Peter Wang

About the author

The first thing I do with a draft is look for the numbers, then check where each one came from. Current, voltage, torque, strip length, conductor size, IP rating. If nobody can say which test record a figure came from, the figure comes out and goes back as “confirm with the drawing for your model”.

I have worked on connector design and production at our plant in Shenzhen for eight years, and I do the technical review on what we publish. Ryan and Sophie write the guides, and the numbers in them are mine to sign off.

Every draft gets the same technical review whether it runs to two thousand words or answers one question in two paragraphs.

Where the time goes

Certification claims take the longest part of a technical review I do. Putting UL 4128 on a page takes a second, and verifying it is the slow part. What I have to see is the certificate itself, the model printed on it, and the current rating on it. A certificate can name a series and still leave out the variant the page is selling, and that gap is the first thing I look for.

Standard numbers trip people up. IEC 60529 tells you how an IP code is tested. It does not say a connector suits a wet location, though I have seen it used that way on pages that should know better. Quoting IEC 62852 on an connecteur pour système de stockage d'énergie is a different mistake, and a buyer’s engineer will notice it before we do.

Torque values worry me more than anything else, because a customer reads 2.5 N·m and puts a wrench on it at 2.5 N·m. One digit wrong, and the joint is what fails first, usually a year later, when nobody remembers the batch.

Strip length comes up the same way. Two cables both marked 4 mm² can want different strip lengths, because jacket thickness does not appear in that number.

When I send a draft back

What comes back from me is usually a question about a figure, not a rejection. It comes down to one number: where it came from, which model it applies to, what happens outside that range. If the answer is a guess, the claim goes out as a range or it does not go out. A technical review that signs a number nobody can trace is worse than no review at all.

Where I work from

The plant is where I spend most of my week, and the test records I check against come off our own line. Either the record exists for that model, or it does not. The contrôle qualité files are where I find out which.

What I cannot answer

I can tell you what a test record says. I can also tell you which certificate covers which model, and where a standard stops applying. What I cannot tell you is what your installation will do over 25 years. That depends on the array, the wrench and the person holding it. When a question turns on your site, I say so instead of inventing a figure.

If a number here does not match what you measured, send the model and the batch with it. Ryan’s author page lists the guides, and the ones I have signed carry my name and the date at the top. For a technical review of your own drawing, send it through the formulaire de devis and mark it for me.