Your driver line runs to a 40-second takt time. The tunnel oven is already at full tray density for the surround bond, and the engineering change on the new model adds a centring adhesive whose cure schedule wants 30 minutes at 80 °C. Nobody is going to sign off a second oven this quarter. So the adhesive choice stops being a chemistry question and becomes a throughput question — and the buyer who understands that will get a workable quote out of the supplier while the buyer who does not will get a datasheet that does not fit the plant.
This is where most speaker adhesive sourcing goes sideways. The specification travels in the wrong direction. A buyer describes the bond, the supplier recommends a grade, and the cure schedule arrives at the end as a fact of nature. Aosibo Adhesives, a Dongguan supplier whose range is built around the loudspeaker driver, states the sequence the other way round: send the two substrates and the cure window you have available. That instruction is worth more than it looks. Leading with the constraint is what turns a catalogue conversation into a specification conversation.
The arithmetic nobody does before the RFQ
Takt time and cure time do not have to match. What has to match is the amount of work-in-progress your floor can physically hold while the bond fixtures.
The relationship is one line of arithmetic:
Units held in buffer = fixture time ÷ takt time
At a 40-second takt with a 20-minute fixture time, that is 1,200 seconds ÷ 40 seconds = 30 units sitting somewhere between the dispense station and the next operation. Thirty is trivial — a rack. At a 20-second takt with a 4-hour ambient fixture, it is 720 units, which is a room. Same adhesive, same bond, completely different capital decision.
Run the same calculation for full cure if any downstream operation needs full strength — final assembly torque, a press-fit, a drop test — because that buffer sits on top of the first one. If final QC can be done after fixture and before full cure, say so explicitly in the specification, because it removes an entire storage requirement.
Then the oven, which has its own two numbers:
Tunnel throughput = tray capacity ÷ dwell time, where dwell = heated length ÷ belt speed
A tunnel with 6 metres of heated length running at 0.5 m/min gives 12 minutes of dwell. If your cure needs 30 minutes at that temperature, you either slow the belt to 0.2 m/min — which cuts throughput by more than half for every product on that oven, not just the new one — or you find a different cure route. Batch ovens have the same trade in a different shape: cycle time is dwell plus load plus unload plus the ramp back up to temperature after the door opens, and that last term is the one that gets forgotten.
Do this arithmetic before you send the RFQ, not after the samples arrive. It converts "what cure does this need?" into "here is the cure window I have — what fits in it?", which is a question a competent adhesive supplier can actually answer.
Four ways out when the oven is full
| Route | Typical fixture behaviour | Equipment needed | WIP consequence | What to ask the supplier |
|---|---|---|---|---|
| Hotter, shorter oven schedule on the same grade | Shorter dwell at higher temperature | Existing oven, if it reaches the temperature | Unchanged buffer, more oven throughput | The maximum temperature the grade tolerates, and whether properties change at the top of the range |
| Ambient-cure or room-temperature grade | Longer fixture, no oven at all | Racking, trays, floor space, FIFO discipline | Large buffer, zero oven load | Fixture time at your actual plant temperature, not 23 °C |
| Two-part system with an accelerated hardener | Fixture in minutes, full cure over hours | Metering or cartridge dispensing | Small buffer, new metering discipline | Mix ratio basis, pot life at your mixed mass and temperature |
| Different cure chemistry entirely for that bond | Seconds to minutes | Depends entirely on the chemistry | Potentially near-zero buffer | Whether such a grade exists in the range at all, and what it costs in substrate compatibility |
The constraint that quietly kills route one is the substrate, not the adhesive. A driver is a stack of heat-sensitive materials: treated paper cone, foam or rubber surround, adhesive already cured on an earlier station, and on micro-drivers for headphones and earphones, components where cure temperature and outgassing both matter. Raising the oven from 80 °C to 120 °C to halve the dwell is a fine idea until the surround takes a set or a previously cured bond passes its glass transition and creeps under the fixture load. Before you ask the adhesive supplier what temperature the glue can take, establish what temperature the assembly can take. That number is yours to supply, and it belongs in the RFQ.
Route two is the one buyers underestimate. Ambient cure looks free because it uses no energy, but 720 units of buffer is floor space you are not currently paying for, plus a first-in-first-out discipline that has to survive a shift change. The failure mode is not technical — it is a rack that gets pulled forward by an operator under schedule pressure, sending under-fixtured drivers to the next station. If you go this route, the counter-measure is a physical one: time-stamped rack cards, or lanes sized so that the only rack you can physically reach is the oldest one.
Where the range's own structure helps
Aosibo lists its centring adhesive in one-part, epoxy and halogen-free versions, and describes the reason in exactly the terms this article is about: the same bond can be specified to a customer's cure schedule or material restriction. That is a useful thing to find in a supplier's range, because it means the cure-window conversation has somewhere to go. If the one-part version wants an oven you do not have, there is a second chemistry for the same joint to evaluate rather than a dead end.
The same logic applies across the bond schedule. The core range maps onto a driver bond by bond — lead-wire, damper (spider), water-based foam and rubber surround, magnet-circuit and centring — and a speaker line usually buys these as a matched set rather than one at a time. Buying the set from one source is not just a commercial convenience; it is what lets you sequence the cure schedules against each other. Two grades that each want an oven pass are a scheduling problem. Two grades where one fixtures at ambient while the other takes the oven slot are a line that runs.
Note what is not available to you here: no technical data sheet was published for any of these grades at review, and the company states that cure schedule and related data are shared at inquiry. So the ambient-versus-oven options per grade, with their temperature and dwell, are something you have to extract in writing — the six fields to demand are set out in reading a speaker adhesive TDS. Do not proceed on a verbal answer for a cure schedule. It is the number your line is built around.
Costing the three routes against each other
Once the technical options are on the table, the decision is a cost comparison with three different units, which is why it usually gets decided by whoever shouts loudest rather than by arithmetic. Put them in the same currency.
- Line seconds. If the cure route forces a slower belt on a shared oven, the cost is not the new product's seconds — it is every product on that oven. Compute the throughput loss across the whole oven load, annualised. This is usually the biggest number on the page, and usually the one nobody computed.
- Floor space and racking. Buffer units × footprint per unit × the cost per square metre of the space you would otherwise use, plus the rack capital, plus the labour of moving trays. Ambient cure moves cost here.
- Energy and load. Oven kWh per cured unit, plus the cost of holding a batch oven at temperature through a changeover. Smaller than the first two in most plants, but it is the one that scales directly with volume.
- Scrap risk. Two-part metering adds a scrap mode that ambient one-part cure does not have; ambient racking adds a FIFO-violation mode that an oven pass does not have. Price each at your realistic defect rate, not at zero.
Three insider details that change these numbers more than people expect:
- A tunnel oven's stated heated length is not its effective dwell length. The first and last portions of the tunnel are ramp zones where the part is not yet at cure temperature. Effective dwell is measured with a thermocouple on an actual part travelling the belt — a profile run — not read off the oven nameplate. Lines routinely discover they have twenty per cent less dwell than they thought, and the discovery usually arrives as a field-return wave rather than as a measurement.
- Fixture time and full cure scale differently with temperature. Raising temperature buys you a lot on fixture time and proportionally less on the last stretch toward full properties. A schedule tuned to hit fixture faster can leave you shipping product that is fixtured but not fully cured, which is fine if nothing downstream loads the joint and expensive if something does.
- Ambient cure is not temperature-independent. "Room temperature cure" in a specification means whatever your room actually is. A plant that runs at 18 °C in January and 33 °C in July has a fixture time that moves by a factor of two or more across the year, and a buffer sized for summer will overflow in winter. Ask for the ambient fixture time at two temperatures, and size the racking for the cold one.
Turning this into an RFQ the supplier can answer
Send the constraint first. A cure-window specification is one page and it changes the quality of every reply you get.
State, per bond:
- Takt time of the line the bond sits on, in seconds.
- Available cure equipment: batch oven or tunnel; heated length and belt speed range, or chamber volume and cycle overhead; maximum stable temperature.
- Maximum temperature the assembly tolerates, and for how long — set by the most heat-sensitive component in the stack, not by the adhesive.
- Buffer you can physically hold, in units, between the dispense station and the next operation — and separately, before full-strength operations if any exist.
- Both substrates by name, with any surface treatment.
- Plant temperature and humidity range across the year, if the line is not climate-controlled.
- The question, stated plainly: which grades in your range fixture inside this window on these substrates, and what is the full-cure schedule for each?
Then ask for the cure data in writing: ambient route and oven route per grade, each with temperature and dwell, plus the fixture-time threshold used. And ask whether any grade in the range cures by a mechanism other than heat or time, because if one does it reshapes the whole calculation — that is a question worth asking directly rather than assuming either answer.
Aosibo's stated commercial terms give sampling at 7 days and production at 15 days, both company-stated and both worth confirming in writing. Plan the cure trial into that sample window deliberately: a sample cycle that only proves the material bonds has wasted the most valuable week of the evaluation. Run the fixture-time measurement at your own plant temperature, run a profile on the oven with a thermocouple on a real part, and run an aged set. The framing in how to request a sample from a Chinese factory applies directly — define the acceptance criteria before the sample ships.
One commercial detail belongs in the same message. The stated MOQ is 1,000 units, and the source does not define whether a unit is a bottle, a cartridge, a syringe or a kilogram — which, when your consumption is grams per driver, is the difference between a month of production and several years of it. Confirm the unit of measure, pack size and price basis alongside the cure data; the general discipline for pinning down MOQ and lead-time data from a Chinese factory is the right one here. The Aosibo factory profile on ChinaMakersHub lists the full bond schedule these questions apply to, including the damper (spider) adhesive on the suspension joint, which is usually the bond where the cure-window conflict shows up first.
Common questions
Can I just run the adhesive hotter to fit my existing oven dwell?
Only if two things hold: the adhesive's properties survive the higher temperature, and the assembly does. The second is the binding constraint on a driver, because the cone, surround and any previously cured bonds are all in the oven with it. Ask the supplier for the maximum recommended cure temperature and what changes above it, then check that against the most heat-sensitive item in your stack. If the margin is thin, a hotter schedule buys throughput and spends reliability.
How do I compare a fixture time from one supplier against another?
Only after you know the strength threshold behind each. Fixture time is defined as the point where the joint reaches a nominal handling strength, but suppliers do not all use the same threshold, so a shorter quoted fixture time can simply mean a looser definition. Ask each supplier for the threshold value and the substrate pair, then compare. If one supplier cannot tell you, treat their number as indicative rather than comparable.
What should I measure during the sample cycle to validate the cure window?
Three things. First, an oven profile with a thermocouple attached to a real part, to establish effective dwell at cure temperature rather than nameplate dwell. Second, fixture time measured at your own plant temperature on your own substrate pair, with the threshold you will use in production. Third, a set aged under heat and humidity and tested against a fresh set, because a schedule that reaches fixture but not full cure will pass a same-day test and fail months later. Those three results are what let you size the buffer and the oven slot with numbers instead of assumptions.