Sintered Ferrite Magnets: Grades, Agency and the Cost of Buying on Grade Name Alone

Ferrite magnets are the least glamorous product in any sourcing catalogue. They have a fraction of the energy product of neodymium, they're cheap, they weigh a lot, and nobody puts them on a slide. Which is exactly why they get bought carelessly — and why the argument usually happens later, in the motor, when the Kv curve doesn't match what the buyer calculated.

China makes the overwhelming majority of the world's sintered ferrite. The clusters are well established: Zhejiang (Hangzhou, Ningbo, Dongyang), Guangdong, Jiangsu and Sichuan. This is a mature, technically competent industry. The problem is rarely capability; it's that the specification exchanged between buyer and factory is three words long.

What a ferrite grade code actually means

The common designations are Y30, Y30BH, Y33, Y35 and equivalents under the IEC 60404-8-1 series, which uses grade numbers like Y30H-1 or Y30BH. The code is a shorthand for a band of magnetic values: residual flux density (Br), coercivity (Hcb and Hcj) and maximum energy product (BH)max, all at a defined temperature.

Here is the part that trips people up. A grade name describes a performance band, not a specific material, a specific manufacturer or a specific process. Two factories can both ship "Y30BH" and differ by 5% on (BH)max, because they're both inside the band. If your design has margin, nobody cares. If your design is marginal — a small motor where every millitesla is doing work — the difference is a re-design.

Worse, the published grade data is usually at 20 °C. A ferrite motor running at 80 °C loses Br, and the shape of that loss depends on the material and the coercivity. If you're building a fan motor, a pump motor or anything inside an enclosure, ask for the temperature coefficients, and ask at the temperature range you actually operate.

Magnetisation direction and the drawings

Ferrite is anisotropic or isotropic. Anisotropic material has much better properties, but only along the axis it was pressed. You cannot magnetise it in a different direction later and get the grade data — you get a fraction of it, and the factory will not necessarily flag it.

So the drawing must specify: the pressing orientation, the magnetisation direction (through thickness, multi-pole on a face, radial for ring magnets), the pole pattern, and the pole count. A multi-pole ring with eight poles on the outer diameter is a completely different tooling requirement from an eight-pole face magnet, and both are "a ring magnet."

Common new-buyer mistakes I've seen on drawings actually sent to Chinese factories:

A drawing with no orientation axis, so the factory presses it the convenient way and the buyer gets 40% of the expected flux.
Pole pattern described only as "poles as per sample", with no sample and no number.
Tolerances written as general machining tolerance, when the critical dimension is a bore or a mating face at ±0.05 mm and everything else can be loose.
No specification for the ground/chamfer edges, which is where chipping starts and where a lot of the visual reject rate comes from.

Ferrite is brittle and dimensionally stable but not precision-machined cheaply. Grinding a bore after sintering adds cost quickly. If you can design around as-sintered dimensions and use a locating feature instead, you'll save more money than by arguing about the unit price.

What the buyer has to test themselves

In practice, the factory will supply a magnetised product and may supply a test sheet. The test sheet will describe a sample, not your shipment. On a magnet order there is almost no such thing as a certify-every-piece option, so the testing plan is the buyer's.

What works for most small and mid-size buyers:

Incoming check on a sample basis. A flux meter or a Helmholtz coil gauss meter against a retained reference magnet from the approved first article. Compare total flux, not the single-point reading, and record it per lot.
One physical geometry check per lot. Key dimensions with a caliper on a defined number of pieces, plus a visual check for chips and cracks under oblique light.
Retain the first article. Keep one magnetised first-article piece and one demagnetised piece in a bag with the lot number. It is the only reliable way to resolve an argument a year later.

The gauss meter reading on a bare magnet varies with probe position enough that a careless operator can make a good magnet look bad. Write the jig down, not just the pass/fail number.

Coating, weight and freight

Bare sintered ferrite has mediocre corrosion resistance and a low surface resistivity. Many applications use a coating: epoxy, a nickel or zinc plating, or a spray coating. If your application involves humidity, salt or condensation, tell the factory and ask for the salt spray result on the actual coating they will apply. Uncoated ferrite in a wet environment will bloom rust from the surface within months.

Freight is where the economics bite. Ferrite density is roughly 4.9-5.0 g/cm³, so a pallet-sized carton of motor magnets reaches 1,000 kg quickly. Sea freight per kilogram is fine; the insurance and handling assumptions change if you're used to shipping foam or garments. And because the material is brittle, packaging quality — foam slots, layer cards, corner protection — matters more than the box. Ask for the export carton spec and, for anything fragile-shaped, the drop test result.

Ordering well, in one paragraph

Pick the grade as a performance band, not a label; ask for data at your operating temperature; put the orientation axis and pole pattern on a dimensioned drawing; agree which dimensions are critical and grind only those; decide the coating against your actual environment; agree the test plan and the retained samples before the first shipment; and price the freight by weight and volume, not by piece. Do that and ferrite becomes exactly what it should be — a cheap, boring, reliable component.