Magnet-circuit adhesive is the bond that holds the top plate, magnet and yoke of a loudspeaker driver in a stack, and bonding that stack is the step most buyers never write a specification for. The cone gets a drawing. The spider gets a compliance figure. The magnet assembly usually arrives on the purchase order as "magnet circuit, bonded" and nobody asks what is holding it together, because the magnetic attraction between the parts is so strong that the stack feels permanent in the hand.
That feeling is misleading. The attraction clamps the joint in compression along one axis and does almost nothing for the other five degrees of freedom. What resists lateral slip, torsion and shock is the adhesive, and the tolerance it has to hold is not measured in millimetres. In a typical driver the voice coil runs in a gap with a few tenths of a millimetre of clearance on each side. A top plate that walks 0.15 mm off-centre during a drop test does not separate, does not rattle in a static check, and produces a driver that rubs at high excursion and gets returned six months later.
What the magnet-circuit bond actually carries
The stack is three parts. The yoke (also called the back plate or T-yoke) carries the pole piece. The magnet — a ferrite ring, or a smaller neodymium slug in a modern design — sits on it. The top plate closes the circuit and its inner diameter forms the outer wall of the magnetic gap. Two bond lines, both flat annular joints, both loaded in ways that a room-temperature tensile figure will not describe.
The loads, in rough order of how often they cause trouble:
- Lateral shear from handling and shipping. A driver dropped in its carton decelerates hard. The magnet mass times that deceleration goes straight into the bond line as shear. On a large ferrite motor the magnet can be the heaviest part of the driver.
- Torsion. Assembly fixtures, press-fit baskets and the occasional operator twisting a stack to seat it apply a moment about the axis. A joint with good tensile numbers and poor shear behaviour gives here first.
- Thermal cycling. Steel, ferrite and neodymium have different coefficients of thermal expansion. Every power cycle works the bond line a little. The joint that survives 500 cycles and fails at 3,000 is a creep and fatigue problem, not a strength problem.
- Sustained compression. The magnetic attraction is permanent. An adhesive that creeps under constant load slowly thins the bond line, and as the line thins the stack's alignment moves. This is a different duty from the flex-fatigue problem at the tinsel lead-wire bond, and it wants a different grade.
Here is the detail that catches people out: because attraction supplies most of the clamping force, a magnet-circuit joint can pass a straight pull test convincingly and still be wrong. Pulling the stack apart tests the one direction the magnet is already helping with. Ask for shear data, ask for it after thermal ageing, and if you are running the test yourself, test in shear.
Surface prep on plated steel decides the joint
Yokes and top plates are low-carbon steel, and almost none of them are bare. Zinc plating, nickel plating, electrophoretic coating or a phosphate conversion layer goes on for corrosion resistance before assembly. Your adhesive is not bonding to steel. It is bonding to whatever is on top of the steel, and the strength of the joint is capped by the adhesion of that layer to the substrate underneath it.
This produces the second detail worth knowing: read the fracture surface, not just the number. After you break a test joint, look at the two faces.
- Adhesive left on both faces, torn through its own thickness — cohesive failure. The joint failed at the adhesive's real strength. This is the result you want, and the number is meaningful.
- One face clean metal, the other carrying all the adhesive — adhesive failure at the interface. Usually a surface-prep, contamination or wetting problem, not an adhesive-selection problem. Changing to a stronger grade will not fix it.
- Plating peeled off with the adhesive still attached to it — the plating adhesion is your limit. No adhesive will improve this joint; the plating specification has to change.
The contamination sources are predictable. Deep-drawing oils sit in the pores of a phosphate layer and do not come off in a quick solvent wipe. Rust preventive sprayed on before a sea crossing is designed to be persistent. Plating brighteners and residual chromate leave a weakly bound surface layer that can hydrate in humidity and take the bond with it — which is why a driver that passed at build fails a damp-heat test weeks later.
Magnets have their own version of the problem. Sintered ferrite is porous and friable; the outer skin of loose particles is part of the joint whether you want it to be or not, and a wetting adhesive that penetrates the pores behaves differently from one that sits on top. Sintered neodymium is corrosion-prone and therefore comes coated — nickel-copper-nickel, zinc, or an epoxy coating — so once again the bond is to a coating, and the coating's own adhesion is the ceiling.
When you send an inquiry, send the plating and coating specification along with the base metal. Aosibo Adhesives lists magnet-circuit adhesive for the top plate, magnet and yoke stack as part of its loudspeaker range, and states that viscosity, cure schedule, pot life and substrate data are shared at inquiry rather than published — which means a substrate description of "steel and ferrite" will get you a generic answer. "Zinc-plated low-carbon steel, alkaline-degreased, and sintered ferrite Y30" will get you a specific one. The same discipline that applies to a product specification sheet for any Chinese supplier applies here: the quality of the answer tracks the quality of the input.
Bond-gap control, and the two ways it goes wrong
An adhesive joint needs a bond line. Metal-to-metal contact with adhesive only in the surrounding voids is not a thin bond, it is a starved one — the load concentrates at the contact points and the joint fails at a fraction of its rated strength.
This is harder to avoid in a magnet circuit than almost anywhere else, and here is the third detail: the assembly self-clamps. As soon as the top plate comes within a few millimetres of the magnet, the attraction snaps it home with a force nothing in the fixture is resisting. Uncured adhesive gets squeezed out. On a large ferrite motor that closing force is substantial and it acts for the entire cure, so even a joint that starts with a good bond line can thin toward metal contact while it sits in the oven.
Two counters are common in driver plants, and both are worth asking your supplier about:
- Gap-control features on the parts. Stamped dimples, a shallow ring groove, or a stepped seat that mechanically limits how close the faces can get. This is a part-drawing change, so it belongs in the discussion with your magnet-assembly supplier, not just your adhesive supplier.
- Gap-control filler in the adhesive. Some structural grades carry glass beads or a rigid filler at a controlled diameter, which sets a minimum bond line no matter how hard the joint is clamped. Whether any grade in a given range contains one is a formulation question. Ask directly; it is not something you can infer from a viscosity figure.
The second failure mode is squeeze-out into the gap. Adhesive pushed radially inward from the top plate's inner diameter ends up in the magnetic gap, where the voice coil moves. A cured bead in the gap is unrepairable scrap, and it is a scrap mode that shows up in batches when a dispense pattern drifts or a viscosity lot changes. Set the dispense pattern back from the inner diameter, and treat any change in dispensed viscosity as a change that needs re-validation, not a cosmetic difference. This is one of the checks worth writing into a pre-shipment inspection scope if you are buying finished magnet assemblies rather than bonding them yourself.
Failure modes, and how to specify against them
What a failed yoke bond does to a driver
Failures here rarely announce themselves. The stack does not fall apart; it moves.
| Bond condition | What changes in the motor | What the buyer hears or measures |
|---|---|---|
| Top plate shifts laterally 0.05–0.1 mm | Gap becomes asymmetric around the circumference | Rub and buzz at high excursion only; passes a low-level sweep |
| Asymmetric gap, coil off-centre | Force factor varies with position and direction | Rising even-order harmonic distortion, especially second harmonic |
| Progressive creep under thermal cycling | Gap narrows on one side over months | Field returns clustered by production date, clean at incoming QC |
| Magnet shifts axially | Coil rest position moves relative to the gap centre | Asymmetric excursion, early bottoming on one half-cycle |
| Bond releases under shock | Stack loses alignment entirely | Coil seizes; driver is dead, and the carton looks fine |
The reason this matters commercially is that none of these show up in a functional check at the end of the line. A driver with a marginal magnet-circuit bond plays. It measures acceptably at the levels a production test uses. The defect is latent, it is correlated with the batch, and by the time it surfaces you have shipped several months of it.
The fields to ask for when there is no data sheet
For most Chinese speaker-adhesive suppliers, and for this one specifically, there is no published technical data sheet to read. Shear strength on plated steel, service temperature, fixture time and the full-cure schedule for the magnet-circuit grade were not available at review, so treat every one of them as a question to put in writing rather than a number to assume.
| What to ask for | Why it decides something | How to phrase the request |
|---|---|---|
| Lap-shear strength on your plating, not on bare steel | Plating adhesion is the real ceiling | "Shear on zinc-plated CRS to sintered ferrite, both in as-received condition" |
| Shear retained after thermal ageing | Predicts the field-return failure, not the day-one one | "Shear after 1,000 h at your stated continuous service temperature" |
| Continuous and peak service temperature | The motor runs hot under continuous power | "Continuous rating, peak rating, and the duration the peak assumes" |
| Fixture time and full-cure schedule | Sets your line takt and oven loading | "Time to handling strength at 25 °C, and the oven schedule for full cure" |
| Recommended bond-line thickness | Decides whether you need gap-control features | "Minimum and maximum bond line, and whether the grade carries a spacer filler" |
| Cure shrinkage | Shrinkage moves alignment during cure | "Linear cure shrinkage, and any post-cure dimensional change" |
| Shelf life from date of manufacture | Decides what arrives usable | "Shelf life from manufacture date, storage temperature, and remaining life on arrival" |
Note also what a quotation does not tell you. Aosibo's stated commercial terms are a minimum order of 1,000 units, sampling in 7 and production in 15 — the two lead times given as bare numbers on its application and read as days. The application does not define what a unit is, and for adhesives a bottle, a syringe, a cartridge and a kilogram differ by orders of magnitude in value. Confirm the unit of measure and the pack size before you compare that quotation to anyone else's.
Common questions
Does the magnet's attraction mean I can use a weaker adhesive?
No, and it is the reverse of the right instinct. Attraction loads the joint in compression, which is the direction adhesives are strongest and least likely to fail. It contributes nothing against lateral shear, torsion or the slow creep that misaligns a gap. Specify against shear at temperature and after ageing.
Should the magnet-circuit adhesive match the other adhesives on the driver?
It does not have to be the same chemistry, but there is a practical argument for buying the set from one source. A driver line typically buys lead-wire, damper, surround, magnet-circuit and centring adhesives together rather than one at a time, and a supplier working across the whole bond schedule can tell you when one grade's cure schedule conflicts with another's temperature limit. The bonds interact through the oven whether the purchase orders do or not — the damper and spider joint in particular shares oven time with the magnet stack in many process flows.
What surface preparation should I specify?
At minimum, a defined degreasing step with a named cleaner, a defined maximum dwell between cleaning and bonding, and an incoming check on the plating. If you buy bonded magnet assemblies rather than bonding them yourself, ask for the cleaning step to be named in the process sheet you are shown, and ask what happens to parts that sit overnight between cleaning and bonding.
What to ask the supplier next
Before you commit a magnet-circuit bond to any adhesive supplier, send this list and treat the completeness of the reply as data in itself:
- The plating or coating specification on your yoke and top plate, and your magnet grade and coating, with a request for shear data on those exact surfaces.
- Fixture time at your shop temperature, and the full-cure schedule, stated as temperature and dwell.
- Continuous and peak service temperature, and the shear strength retained at continuous temperature.
- Recommended bond-line thickness, and whether the grade contains a gap-control filler.
- Cure shrinkage and any post-cure dimensional movement.
- Shelf life from date of manufacture, storage temperature, and a minimum remaining shelf life on arrival written as a contract term.
- The unit of measure behind the minimum order quantity, the pack size, and the price basis, so quotations are comparable.
- Whether a formulation change would be notified before shipment, and what batch documentation travels with each delivery.
- A sample large enough to bond coupons in your own plating and magnet material. Test in shear rather than tension, split the batch, age half at your expected continuous temperature, and compare — the gap between the two numbers is more useful than either alone.
Aosibo Adhesives is a useful example of what this looks like in practice: a Dongguan supplier whose range maps onto a driver bond by bond, including the magnet-circuit grade, with viscosity, cure schedule, pot life and substrate data described as shared at inquiry rather than published in advance. That is a normal arrangement for speaker adhesives, and it puts the burden on your inquiry. The profile at Aosibo Adhesives sets out what the company states about its range and terms and which items were not documented at review, which is the right starting point for the questions above.