Five details to check on an IEC 62133-2 test report

The five details worth checking on an IEC 62133-2 test report before a PPAP meeting are: whether the item tested was a cell or a battery; the edition and amendment against the requirement you are actually held to; the current status of the certificate the report hangs off; the product description, ratings, markings, and photo pages against the configuration you are buying; and whether the report is complete as a document. Any one of those can fail while the front page still looks immaculate, which is why the review has to be mechanical instead of a read-through.

I run this at one desk with two windows open, the supplier's PDF on the left and the specification on the right, and I give it twenty minutes. Not five. Five minutes is what produces the email that says "certificate looks good," and then four months later, a line-down argument about a model number suffix.

The file usually lands the evening before, subject line something like "62133 cert," one attachment, no message body. That is the moment to get suspicious, not because anyone is lying, but because a document that arrives with no commentary has not been read by anyone yet, including the person who sent it.

Five passes, always in the same order: scope, edition, status, physical match, completeness.

Start where the document is weakest

Scroll past the certificate page into the body of the test report and find the scope statement. Somewhere in the first few pages the report says what was tested. Read that line twice. It is the single fastest way to know what you actually hold. A cell and a battery are not the same article under IEC 62133-2, and the report will name which one it covered. If your deliverable is a pack with a BMS, a housing, and a connector, a report on the bare cells inside it is evidence about the cells. It is not evidence about the pack. That is not a technicality to wave through in the room. It is the first question a competent customer engineer asks, and you want the answer before they do.

Then the edition. The header should name the standard with its edition, and typically the amendment. Compare that string against what the specification requires, character for character. An older edition is not automatically a failure. It depends entirely on what the requirement says, and sometimes the honest answer is "acceptable, documented, move on." Sometimes it isn't, because the amendment changed something substantive rather than cosmetic. What you cannot do is assume. I have watched programs stall over a document everyone had already approved emotionally and nobody had read past the title.

A report with no edition stated in the header is not a pass with a caveat. It is an incomplete document.

The certificate is a snapshot, not a status

A PDF is a picture of a certificate taken on the day someone downloaded it. Certificates get withdrawn, suspended, and superseded. The copy in your hand can be entirely genuine and completely stale.

So the third pass is a lookup, not a reading. Take the certificate number and the issuing body named on the page, and check current status in that body's public certificate database, or ask the supplier for a status confirmation dated inside the last quarter. Two things fall out of that lookup. One, whether the certificate is live. Two, who the certificate holder actually is, and whether that entity is the manufacturer, an affiliate, or someone else entirely. If the holder is not the company on your purchase order, that is a question to raise now, calmly, rather than a surprise to absorb during a quality escalation.

I keep the result in a one-line note with the date and who ran it. Not bureaucracy. It is the difference between "we checked" and "we think we checked."

Where a mismatch shows itself without anyone saying a word

The product description and photo pages are where a report quietly stops matching reality. Look at the model designation on the certificate and compare it to the part number on the drawing. Watch for the words that broaden a document: series, family, type. Those words are legitimate in themselves, and also exactly where a certificate can end up covering something adjacent to what you are buying.

Then look at the photographs and the rating markings. What is in the picture, whether a bare cell, a shrink-wrapped module, or a cabled assembly, should be the thing you are contracting for. The label visible in the photo should carry the voltage and capacity your drawing carries. If the photo shows a marking label at one voltage and your specification says another, that is the whole review right there. Not a formatting issue. A different product.

The last pass is the least glamorous one, which is why it catches people: count the pages. Your PDF viewer reports how many pages the file contains. The footer reports how the report is numbered. Those two numbers agreeing proves nothing, but when they disagree you have a gap, and the gap is often the middle of the document where the test data lives. While you are in there, confirm that the clauses the report says it applied are followed by actual measured values and data sheets, not just a table of clause numbers with a checkmark beside each. A clause list with nothing behind it is a cover sheet pretending to be a report.

Missing pages, no stated edition, no data sheets. Any of those three means the correct output of this review is a request for more, not an approval.

What you do with the result

The output is short. Either the document is clear enough to present, or it isn't, and if it isn't, you send one message listing the specific gaps and the specific pages, asking for a corrected or supplemented report. Suppliers respond to that far better than to "the certificate seems off," because it tells them which drawer to open.

Two honest caveats. First, none of this decides whether a given certificate satisfies a given customer's requirement. That conversation belongs to the people who wrote the requirement, and the answer genuinely varies by program. Second, this checklist clears documents. It does not clear product. A clean 62133-2 report says the tested article passed a defined set of tests under a defined edition. It says nothing about whether the pack arriving in month nine came off the same line as the pack in the photo.

Which is why I still do this twenty minutes before the meeting instead of the week before. It is the last cheap moment to catch a mismatch. After the PPAP is signed, the same discovery costs a great deal more.