EN 300 220-2 V3.3.1: Does Your Existing Product Really Need to Be Retested?

Thousands of Short Range Devices already on the European market were tested against earlier versions of the EN 300 220 series.

A new version of a harmonised standard therefore creates an immediate concern for manufacturers:

Do we have to test everything again?

With ETSI EN 300 220-2 V3.3.1, the answer is not as simple as yes or no. The new standard represents a substantial development of the previous requirements. Some tests have changed mainly in structure or methodology. Some requirements have been separated into dedicated assessments. And, particularly on the receiver side, genuinely new requirements can become applicable.

But that does not automatically make an existing test report worthless.

Our detailed investigation of the changes, combined with assessment of existing EN 300 220 test reports, shows that the most efficient approach is often:

Reassess first. Retest where necessary.

EN 300 220-2 V3.3.1 reorganizes a significant part of the technical framework for non-specific Short Range Devices operating from 25 MHz to 1 GHz.

The changes include a stronger Operating Channel concept, separation of transmitter frequency stability from occupied bandwidth, revised treatment of out-of-band emissions, a dedicated spectrum mask at permitted frequency-band edges, changes to receiver requirements and additional provisions for equipment using polite spectrum access.

The new standard also defines the information needed to execute its conformance tests in considerably more detail. Relevant EUT properties include the Permitted Frequency Band, nominal operating frequencies, Operating Channel edges, test signals, disregard time, environmental category, voltage range, spectrum-access mechanism and, where relevant, FHSS parameters.

This matters when looking at an old test report.

The question is no longer simply: "Was this product tested to EN 300 220?"

It becomes: "Does the evidence in that report demonstrate what V3.3.1 now requires?"

One of the most important findings from our investigation is that a changed requirement does not automatically mean a new measurement is necessary.

Take transmitter frequency stability.

V3.3.1 treats frequency stability as a separate transmitter requirement. An older report may not contain a test carrying that exact title. But it may contain extensive occupied-bandwidth or frequency-edge measurements over temperature and supply voltage.

Those measurements could potentially provide useful evidence. Whether they are sufficient requires assessment.

The same principle can apply to Effective Radiated Power, occupied bandwidth, transient power and certain emissions measurements. If the original report contains the right measurement conditions, data and sufficient compliance margin, some evidence may remain technically useful.

In practical legacy-report assessments, we have seen existing ERP measurements with substantial compliance margin and extensive environmental occupied-bandwidth data that were strong candidates for reuse following reassessment rather than automatic repetition.

That can make a considerable difference to the scope of a new laboratory programme.

Not every requirement can be treated that way.

Our detailed GAP investigation identified several areas where manufacturers should be particularly careful.

One is the dedicated transmitter spectrum mask at the permitted frequency-band edges.

Older reports may already contain measurements around band edges as part of previous out-of-band testing. But having an old plot around the same frequencies does not automatically mean that the new dedicated requirement and its conformance procedure have been demonstrated.

Another example is transmitter behaviour under low-voltage conditions. The newer procedure is more specifically defined. An old report containing a measurement at the minimum declared voltage may therefore not necessarily demonstrate the complete new procedure.

And even where the numerical limit appears familiar, the measurement methodology can matter.

This is exactly why a GAP analysis should compare more than two limit tables.

For products containing a receiver, the assessment can become considerably more important.

Our investigation identified important changes affecting receiver requirements, including receiver dynamic range, adjacent channel selectivity, spurious response rejection and intermodulation rejection.

Receiver intermodulation rejection is a particularly clear example of why manufacturers should not assume all V3.3.1 changes can be covered by paperwork.

Where applicable, V3.3.1 introduces an explicit receiver intermodulation rejection requirement with a dedicated conformance procedure. If an equivalent requirement was not previously applicable, there may simply be no legacy measurement to reuse.

That is fundamentally different from a requirement where only the test setup or reporting method has changed.

The reverse can also happen.

Short Term Behaviour illustrates this well.

The requirement has been restructured from the previous Duty Cycle Template-oriented approach into a dedicated Short Term Behaviour requirement. Applicability is now connected to the relevant permitted-frequency-band or National Radio Interface conditions, while the new standard directly expresses the Ton and Toff behaviour.

That looks like a significant structural change.

However, our detailed comparison found that the fundamental Ton/Toff principles and measurement concept remain substantially equivalent. The appropriate action may therefore be reassessment of existing evidence rather than automatic retesting.

This is precisely why judging the impact from the number of changed pages, or simply from a new clause number, can be misleading.

A good legacy test report can contain far more information than the final PASS result.

It may contain:

• measurements at several operating frequencies;

• radiated and conducted results;

• spectrum plots and analyzer settings;

• occupied-bandwidth edges;

• temperature measurements;

• minimum and maximum voltage measurements;

• antenna configurations;

• timing information; and

• substantial compliance margins.

That information can be extremely valuable during a standards transition.

In practical assessments of existing EN 300 220 reports, the result is not necessarily 'repeat everything'. Requirements can fall into different categories: existing evidence may be reusable after reassessment; some requirements need assessment first with testing only if the evidence proves insufficient; and a smaller group may justify targeted new or repeat testing.

That is a very different outcome from commissioning a complete radio test programme from the beginning.

There is another lesson from looking at actual reports rather than only standards.

Imagine an old measurement still appears technically compatible with the new requirement.

If the old result is 20 dB below the applicable limit, the engineering discussion is very different from a measurement only 0.5 dB inside the limit.

In the latter case, even a relatively modest change in measurement procedure may make relying solely on the old result difficult to justify.

This means a useful GAP assessment considers not only "Was this measured?" but also "How was it measured, and how convincing is the existing evidence?"

There is no universal answer.

For an existing product, there are four practical outcomes:

• Requirement essentially equivalent and evidence sufficient - reuse existing evidence.

• Requirement or method changed but old data may demonstrate compliance - perform a technical reassessment.

• Existing evidence partly covers the new requirement - perform targeted supplementary testing.

• New applicable requirement with no equivalent legacy evidence - new testing is required.

And this is where product configuration becomes critical.

A transmitter-only fixed-frequency product may have a very different transition path from a transceiver using FHSS or polite spectrum access.

Two products can therefore have reports against exactly the same old standard and still require completely different actions under V3.3.1.

This is perhaps the most important distinction.

We can compare V3.2.1 with V3.3.1 and identify every technical change. That gives us the standards GAP.

But it still does not tell a manufacturer exactly what to retest.

For that we need to add:

Product configuration + Existing test report + Available measurement evidence

Only then do we have the product GAP.

This is also why simply asking a laboratory for a quotation against the latest standard can sometimes result in more testing than is technically necessary.

The better sequence is:

Understand the changes -> determine applicability -> assess existing evidence -> identify the actual missing evidence -> define the test programme.

EN 300 220-2 V3.3.1 contains meaningful changes, and manufacturers relying on older EN 300 220 reports should assess their technical documentation.

But "new standard" should not automatically translate into "new full test report."

Existing evidence may still have considerable value.

At the same time, assuming that an old PASS result automatically covers V3.3.1 can leave genuine gaps, particularly where applicability, test methodology or receiver requirements have changed.

The right answer lies between those extremes.

IoT Consulting Partners can compare your existing test report and product configuration against EN 300 220-2 V3.3.1 and identify:

• what can be reused;

• what needs to be reassessed; and

• what actually needs to be tested.

Before committing an existing product, or an entire product family, to a new laboratory programme, it may be worth finding out how much valid evidence you already have.

Because the most useful question is not:

"Do we need to retest?"

It is:

"What do we actually need to retest?"

This article is based on a detailed technical comparison of ETSI EN 300 220-2 V3.3.1 against earlier EN 300 220 requirements and practical assessment of legacy radio test evidence. Customer identities, product names and confidential report details have intentionally not been disclosed.