Insights › Magnetic cases

MagSafe Phone Case Manufacturing: Magnets, Shielding and What to Test

9 November 2026

A magnetic case looks like a plain case with a ring glued in. In production the ring is the hardest part of the build, because it has to hold a charger, pass power through, survive a drop and stay aligned to a coil it never touches. Buyers who order these cases without a drawing usually find out at the sample stage that the magnets are in the wrong place, or that the case gets warm and the phone stops charging at 7.5 W instead of 15 W.

This is the build sequence we run, and the points where a cheap version fails.

Magnet grade comes before magnet count

Almost every magnetic case uses sintered NdFeB, nickel-copper-nickel plated. The grade decides how much pull force you get per millimetre of thickness, and thickness is what the case cannot spare. N52 in 0.5 mm holds roughly what N35 holds in 0.8 mm, so a thin case has to use a high grade or accept a weak ring.

Count is a secondary decision. The iPhone ring is a 18-magnet layout around a central alignment magnet; a 16-magnet ring can hit the same pull force if the grade is higher, but the magnetic field becomes less even, and uneven field is what causes a charger to disconnect when the phone shifts a millimetre. For a case that will be sold as MagSafe compatible we keep the 18-magnet geometry and tune grade instead of count.

Pull force is the number to put in the specification, not magnet count. Measure it with a digital force gauge on the finished case, mated to a bare reference phone or a steel plate at the same distance as the real coil gap. Typical commercial targets sit between 800 gf and 1,200 gf. Below 800 gf the case slides off a car mount on a bump; above about 1,400 gf the user has to peel the phone off, and the ring starts to pull credit cards out of a wallet.

Shielding is what protects the phone, not the magnet

Behind the magnet array there is a ferrite or nanocrystalline sheet. It does two jobs: it shapes the field toward the charger so less flux leaks sideways, and it protects the phone's own electronics and any cards in the same pocket. Skip it, or use a thin painted-on coating instead of a real sheet, and the case will still snap to a charger but will interfere with the magnetometer and with RFID cards.

The shield sits between the magnets and the phone, never between the magnets and the charger. On a case with a microfiber lining, the stack from outside in is: shell, magnet array, shield, lining. If a supplier puts the shield on the outer face because it is easier to assemble, the field is blocked and charging power drops sharply.

Recess tolerance is where most failures start

The magnets sit in a recess moulded into the TPU or PC shell. That recess is typically 0.3 to 0.5 mm deep and the magnet is a press fit. Two tolerances matter:

• Depth. If the recess is too shallow the magnet stands proud and the phone rocks; if it is too deep the magnet sits further from the coil and charging efficiency falls. Hold depth to plus or minus 0.05 mm.
• Position. The ring centre has to sit on the coil axis within about 0.3 mm. A ring that is off-centre still charges, but at reduced power, and the customer experiences it as a slow charger rather than a defective case, which is the worst kind of complaint because nobody can reproduce it.

Both are checked on the first article, not on a finished carton. If your supplier only measures the case outline and not the magnet pocket, the pocket is uncontrolled.

Three assembly routes, three different failure modes

Gluing into a moulded recess

Most common, cheapest tooling. The ring is placed by hand or by a jig and held with structural adhesive. Failure mode is adhesive creep under heat: a case left in a car reaches 70 to 80 degrees Celsius, the glue softens, and the ring shifts a fraction of a millimetre. Specify an adhesive rated above 100 degrees Celsius and ask for a heat soak test at 80 degrees for 24 hours with a pull force measurement before and after.

Insert moulding

The magnet assembly is placed in the tool and the shell is moulded around it. No adhesive, no creep, and the ring cannot come out. The trade-offs are tooling cost and the risk of the insert shifting under injection pressure; that risk is real and shows up as a ring that is consistently 0.2 mm off in one direction. Insert moulding is worth it above roughly 30,000 units per colour.

Sandwiched between shell and liner

A two-piece construction: an outer shell, a magnet carrier plate, and an inner liner that clips or is glued over it. The magnet never touches the user's phone directly, which is the safest for cards. This is the route for cases with leather or fabric backs, where the ring is hidden under the material and thickness is available.

Four tests that tell you whether the ring works

1. Pull force, finished goods. Force gauge, defined mating surface, five samples per batch. Record the minimum, not the average.
2. Charging power and thermal rise. Charge a reference phone from 20 to 60 per cent with the case on and off, log watts and case surface temperature. A drop of more than about 15 per cent in peak power means the stack is losing flux or the shield is misplaced.
3. Drop retention. Drop the assembled case and phone from 1.2 m onto concrete, six faces. A ring that separates is a safety issue, not a cosmetic one. The drop test article covers the method in full.
4. Heat ageing. 80 degrees Celsius for 24 hours, then repeat the pull force and the charging test. This is the test that catches cheap adhesive, and almost no supplier runs it unless asked.

What the magnet adds to cost

Material is not the main line. An 18-magnet N52 array with a proper shield lands between roughly 0.15 and 0.35 US dollars per case depending on grade and shield type, and insert moulding adds tooling rather than unit cost. The real cost is yield: a magnetic case has more assembly steps than a plain one, and every step that is done by hand is a step that can put a ring off-centre. When a quotation for a magnetic case comes in below the plain case price, the difference is being taken out of the shield or the adhesive.

If you are building a magnetic case for the first time, the mould timeline applies, plus one extra week for the stack validation. Order the magnet and shield to drawing before the tool is cut; changing magnet thickness after the recess is machined means recutting the insert.

We mould magnetic cases on our own tools, TPU or PC shell, 18-magnet array, ferrite shield, with pull force and heat ageing reported per batch. If you have a coil drawing or a target pull force, send it and we will quote the stack rather than the case.

Two magnetic designs described above are already in the catalogue: Aegis Soft-Touch Matte Dual-Layer Magnetic for iPhone 18 Pro Max and Aegis Soft-Touch Matte Dual-Layer Magnetic for Galaxy S26 Ultra. Both ship from stock colourways at MOQ 50.