Damper (Spider) Adhesive: Bonding the Suspension Joint

Damper spider adhesive for loudspeaker suspension bonding is the least glamorous line item on a driver's bill of materials and the one that decides whether your Thiele-Small parameters hold their spread. The spider is the part that positions the voice coil in a magnetic gap frequently 0.3 to 0.6 mm wide, with a couple of tenths of a millimetre of clearance per side. Two adhesive joints hold that geometry: the inner diameter to the former, and the outer diameter to the frame landing. Neither joint is asked to be strong in the way a structural bond is strong. Both are asked to be dimensionally honest, at temperature, for years.

Get this bond wrong and the symptoms do not look like adhesive problems. They look like a resonance frequency drifting across a production lot, a rub that appears only above two-thirds of rated excursion, or drivers that measure clean at the plant and buzz after a container crossing. All three trace back to glue often enough to check it first.

Two joints, two completely different problems

Plants that buy one adhesive and use it at both ends usually get away with it, then cannot explain why one end fails and the other does not. The joints load differently, sit at different temperatures and see different substrates.

The inner-diameter joint wraps the spider's centre hole onto the voice-coil former. It is in direct thermal contact with the part of the driver that gets hottest after the coil itself, because the former conducts coil heat straight into it. Mechanically it is loaded in cleavage and peel at the excursion extremes, not in shear — the spider tries to fold away from the former at the top and bottom of the stroke. A lap-shear figure measured on steel coupons predicts almost nothing about this joint.

The outer-diameter joint bonds the spider's flat rim to a landing on the frame, which is either a machined shelf on a cast basket or a stamped shelf on a pressed-steel one. It runs cooler and is loaded closer to pure shear, but it has a geometry problem the ID joint does not: the landing may not be flat. A stamped steel basket landing can carry a couple of tenths of a millimetre of waviness, and the adhesive is the only thing available to fill it.

The glue sets the compliance, not just the joint

Here is the detail that costs driver plants real money and gets blamed on everything else. The spider's compliance comes from its corrugations flexing. The adhesive at the inner diameter defines where the flexible region starts — because any corrugation the adhesive has wicked into is no longer flexing.

Put a slightly generous bead on the ID and resin climbs the first corrugation: the effective working radius shrinks, the suspension gets stiffer, resonance frequency goes up. Put a lean bead on the next unit and Fs comes back down. Nothing in the drawing changed, nothing in the fabric changed, and the end-of-line parameter test still shows a spread wide enough to fail your customer's incoming tolerance.

If your Fs distribution is bimodal or drifting through a shift, look at the dispenser and the adhesive's viscosity before you look at the spider supplier. The fabric is porous by design — phenolic-impregnated cotton in particular absorbs adhesive readily — so viscosity, solids content and open time on this grade are compliance parameters, not just handling parameters. Ask for them in writing, along with the recommended bead volume for your former diameter, and check them on incoming lots.

Cure shrinkage moves the coil before the driver ever plays

The second one. An adhesive that loses a carrier as it cures — water, solvent — shrinks. Solids content tells you roughly how much material stays behind. If the spider is clamped in a jig while that shrinkage happens, the rim is held and the shrinkage goes somewhere else: into doming or dishing of the spider, which shifts the coil's rest position in the gap by a few tenths of a millimetre.

That offset does not stop the driver working; it makes it asymmetric. The coil sits off-centre in the gap axially, the BL curve is no longer symmetric about rest, and you get second-harmonic distortion rising with level — an audible hardness the listening panel calls "boxy." Meanwhile the driver passes every impedance sweep you run.

Two practical defences: ask for cure shrinkage and solids content as numbers, and check coil rest position on cured samples rather than on jig-clamped ones. If a supplier has not been asked this before, that in itself tells you something about whether they have worked with driver lines.

Rocking modes come from the OD joint

Third detail. If the OD bond makes full contact around three-quarters of the circumference and partial contact on the rest — because the stamped landing is wavy and the adhesive could not bridge the gap — the suspension is stiffer on one side than the other. The moving assembly then develops a rocking mode: a tilting resonance, usually somewhere in the low hundreds of hertz, where the coil swings rather than pistons.

At low level nothing happens. At high excursion the coil touches the pole or top plate on one side and you get an intermittent rub that appears only at power — the classic "passed at the factory, buzzes at the customer" failure, and a gap-fill problem. What you need is enough body to bridge worst-case landing flatness at your available clamp pressure, and enough green strength to hold it there while it cures. Measure flatness on twenty baskets before specifying anything, then give the supplier the number.

Why 23 °C qualification is not qualification

Suspension bonds are almost always qualified cold — a pull test on a cured sample at room temperature, a pass, and into production. Then the driver runs at rated power, the coil heats, the former conducts that heat into the ID joint, and the adhesive crosses its glass transition.

Above that transition the modulus falls steeply and creep governs. The spider stops being rigidly located on the former and migrates under the peel loading of each excursion cycle. In the field: rub after extended high-level use, or a resonance frequency permanently lower when a returned unit is retested. The bond never let go in a way a pull test would catch — it crept.

So the numbers to ask for on the damper grade are the ones nobody volunteers: glass transition or continuous service temperature, and a hot creep result under representative load at that temperature. For micro-drivers — earphone and TWS assemblies, where cure temperature and outgassing both matter because the parts are small and heat-sensitive — you also want the lowest cure temperature the grade will tolerate, and condensable-volatiles data if the driver is in a sealed cavity.

None of that is published for this range. Aosibo Adhesives, a Dongguan supplier whose catalogue is written bond by bond onto a driver — lead-wire, damper, surround, magnet-circuit and centring grades — states that viscosity, cure schedule, pot life and substrate data are shared at enquiry rather than in a public datasheet, and asks buyers to send both substrates and the available cure window with the enquiry. That is a workable process, and it puts the burden where it belongs: on you to define the joint precisely, and on them to answer with figures you can put in a contract.

Cure schedule, line takt and qualification

The spider bond usually sits early in the build sequence, which means everything downstream waits on it. Three arrangements are common, and they cost different amounts:

• Ambient cure with a fixture. Cheapest in capital, expensive in floor space and work-in-progress. You need enough jigs to cover the full fixture time at your takt rate — at a 40-second takt and a 30-minute fixture time, that is roughly 45 jigs in circulation just for this station.
• Oven cure. Fast and repeatable, but the oven becomes the constraint on the whole line, and the spider is competing for that oven with the centring and magnet bonds. Ask whether cure schedules across the range can share a single pass, because two ovens is a capital decision.
• Two-stage: fixture tack, then batch cure. Common on high-volume driver lines. The bond reaches handling strength in seconds, the assembly indexes on, and full cure happens in a batch oven or on a rack. This needs the supplier to state both times separately — most quote only the full cure.

This is one reason a supplier offering the same bond in more than one chemistry is useful rather than just confusing. Aosibo lists its centring adhesive in one-part, epoxy and halogen-free versions specifically so the same joint can be matched to a customer's cure schedule or material restriction. The right question to bring to a spider bond is the same: what does the range offer for the cure window I have, at the temperature I can afford?

Bear in mind that adhesive quotes for the same nominal bond come back at wildly different prices for exactly this reason — one supplier is quoting a fast-cure system and another an overnight ambient one. The reasons quotes for the same spec differ apply as much to a drum of adhesive as to a moulded part, and the fix is the same: specify the cure window before you ask for the price.

Qualifying a spider adhesive: a runnable plan

With sampling stated at 7 days and production at 15 on this listing, a two-to-three week evaluation fits comfortably inside the gap before you commit — and running it properly is what keeps the sample-to-mass-production step from becoming the place your schedule collapses.

1. Incoming characterisation. Viscosity at 25 °C, solids content by oven-dry, appearance, and date of manufacture on the pack. Record these; they are your baseline for every future lot.
2. Wick and bead study. Bond twenty spiders at the ID with your production dispenser settings. Section five and measure how far resin climbed into the first corrugation. Vary bead volume by ±20 % and record what Fs does. That curve tells you your process window.
3. Free-air parameter spread. Build thirty units. Measure Fs, Qts and Cms. A tight distribution here is the single best evidence the adhesive is behaving.
4. Coil rest position after cure. Release from the jig, then measure coil position in the gap. Compare against the design intent — this catches cure shrinkage.
5. Landing contact check. Section OD joints on the worst three baskets in your flatness sample. Look for voids at the wavy points.
6. Hot creep. Hold assemblies at your measured former temperature under representative load for a stated dwell, then remeasure parameters and check rest position.
7. Damp-heat ageing, per your customer's spec, followed by a rub-and-buzz sweep at full excursion.
8. Line trial. A short run at real takt with real fixtures, ending with the same parameter sweep, because handling behaviour on a bench never matches handling behaviour under pressure.

Common questions

Can I use one adhesive for both spider joints?

Frequently yes, and plenty of lines do. But specify against the harder of the two joints — the inner diameter, because of temperature and cleavage loading — and then confirm the same grade gives you the gap-fill you need at the outer diameter on your actual frame finish. A grade that satisfies the ID and cannot bridge a wavy stamped landing will hand you rocking modes.

What should the supplier tell me about substrates?

Name every material in the joint and expect a compatibility answer for each pair. Phenolic-impregnated cotton, aramid non-woven, Kapton, phenolic and aluminium formers, and painted, plated or bare frame landings all behave differently — Kapton in particular is a difficult surface for many chemistries and may need a specific grade or a surface preparation step. If the answer to a substrate question is a general reassurance rather than a named grade, ask again in writing.

Is a supplier who specialises in speaker adhesives actually better than a large general chemical house?

They are answering a narrower question, which cuts both ways. A supplier whose range exists to bond leads, spider, surround, magnet stack and centring already knows what a spider bond has to survive, and that is a different conversation from a trader selling one general-purpose adhesive across ten industries. What specialisation does not substitute for is documentation: you still need per-grade figures, a safety data sheet, shelf life from date of manufacture, and a formulation-change notification commitment in writing.

What to ask the supplier next

• What is the recommended bead volume for my former diameter, and what viscosity and solids content back it up?
• Cure shrinkage of the damper grade, as a number.
• Glass transition or continuous service temperature, plus any hot creep result you have.
• Peel or cleavage performance on my actual substrate pair — not lap shear on steel.
• Fixture time and full cure time, stated separately, with the ambient option and the oven option side by side.
• Can the damper, centring and magnet-circuit grades share one oven pass, and at what schedule?
• Gap-fill capability at my measured landing flatness and available clamp pressure.
• Substrate compatibility, named material by named material, including the frame coating.
• Shelf life from date of manufacture, storage temperature, and whether the batch date is printed on each pack.
• Safety data sheet in the destination language, plus UN number, class and packing group for each grade quoted.

Aosibo Adhesives is a useful example to test that list against: a Dongguan house registered in 2011 whose product range is organised as a driver's bond schedule rather than a general catalogue, with sampling stated at 7 days and production at 15. Its factory profile separates what the company stated on its application from what was not verified at review — which is exactly the distinction you should be drawing yourself before a suspension bond enters your bill of materials.