What a CR80 specification must add to the nominal size: acceptance limits, corner and slot geometry, keep-outs and a feed trial on finished cards.

A nominal card size is only the beginning of a mechanical specification. A buyer also needs to know whether the finished card will feed, sit and move correctly in the equipment used by the client.
CR80 is the shorthand most quotations use for the credit-card-sized format, and as shorthand it works. It is not an acceptance specification. Two cards can both be sold as CR80 and still behave differently in a dispenser, a badge holder or a desktop printer, because the properties that decide fit are the permitted variation around the nominal size, the corner geometry, the position of any slot, the thickness build and the surface finish. None of those is settled by the name.
This guide covers the dimensional part of a card order: what belongs on the drawing, how to express acceptance limits, what to test on a finished card and what a dimensional report should contain. It deals with plan geometry and fit. The layer build and the ways a laminated card fails are covered in the card thickness and lamination guide.
CR80 refers to the credit-card-sized ID-1 format described in the ISO/IEC 7810 family of standards. The standard gives a nominal outline and a nominal thickness for the format, which is why a card from one factory and a card from another normally drop into the same wallet. That much is genuinely standard.
What the standard does not describe is your client's equipment: how much clearance a particular dispenser needs, how a holder grips a slot, how a coated surface behaves against a feed roller, or what happens when a raised feature meets a thickness gate. Those requirements come from the installed base, not from the format.
So treat the word standard on a quotation as a starting point. Put the required finished dimensions on a controlled drawing, name the edition of the standard you are working to, and add any constraint imposed by the printer, dispenser or holder manufacturer. A supplier cannot be held to a requirement that was never written down.
Before writing limits, list every piece of equipment the finished card will touch: the printer or encoder used to personalise it, any dispenser or stacker, the holder, clip or wallet it will live in, and the reader it has to work with. Each interface constrains a different property, and the strictest one governs the specification.
A card printer constrains thickness, flatness and edge quality, because the card has to separate from a stack and pass a gate. A dispenser constrains the outline and the warp, because a bowed card in a spring stack either jams or feeds two at a time. A holder constrains the outline, the corner radius and the slot. A contactless reader constrains where the antenna sits and how thick the body is around it, which is why antenna and read range is a separate question from fit.
Equipment families change over their life. Asking for the model, the revision and the manufacturer's card requirement where one exists turns a vague compatibility question into a list of figures you can put on a drawing. Where the client cannot supply it, plan a feed trial instead and say so in the order.
A nominal dimension describes the target. A tolerance describes the variation accepted around it. A specification that gives only the nominal has not defined a pass, and a batch measured against it can only be released on opinion.
Give width, length and thickness their own limits rather than copying one figure across all three: the equipment constrains each differently, and the processes that set them are different too. Add the characteristics the outline does not cover, in particular flatness or warp, edge finish and burr, and squareness. Where a characteristic matters enough to reject a lot, it needs a limit and a method, not an adjective.
Measurement uncertainty belongs in the same conversation. If the method cannot distinguish a result near the agreed limit from one just outside it, that has to be resolved before the reading is used to accept or reject a lot.
Geometry is fixed at the punch, and the tool is part of the specification. For a standard outline the corner radius comes with the format, but it is still worth stating on the drawing, because it is the feature most often assumed rather than agreed. A radius that is too sharp wears a holder and can catch in a feed path. One that is too generous lets the card move inside a slot.
Slots and holes need a position, an orientation and a datum. Dimension them from a named edge rather than from the centre of the artwork, so that the drawing and the punch agree. Then check the slot against what sits under it: the antenna path in an RFID card, the stripe in a magnetic stripe card, or a signature panel. A slot that crosses an antenna is a scrapped sheet, not a late revision.
For a custom outline, agree the drawing and the inspection method before tooling is approved. The tool is made once and reused for every reorder, so a shape accepted informally on the first run becomes the standard for every batch after it. The same applies to moulded formats such as key fobs, where the mould, not the drawing, decides what arrives.
Run decorated, production-representative cards through the actual feed path. A blank white sample proves very little: it has none of the decoration, none of the raised features and often none of the surface finish of the card that will ship. A matt or textured laminate changes friction in a hopper. A holographic patch, an embossed character or a signature panel changes the local thickness. The guide to printing methods and finishes covers which decorations sit on the surface and which are built into the construction.
Record what happens in categories rather than as a verdict. Jams, double feeds, misreads and physical damage have different causes and different fixes, and a trial that ends in a single word tells you nothing when the second batch behaves differently. Note the equipment, the sample size and the conditions, because temperature and humidity affect how a stack separates. Mechanical acceptance and functional acceptance are separate tests, even on the same sample: a card can fit perfectly and still fail to read.
Ask for a report that names the drawing revision it was measured against, the characteristics measured, the locations on the card, the method and instrument, the sample size and the observed results. A single figure with the word pass beside it is a statement of confidence, not evidence.
Results near a limit deserve explicit treatment: state whether the reading stays inside the limit once measurement uncertainty is taken into account, and agree in advance what happens when it does not. Keep the fit evidence, the feed trial and the equipment used alongside the dimensional record, because a compliant outline does not prove that every holder, decoration or downstream process works. The delivery inspection checklist covers what to check on receipt, and specifying a card order covers the brief as a whole.
Illustrative scenario, not a customer case. A distributor moves a membership programme from a plain printed card to one with an embossed number. The outline does not change, the artwork is approved quickly, and the first batch passes a dimensional check against the same drawing as before.
In the field the cards work in the reader and sit correctly in the wallet, but the desktop printer the client uses for replacements starts double feeding. The cause is not the outline. It is the raised characters, which change how cards separate in the hopper. The check that would have caught it is a feed trial with decorated cards in the client's own printer, before the run was authorised, plus a drawing that treated local thickness over a raised feature as a characteristic in its own right.
Kaway manufactures in its own plant in Dongguan, China, and works from a drawing rather than a description. The most useful thing to send with a dimensional enquiry is the outline, the corner and slot geometry, the thickness build and the list of equipment the card has to pass, so that tooling, inlay position and print layout are settled in one revision.
Minimum order is 500 units. Standard production is 13 calendar days, 8 calendar days on express and 12 calendar days for wooden cards. A free sample is available where the request is qualified, which is the practical way to run a feed trial before committing to a production run. Quote requests are answered the same working day. The standard constructions are on the PVC card page and the RFID and NFC card page.
Discuss your card construction and equipment requirements with Kaway Group.
No. It is a size description. The order still needs the agreed drawing, the tolerances and any constraint imposed by holders or processing equipment.
In everyday purchasing they are used for the same credit-card-sized format. ID-1 is the designation used in the ISO/IEC 7810 family, and CR80 is the trade name that grew up around it. Name the standard and the edition on the drawing so there is no ambiguity.
Usually yes, as a starting point. Check that it still matches the equipment in use and that it covers the characteristics you now care about: a drawing written for a plain card may say nothing about a slot, a keep-out or a raised feature.
Agree it with the manufacturer against the strictest interface you can identify, then prove it with a feed trial on finished cards. An arbitrary tight tolerance raises cost without improving fit, and no tolerance at all leaves the acceptance rule undefined.
About this guide
Published by Kaway Group. The date on this page shows when this guide was last updated. If anything here is out of date, or contradicts what your integrator tells you, tell us and we will correct it.
Send the reader or lock model, the quantity and the date you need them. That is enough for us to answer.
Talk to us