Most of these cases arrive as an assumption rather than a question. The second colour is ready, or a distributor asks for a higher-capacity version of something that already has a report, and someone in the company says: it is the same product, the report covers it. Sometimes that is right. Sometimes it holds until a marketplace compliance team, an importer or a buyer's quality engineer opens page one of the report, reads the model list, and finds a model name that is not there.
Whether a model can be added to an existing test report depends on one thing: what actually differs between the two models, and whether that difference touches anything the laboratory measured. Everything else follows from that - the cost, the timeline, and whether the answer is a piece of paperwork or a bench booking.
A test report covers the models it tested
A test report is a record of a measurement. It names the sample that was on the bench: its model designation, electrical ratings, mechanical construction, firmware version, the accessories used, and the radio modes that were exercised. When a report lists several model names, those are the variants that were declared before testing started and assessed as one family, usually with the most severe of them tested as the representative sample.
That is not the same as holding a certificate. The documents in this area do different jobs, and treating them as interchangeable is where most of these problems begin - the roles of a report, a declaration and a third-party certificate are set out in test report vs certificate. The point here is narrower: a report is not a licence that you extend by printing a new name on the carton. If the model is not in the list, the report does not cover it.
Two situations that get confused
Nearly every "can we add a model" question is really one of two different cases.
- Variants declared before testing. The factory said up front that the product ships as four models, differing only in colour or plug type. The laboratory picked the worst case, tested it, and listed all four names in the report. Nothing has to be added later, because they were covered from the start. This is by far the cheapest route, and it costs nothing more than writing the full model list on the day the project is quoted.
- Variants added after the report is issued. The report exists with one model name on it, and a second one appears later - a new market needs a different plug, a customer wants their own brand on the enclosure, or the range simply grew. Now the question is whether the difference is something the tested sample already represents.
Only the second case needs a decision, and that decision is a comparison, not a formality.
What varies between models, and what it means for the report
| What differs between the two models | Usual effect on the report |
|---|---|
| Brand or model name only - the hardware is identical | No bench work. The added model is documented and the report is amended to list it, provided the identical hardware is evidenced |
| Colour, paint, trim, packaging | Normally documentation only, with the more severe variant as the tested sample |
| Enclosure material or shape, internal layout | Reassessment. Shielding, creepage and clearance distances move with the construction |
| Output power, current or voltage rating | Retest of the affected items - ratings are measured values, not label text |
| Different battery cell, or a higher-capacity pack | The battery items are tied to the cell and the pack; the EMC and safety items depend on what else moved |
| Different supplied adapter or charger | The adapter is part of the tested configuration, so the configuration changes with it |
| An added radio mode, or a changed antenna | New radio testing. Radio behaviour is tested mode by mode |
| New firmware version, power table or band list | The tested firmware version is recorded, and RF behaviour can change with it |
The pattern is straightforward: names and cosmetics are cheap; anything electrical, mechanical or radio is not.
Why one sample can represent several models
This is the logic that makes family coverage possible at all. A laboratory does not test every variant. It identifies the worst case in each direction - the variant most likely to produce the highest emissions, the one drawing the most current, the one with the tightest clearances - and tests that one. If every other variant in the range is the same or less severe in every measured respect, the result for the worst case covers them.
The catch is the word every. A second model can be less severe for emissions and more severe for temperature rise. Family coverage fails the moment one variant is worse in one parameter, because the tested sample is then no longer the worst case for that item. That is why a family assessment is a comparison table rather than a one-line statement that the models are similar, and why a laboratory will ask for both bills of materials before answering.
How an added model is handled in practice
There are three possible routes, and which one applies follows from the table above.
- Amended or re-issued report, no bench work. Where the difference really is name or cosmetic only, the laboratory reviews the evidence, confirms the tested sample still represents the added variant, and issues the report with the additional model in the list - as an amendment or a re-issue with a new date. There is still work in it: someone has to verify that the hardware is identical, and the paperwork has to be traceable.
- Delta test. Where the difference affects some measured items but not all, only those items are repeated. The new report references the original, covers the added model, and states which results carried over and which were measured again. The bench time is a fraction of the original project.
- New report. Where most of the scope is affected, a fresh project is cheaper than a patchwork of carried-over results. This is the exception, not the norm.
Two things do not count as any of these. Editing the model list in your own copy of the report, and relabelling the product. Both leave you with a document that does not describe the product you ship, and it works right up until someone checks - by which point the goods are usually already on the water.
What the other side actually checks
Buyers, importers, marketplaces and customs authorities do not read a report from cover to cover. They open it and look for four things: the model name that matches the shipping carton, the standards applied, the test dates, and the name of the responsible company. A model that is absent from the list fails the first of those checks no matter how similar the product is. It is also why the model list on the report, the model list on the declaration and the label have to agree with each other - the reasons they usually do not are covered in why the laboratory asks for your label and artwork.
What to send for a variant assessment
If you want a firm answer instead of a provisional one, send these five things together:
- The complete model list as it should appear on the report, the existing names and the new ones
- A difference table: what changes between the models, item by item
- The bill of materials for the original and for the new variant, so differences can be confirmed rather than assumed
- The electrical ratings and the label artwork for the new model
- The firmware version, if it differs from the one recorded in the report
A photograph of both units helps more than it sounds. A cosmetic-only change sometimes turns out to include a relocated connector once the two boards are placed side by side.
Situations that come up most often
A private-label version for a distributor. The hardware comes from the same factory and only the brand name on the enclosure changes. This is the easy case, but it needs evidence rather than an email: a statement from the manufacturer plus matching bills of materials. Without that, no laboratory can put a second brand on the report.
A higher-capacity battery version. The instinct is that the case and the electronics are unchanged, so the report should still apply. It does not, because the battery-related items were measured on the smaller pack. Expect the battery tests to repeat and the EMC and safety items to be reviewed against the new pack.
Adding a plug variant for a second market. The electronics are identical and only the supplied cord set differs, yet the cord set is part of the tested configuration, so the configuration has changed. If you know from the start that the product ships with three plug types, declaring all three before testing removes the problem entirely.
A model that only adds a radio mode. Occasionally a new model turns out to switch on a second radio that was already on the board but never declared. Radio behaviour is tested mode by mode, so this is a new radio test - and the honest version of that story is much cheaper than the version a buyer's engineer discovers later.
If what changed instead is a part inside one product - a substitute capacitor, a second-source connector - that is a different question with a different answer, covered in does changing a component require retesting.
Where we come in
On our side, family coverage is normally settled at quoting time. We ask for the full list of models you intend to sell, not just the one you plan to ship first, and we build the test plan around the worst case in the range so the report lists the whole family from the beginning. When a model is added afterwards, the review is a comparison of the two configurations against the tested sample, and the outcome is one of three things: an amended report, a defined delta test, or a clear statement that the new variant is not covered. Whichever it is, the reasoning is written down with the file, so the family rationale still stands up if a buyer or an authority asks about it two years later. That documentation is also what an importer needs when a marketplace requests evidence, since those requests usually arrive with a deadline attached. If it is not obvious which category your case falls into, our testing and certification work covers the EMC, radio, safety and battery scope most of these ranges sit in, and the initial comparison is a short piece of work.
If you are preparing to launch a second variant, send the model list you intend to sell, a short note on what differs between the existing model and the new one, and the original report. We will tell you which of the three routes applies - an amended report, a delta test, or a new project - before anything is booked on the bench.