Sourcing computer mice

How the failure happens. A micro-switch's contact degrades with use and contamination until a single press produces electrical bounce that the firmware reads as two clicks. It is progressive, it starts on the button used most, and it appears months or years in — which is exactly when it costs you a warranty claim and a review rather than an inspection failure.

It is a specifiable part. Switches come from named makers with published ratings, so put the maker and part number in the specification and on the bill of materials. The cycle rating belongs to that datasheet and is not restated here — ask for the datasheet, and note that a rating is a laboratory figure under defined conditions rather than a promise about a customer.

Verify it on a teardown, not on a document. Substituting a visually similar switch is the easiest cost reduction in the product and the hardest to see. Open a unit from each production batch and read the markings on the switch body.

Firmware debounce is the other half. A well-written debounce routine masks early degradation and a badly written one produces double-clicks on healthy switches. Ask who writes the firmware and whether the debounce interval is adjustable — on a product you brand, that is the difference between a fixable complaint and a recall.

And consider whether the switch is socketed. A few designs let a user replace it, which turns the category's defining failure into a repair. That is a genuine differentiator in markets where repairability is a selling point.

Glide feet. Material, thickness and area decide how the mouse moves and how quickly it starts to drag. They are consumable, so ask whether spares are supplied — a cheap inclusion that measurably improves reviews.

The cable, on wired models. Stiffness and memory are what people actually feel; a cable that holds a coil from the packaging drags the mouse and reads as a quality fault. Specify the jacket and the strain relief, and check it after being coiled in a box for weeks.

Shell finish. Soft-touch coatings are pleasant and can degrade into a tacky surface over time in warm, humid conditions — a well-known long-term complaint. If your market is hot and humid, specify accordingly and age a sample deliberately.

Weight and balance. Both are specification lines and both are cheap to get wrong; a mouse whose weight is concentrated at the back feels different from one balanced under the hand even at the same total mass.

The scroll wheel. Encoder type, detent feel and whether the wheel develops a dead zone. It is the second-most complained-about mechanism after the click and is worth its own cycle test.

Wireless adds three obligations: qualification with a Declaration ID against a member company for Bluetooth models, per-market radio approval, and a battery — either cells with their transport documentation or a rechargeable pack with its own specification. Proprietary receivers are a separate radio again and a spare-parts liability if lost.

Reconnection behaviour is the support load. After sleep, after switching hosts, after a battery change. Test all three on every host family you sell alongside, because this is what generates tickets rather than the sensor.

Treat headline sensor resolution as marketing. Beyond a certain point the number stops describing anything a user experiences, and it is routinely inflated by interpolation. What matters is tracking behaviour on the surfaces your customers use — glass, wood, cloth, a desk with a coating — tested on samples rather than quoted from a listing.

Polling rate is similarly oversold and interacts with the host and the connection; if you claim it, measure it on the connection type you ship rather than repeating the chipset's maximum.

Software. If the mouse has configurable buttons or lighting, somebody has to maintain the application across operating-system updates. Ask whose it is, under whose developer account it is distributed, and what happens when the host changes.

Business licence and business scope (经营范围) first, then: whose design and whose firmware, and can the debounce and button mapping be changed for you? A mouse is a small embedded product and that answer decides whether a field complaint is fixable.

Buy several and use them for weeks, on several surfaces and several hosts. Then run a click-cycle test on the primary button — a simple rig is enough — and check the switch markings on a teardown against the bill of materials.

Age one deliberately: heat and humidity for the shell coating, and a coiled cable for the drag test. Both failures are long-term and both are predictable.

Check the scroll wheel across many revolutions, the feet after extended use, and the reconnection behaviour on wireless models after sleep and after a battery change.

Every supplier on this site is a manufacturer, and the registered-scope check is one of the things confirmed before a listing goes live.