The document set a laboratory actually works with, what each file is used for, and why the version of a document often matters more than whether it exists.
In most certification projects, the testing is not what causes the delay. The test plan is agreed in a day or two, the slot on the schedule is booked, and then the project sits still for a week because the manual is not finished, the BOM describes a different product from the samples on the bench, or nobody can find the antenna specification. The documents are the part of the project that clients underestimate, and they are the part that decides whether the report can be issued cleanly at the end.
This article lists the documents a laboratory normally asks for, explains what each one is actually used for - which makes the requests easier to answer - and describes what happens when one of them is missing or out of date.
For almost every electronic product, regardless of market or directive, the laboratory works with these files:
| Document | What the laboratory uses it for | What goes wrong without it |
|---|---|---|
| Specification sheet / datasheet | Building the test plan: rated voltage, power, interfaces, operating modes | The quotation is built on guesses, and the scope changes once the real product is seen |
| User manual (final version) | Referenced by the test report and the technical file - accessories, ratings, warnings, intended use | The report cites a manual that no longer matches the shipped product, and the file has to be corrected later |
| Bill of Materials (BOM) | Identifying material groups for RoHS sampling, checking the tested configuration, worst-case analysis | Sampling misses parts, or the samples turn out to be a different product from the paperwork |
| Schematics and block diagram | Understanding circuits before EMC testing, and diagnosing a failure if one occurs | An EMC failure becomes a guessing game, and debugging time is billed while the lab probes blind |
| PCB layout drawings | Locating clock sources, switching regulators and antenna traces when emissions are high | Failure analysis takes far longer than it needs to |
| Label artwork | Checking mark, model number and responsible-company details against the applicable rules before the report is issued | A label problem is found after testing instead of before, when fixing it is still free |
| Product photos and full model list | Connecting the report to the product and covering every variant you intend to sell | A model you sell is not covered by the report, or the model numbers do not match what the buyer sees |
Two of these deserve a closer look, because they cause the most arguments.
The test report and the technical file reference the manual: the accessories listed in it, the input ratings, the warnings, the intended use. If the manual sent to the laboratory is a draft and the production manual ends up different - a new accessory added, a rating changed, a warning removed - the report is describing a product configuration that no longer exists. The fix is a file correction, which costs time and sometimes money, on a project that was technically finished.
A draft is acceptable for starting the project, as long as the laboratory knows it is a draft and the final version arrives before the report is issued. The failure mode is not "draft" - it is "final manual quietly changed and nobody told the lab".
Laboratories test what arrives on the bench, not what the drawing says should arrive. When the battery, the display module, or the power adapter in the samples differs from the BOM, one of two problems follows: either the report describes the sample, which is not what production ships, or the mismatch is discovered mid-project and the configuration has to be resolved before testing continues. Both are more expensive than checking the BOM against the samples the day before they ship.
If the product contains a radio - Bluetooth, Wi-Fi, a proprietary 2.4 GHz link, cellular - the document set grows:
The radio mode list is worth repeating. In real projects, the enquiry says Bluetooth, the sample turns out to run a second mode, and the project grows a test item in the middle. Declaring every mode in the first email is the single cheapest way to keep a radio project on its original quotation.
Nothing dramatic, which is why it keeps happening. The consequences are quiet and cumulative:
One reassurance on confidentiality: schematics and firmware are handled under the laboratory's standard confidentiality terms. Refusing to send them does not protect the design - it only guarantees that any EMC problem will be slower and more expensive to fix.
Everything a project needs can be sent in one email. Before samples ship, hand the laboratory:
The same list, sent with the quotation enquiry instead of after it, is what makes a first quote accurate - and it connects to the other end of the process, because these are the documents a buyer or marketplace eventually asks to see. Our overview of certifications for selling electronics in Europe covers that side.
We review document sets as a standard part of every project - it is faster to fix a manual or a label before testing than after the report. The testing and certification services page shows the markets we cover, and if you want a second pair of eyes on your paperwork before the samples ship, the details are below.
If you are preparing a submission and not sure whether your document set is complete, send your manual, specification sheet and BOM before the samples ship.
We will check which documents are missing, which versions are out of date, and what the radio and power configuration will add to the test plan - so the project does not stop in the middle.
Testing can often begin with the samples while documents are still being prepared, but the report cannot be issued until the manual, model list and configuration are resolved. Sending the core set with the samples avoids the pause at the end.
The BOM identifies the material groups sampled for RoHS, confirms the tested configuration, and supports worst-case analysis across variants. Without it, sampling can miss parts and the samples may not match the paperwork.
Yes, if it is clearly marked as a draft. The final version has to arrive before the report is issued, because the report and technical file reference it. The problem is never the draft - it is a final manual that changes without the laboratory being told.
Laboratories work under standard confidentiality terms. Schematics are needed to plan EMC testing and to diagnose failures quickly; without them, debugging takes longer and the time is billed. A block diagram is sometimes sufficient to start, but it rarely covers a failure analysis.
No. A pre-certified module simplifies the radio assessment, but the finished product still needs its own documentation - operation description, antenna details, manual, BOM and label review. A component report is not a product file.
Buyers, marketplaces and customs authorities check that the model in the report matches the model on the listing and the label. A mismatch usually means the document is rejected and has to be corrected or reissued, even though the product itself was tested and passed.