them, and through which layer they are referenced in, should be confirmed against the supply arrangement and the target-market requirements, using the current valid version of the standard text and the submission pathway. Do not file them under GB 4824 or CISPR 11 from impression.

There is one more place where wires get crossed: the "group" and "class" in the emission standard describe how radio-frequency energy is produced and where the equipment is intended to be used. They are an entirely separate determination from the regulatory risk class assigned during product registration, and from the applied-part type determination, and none of these substitutes for another. How equipment class and applied-part type are determined in terms of protection against electric shock is discussed in a separate article and is not expanded on here.

Grouping and classification: how the determination should proceed

Every conclusion on the emission side starts from placing the equipment in the correct group and class. Get this step wrong and all the data downstream has to be reinterpreted.

Grouping asks whether the equipment intentionally generates and uses radio-frequency energy internally. The order of questioning is: is there a radio-frequency energy source inside the equipment; is that radio-frequency energy essential to the function of the equipment itself; and is it used for a physical effect within the machine (heating, cutting, imaging excitation) or for communicating externally. Answer those three in order and the group is essentially fixed.

A very common misjudgement here is treating "there is a crystal oscillator, a switch-mode power supply and a wireless module inside" as intentional generation of radio-frequency energy. High-frequency switching in a power supply is a means of converting electrical energy; the radio-frequency content is a by-product, not the functional purpose. The transmitter section of a wireless communication module is normally assessed under a separate system, and its presence does not by itself change the group the device falls into. A genuinely typical intentional source is equipment whose therapeutic or diagnostic function depends on radio-frequency energy in the first place. The determination turns on functional intent, not on whether high-frequency signals exist in the circuit.

Classification looks at the intended place of use and how the equipment connects to the supply. This step often gets pulled off course by sales language: because "we only sell to tertiary hospitals", someone goes straight to the industrial-location class. But the basis is the intended place of use and whether the equipment connects to the public low-voltage supply network feeding domestic premises, not the customer list. Equipment that will be used in homes or community settings, or connected directly to the domestic low-voltage network, has to be considered under the more stringent class. More importantly, this determination has to stay consistent with the intended use and use environment stated in the instructions for use. Reviewers read those two side by side, and the collateral standard layer likewise divides requirements by intended use environment, so the instructions for use, the environment division under the collateral standard and the class determination under the emission standard must all line up.

One further point that gets overlooked is multi-mode equipment. Where a device has standby, normal output, high-power output and charging modes, the grouping and classification determination has to cover all of them, and the test configuration has to be set up around the mode with the higher emission level. Some companies submit only the stable light-load mode to make life easier; the report is obtained, but the configuration description does not match how the product is actually used, which is a liability at the next change or surveillance check.

Conducted and radiated emission: two coupling paths on the same device

The following is engineering analysis rather than a statistical conclusion. It is meant to help judge the direction of remediation; the actual frequency ranges, limits and detector types are governed by the current valid version of the standard text.

Conducted emission travels a conductor path: the mains port, signal ports and some patient-connected ports. In essence, common-mode and differential-mode currents generated inside the equipment propagate along the conductors and are then sampled by the measuring equipment. Radiated emission travels a spatial path: the enclosure, internal wiring, external cables and the earthing system together form an equivalent antenna that radiates energy into space.

From that difference in path, the remediation logic can be derived. Where conducted emission exceeds the limit, look first at whether the filter network at the mains inlet has been weakened by later board revisions, whether the earth return of the common-mode choke and Y capacitors is short enough, whether the input wiring runs alongside high-frequency areas inside the chassis, and whether the heatsink of the power module has created an extra common-mode path. Where radiated emission exceeds the limit, look first at whether the cable shield is terminated as a long thin pigtail, whether enclosure seams and ventilation openings cut across the surface current, and whether a split ground plane has lengthened the return path of internal high-speed signals. These are the recurring causes in emission remediation, and they make engineering sense: when cable length approaches the order of the operating wavelength, the cable itself is a reasonably efficient radiator, and the quality of the shield termination decides whether it acts as a shield or as an antenna.

There is a practical corollary that maps onto the two paths: the same excursion may be judged differently under different classes, because the limit curves vary with the intended place of use in the first place. Getting the classification wrong is therefore not something a new report cover can fix; it changes the conformity conclusion directly. That is why this article puts grouping and classification ahead of remediation.

Can one set of measurements be cited by two reports

Ask the question correctly first. It is not "should we test to GB 4824 or to CISPR 11" — both are cited as sources of limits within their respective reports, so there is no either/or. The real question is whether one set of raw measurement data can be cited separately by a report under YY 9706.102 and a report under IEC 60601-1-2. The answer is that it is technically feasible, but it depends on whether the single measurement campaign covered, in one pass, the conditions needed for both conclusions, and that has to be settled at the ordering stage. The table below is what we actually confirm item by item when planning emission work.

The rows to watch in that table are the first two and the last. The first two say the same thing: the test basis and the source of the limits belong to two different layers, cannot be written interchangeably, and cannot be reduced to one. The "use of the result" row marks the boundary of the conclusion: raw measurement data can be shared, but a conformity conclusion is a judgement made against a specific standard, a specific edition and a specific class. Lifting a conclusion straight into another system is a fairly serious defect in a technical file.

One more reminder: emission is only part of what the collateral standard requires, and the immunity items have to be covered under the same layer of basis. Taking a single set of emission data to a submission produces an incomplete report even if the layering is written correctly. The right approach is to state on the test request that conclusions are needed both for domestic submission and for export customers, so that the laboratory arranges configuration, modes and recording in one test plan. Adding measurements afterwards usually costs far more than the extra communication time at the start.

For how items are scheduled and priced, start with the scope listed on the EMC testing service page, then prepare your documentation against the testing requirements.

There is more to prepare before submission than most people expect

More than half the waiting time on an emission test goes on the condition of the prototype and on missing documentation. The checklist below is organised by what happens when an item is missing.

Three more traps that come up frequently in real projects.

First, the prototype is not in production condition. Hand-soldered flying leads, temporary earth wires and development debug cables left on the machine are all genuine radiators in an emission test; the result represents neither the production unit nor a usable basis for remediation. The prototype should be as close to production condition as possible, and any real differences should be declared at submission.

Second, the configuration gets changed during testing. An engineer tunes while measuring in the chamber, the result improves, but the configuration described in the report no longer matches what is delivered. Every configuration change during an emission test should be recorded; that is both a requirement of the standard and the key to reproducing an issue later.

Third, no remediation window is left. Emission remediation often needs board changes, added filtering and rerouted cables, and the time for one iteration is not the laboratory's to decide. Booking registration testing into a very short window before the submission deadline leaves no room to manoeuvre if a limit is exceeded. For how scheduling works see the testing process, and common questions are collected on the FAQ page.

One sentence that has to be said about accreditation and scope

SUNGO Lab (Shanghai Shage Medical Technology Co., Ltd.) is accredited by CNAS, CMA and IAS (USA), operates laboratories in Shanghai and Hefei, and can undertake emission and immunity testing on active medical devices. One point must be explicit: an accreditation mark only demonstrates that the laboratory holds the corresponding technical competence within its accredited scope; it does not constitute a commitment as to the outcome of market access in any target market. Whether a report is accepted in a given market depends on the local regulatory pathway, the specific requirements of the receiving authority and the completeness of the submission dossier, which is a different question from which accreditation marks the laboratory holds. When planning an export route, assess "technical validity of the report" and "regulatory pathway for market access" separately rather than treating them as one thing.

How to get started when you need support on emission items

If your device is in late-stage development or preparing for registration testing, the recommended sequence on the emission side is: first confirm which layer of basis the report is issued under and whether both the domestic and export routes have to be covered; then determine group and class; then fix the submission configuration and the operating-mode list; and only then book the chamber. That order avoids most of the rework. Before formal submission, SUNGO Lab can go through the layering of the basis and the grouping and classification with you, confirm the test configuration, and issue separate reports for domestic submission and for export customers from the same set of measurements, reducing repeat visits to the chamber. Our wider scope is listed on the services page.

To discuss a project, scheduling or pricing, call +86 132 4819 8029, or request a quote with your product documentation and timing requirements, and we will arrange for an EMC engineer to take it forward.