Micro Speaker RFQ: Which Specs Actually Move Price and Lead Time
Published: 2026-09-09 | Use case: OEM engineers and program procurement who have assembled a micro speaker request for quotation and now need to understand which specification lines actually change the quoted number and the schedule — so they can decide where to hold a spec firmly, where to relax it, and which requests are worth paying for before the quote goes back out.
Not every line on a micro speaker RFQ moves the quote. Take Shenzhen Hongsheng Electronic Industry Co. LTD as an example: across a Hongsheng catalog sample of 70 driver models, changing the termination from solder to a lead-wire voice coil, or adding screw holes, generally sits inside an existing part family and carries no tooling, while asking for a footprint that does not exist anywhere in the family starts a new tool — and those two requests differ by an order of magnitude in both cost and calendar time. The difficulty is that they look similar on a specification sheet. This article sets out the five-rung customization ladder, what the buyer has to supply at each rung, which specifications sit at which rung, and where a tightly written tolerance or test condition quietly adds cost without adding performance.
1. Why Some Specs Move the Quote and Others Don't
Short answer: Because a specification either fits inside an existing part family or it does not. Requests that fit are configuration; requests that do not are tooling, and the two differ by an order of magnitude in both cost and schedule.
Two RFQ lines that look equally innocuous on a spec sheet can sit at opposite ends of the cost spectrum. Asking for a lead-wire voice coil instead of a solder terminal usually stays within an existing part family — the frame, the magnet and the diaphragm are unchanged and only the termination differs. Asking for a footprint that does not exist in any family starts a new tool, with the frame, the tooling and the qualification that follow. Both are one line on a form; one is configuration and the other is development.
The ladder below is a practical way to sort specification requests by consequence. The rungs are not a vendor's official price list — they are a way to ask 'does this request fit inside something that already exists?' before the quote goes out, and the answer predicts both the number and the calendar.
2. The Five-Rung Customization Ladder
Short answer: Stock part, interface and cosmetic, cavity and tuning, motor and materials, then new outline. The first three are usually configuration; the last one is tooling.
Each rung below lists what the buyer has to supply, what typically changes in the quote, and a real example from this catalog sample. Cost and schedule impact are described qualitatively and should be confirmed with the supplier, since they vary with the part, the program and the factory's loading.
Table 1: The five-rung micro speaker customization ladder, with what the buyer supplies and the evidence available at each rung.
Rung | What changes | What the buyer must supply | Typical impact on price / schedule | Catalog evidence |
L0 — Stock part | Nothing; an existing catalog part | Part number, quantity, schedule | Lowest cost, shortest schedule; the reference case | HS151125H, HS121722H, HS201623H, HS251233H |
L1 — Interface and cosmetic | Termination, lead length, connector, mounting holes, labelling | Termination type, lead length and gauge, hole pattern or a drawing | Generally no tooling; modest unit-cost and schedule effect | HS005017H and HS00663H ship with screw holes; HS003050H50 uses a lead-wire voice coil; HS402055H and HS352052H combine spring terminal with lead wire |
L2 — Cavity and tuning | Enclosure geometry, porting, damping, target in-box F0 | Rear cavity volume, front cavity depth, target in-box F0, gasket method | Usually no new tool for a box module; adds engineering and validation time | The 1217 BOX family spans 850 Hz to 1350 Hz F0 on geometry alone at the same electrical and acoustic class |
L3 — Motor and materials | Magnet count and grade, diaphragm material, surround material, voice coil | Target SPL or F0 improvement, material constraints, lifetime expectation | Component and process change; can add qualification time | HS241540H42 to HS241534H34 adds a second magnet for +2 dB at 0.6 mm less height; HS-BX-1217-VV30LT uses a five-magnet motor |
L4 — New outline | Frame, footprint, or a geometry that does not exist in any family | Full mechanical drawing, target specs, volume to amortise the tool | Tooling cost and the longest schedule; the only rung where a miss is expensive | No catalog equivalent — the request defines the part |
The single most useful question a program can ask before finalising an RFQ is which rung each specification line sits at. Lines at L0 through L2 can usually be iterated cheaply and late; a line at L4 cannot, and discovering late that a request was L4 rather than L1 is the most common reason a quotation cycle overruns.
3. L0 Stock Parts: What Minimal Information Is Enough
Short answer: A part number, a quantity and a schedule. Everything else on the RFQ at this rung is for the buyer's own benefit — confirming the part fits the application — not for the vendor's pricing.
For a stock part the vendor needs three things: which part, how many, and when. The eight information blocks from the companion article on RFQ content still matter, but at this rung they are there to confirm the part is right for the application rather than to build a price. A program buying a stock part is buying against the datasheet, and the datasheet already fixes the cavity, the test basis and the materials.
The trap at L0 is accidentally leaving it. A buyer who specifies a stock part number and then adds 'but we need it in a 1 cc cavity' or 'with a two-pin connector' has moved the request up the ladder without changing the part number, and the quote will come back higher than the stock price suggested. Where a program needs something the stock part does not have, say so in the request rather than in an attached condition — the vendor can then say which rung the real request sits at.
4. L1 Interface and Cosmetic: Termination and Mounting
Short answer: Termination, lead length, connectors and mounting holes generally carry no tooling. They are the cheapest requests on the RFQ and the ones most worth specifying precisely.
Interface requests are the best value on the whole ladder: they usually need no new tool, they remove an assembly step on the buyer's line, and they are frequently free to specify. This catalog sample shows the range that exists inside existing families. Termination appears as solder type (the majority of parts, including HS121722H, HS201623H and HS251233H), leaf-spring contacts (HS151125H, HS150827H, HS151130H), lead-wire voice coils (HS003050H50, HS001846H), and spring terminals combined with lead wires (HS402055H, HS352052H, both also IP68). Mounting features appear the same way: HS005017H and HS00663H ship with screw holes as standard, and the BOX family includes variants with two and four screw holes (HS-BX-703314H, HS-BX-703017H) alongside versions without.
What the buyer has to supply at this rung is specific rather than general: the termination type, the lead length and gauge where leads are involved, the connector part number where one is used, and either a hole pattern or a drawing where mounting features are needed. Vague requests cost time here, because 'we need a connector' invites three questions back while 'we need a JST SHR-02V-S-B on a 60 mm lead' can usually be answered yes or no immediately.
5. L2 Cavity and Tuning: What the Buyer Must Specify
Short answer: The rear cavity volume, the front cavity depth, the target in-box F0, and how the assembly seals. Cavity requests usually add engineering and validation time rather than tooling.
Cavity and tuning requests sit one rung above interface work because they change the acoustic result and need validation, even though they usually do not need a new tool. What the vendor needs is the rear cavity volume the host will supply, the front depth from the driver face to the grille, the target in-box F0, and the sealing method. Where the program does not yet know the cavity, a stated range is far more useful than silence.
The 1217 BOX family is the clearest evidence of what this rung can do. Every variant runs the same 8 ohm / 1.0 W / 95 dB electrical and acoustic class, and the published F0 moves across 850 Hz to 1350 Hz purely on enclosure geometry — HS-BX-1217-3813X and HS-BX-1217-X10 at 850 Hz, HS-BX-2030 at 950 Hz, HS-BX-1915 at 1150 Hz, HS-BX-1217-T26 at 1350 Hz. No driver change is involved. For a program whose only open question is the in-box F0, that is a tuning request rather than a development one, and specifying the target in-box F0 on the RFQ is what lets the vendor answer it.
One physical point is worth stating plainly, because getting it wrong on an RFQ produces a quote for the wrong part. A sealed cavity adds acoustic stiffness rather than compliance, so a smaller host volume pushes the system resonance upward: Fc = Fs × √(1 + Vas / Vb). A program that asks for 'a lower F0' without stating the cavity it can supply is asking for two things at once, and the vendor can only quote one of them.
6. L3 Motor and Materials: Magnet, Diaphragm and Surround
Short answer: Magnet count and grade, diaphragm material and surround material change components and process. They add capability and, where qualification is required, schedule.
Motor and material requests change what the part is made of, which puts them above cavity work on the ladder. They are still usually within an existing frame and family, so they are not tooling, but they do change the bill of materials and the process, and where the program needs lifetime or environmental qualification they add validation time.
The magnet lever is well documented in this catalog sample. HS241540H42 (24 × 15 × 4.0 mm) and HS241534H34 (24 × 15 × 3.4 mm) share a footprint and differ by motor: the dual-magnet part reaches 95 dB against 93 dB and is 0.6 mm thinner. HS-BX-1217-VV30LT uses a five-magnet motor to reach 97 dB and 800 Hz against the family's 95 dB. HS402055H and HS352052H add a secondary magnet alongside IP68 construction. Where a program is tooled and needs more output, this is often the only lever left, and it belongs on the RFQ as a target rather than as a part number.
The material lever works the same way and is usually driven by lifetime or by tone. Cloth edge with a paper diaphragm appears on HS-BX-203008H, HS003021H and HS003058H; foam edge on HS002628H28, HS0028110H110 and HS-BX-282813H; PU on HS0023123H123. The right way to specify this rung is to state the outcome the program needs — a lifetime expectation, a temperature range, a tone preference — and let the vendor propose the material, rather than naming a material whose acoustic consequence the program has not modelled.
7. L4 New Outline: When Tooling Is Unavoidable
Short answer: A footprint or geometry that does not exist in any family starts a new tool. This is the only rung where a mistake is expensive, and it should be entered deliberately and with the volume to amortise it.
A new outline is the one rung where the answer to 'does this fit inside something that already exists' is no. The frame, the tooling and the qualification all have to be created, and the schedule is set by tooling and validation rather than by assembly. Two practical consequences follow. First, the request should be entered deliberately, after the L0 through L3 options have been ruled out for a stated reason — a mechanical envelope that no family meets, or a performance target that no stock part reaches. Second, the RFQ should carry the volume the program expects, because tooling is amortised across it and a vendor cannot price a tooled part without knowing the quantity.
Where a program is unsure whether its request is L3 or L4, asking is cheap and assuming is not. A vendor can usually say quickly whether a target footprint exists inside a family or just outside it, and that one question can move a request down a rung. The 1217 family again illustrates the point: a program needing a 38 × 18 × 3.5 mm side-fire part with a 850 Hz F0 is asking for an existing L2 variant, while a program needing 42 × 18 × 3.5 mm with the same acoustics is asking for L4.
8. Tolerance and Test-Condition Requests That Add Cost
Short answer: Tighter tolerances, extra test conditions and bespoke qualification add cost without necessarily adding performance. Specify the tolerance the application actually needs, and ask what a looser one saves.
Some of the most expensive lines on an RFQ are not requests for a different part — they are requests for a tighter number or an extra test on the same part. Three categories are worth reviewing before the request goes out. First, tolerance: specifying a sensitivity window tighter than the application needs raises sorting cost and yield loss for no audible benefit. Second, test conditions: asking for every unit to be measured in the production cavity rather than on a standard jig adds test time per unit, and is only worth it where the in-box result genuinely gates the product. Third, bespoke qualification: a custom thermal cycle, a custom salt-spray duration or a custom life test is a real line item, and it is worth asking what standard test it replaces.
The useful habit is to mark each tolerance and test line on the RFQ with the reason it is there. Where the reason is 'the datasheet said so' rather than 'the application needs it', the line is a candidate for relaxation, and asking the vendor what a looser specification saves is a reasonable question that most buyers never ask.
9. Project Case: Downgrading a Request From L4 to L2
Short answer: A control-panel program was about to request a new outline for a cavity target that an existing 1217 family variant already met. Stating the target instead of the geometry moved the request down two rungs.
Project snapshot
A wall-mounted control panel program came to Hongsheng with a mechanical drawing for a 39 × 19 × 3.6 mm side-fire enclosure and an 870 Hz in-box F0 target, expecting a new tool because the geometry sat just outside every family the team had reviewed. The Hongsheng engineering review read the acoustic target rather than the drawing: HS-BX-1217-3813X (38 × 18 × 3.5 mm, 8 ohm, 1.0 W, 95 dB) already published 850 Hz on a side-fire enclosure, and the 1 mm of outline difference was in a non-acoustic dimension that the host bezel could absorb. The request moved from a new outline to a cavity-and-tuning variant with a slightly revised bezel, and the program kept both the schedule and the tooling budget. The lesson generalises: stating the acoustic target alongside the drawing lets the vendor say whether the geometry or only the drawing is doing the work.
10. FAQ on Specs That Move a Micro Speaker Quote
Short answer: Whether a footprint change means tooling, what a magnet upgrade costs, how tight a tolerance to specify, and what to do when the quote comes back higher than expected.
Q1. Does changing the footprint always mean new tooling?
A1. Not always, but usually. Small changes to a non-acoustic dimension can often be absorbed by the host bezel or by an existing family variant, while a change to the frame, the magnet pocket or the diaphragm area generally cannot. Ask the vendor which it is before assuming — the 1217 family covers a wide geometry range on one driver, and a request that looks new is sometimes a variant.
Q2. Is a lead-wire termination more expensive than a solder type?
A2. It is usually a modest unit-cost difference rather than a tooling one, since the frame and motor are unchanged. This catalog sample carries both inside existing families — HS003050H50 uses a lead-wire voice coil, while HS003050H is the comparable solder type. Confirm the specific delta with the supplier, as it depends on the part and the lead specification.
Q3. How much does a magnet upgrade add?
A3. In this catalog sample a dual-magnet version of the same footprint (HS241540H42 to HS241534H34) gains roughly 2 dB and 0.6 mm of height for the same outline. The unit-cost delta depends on magnet grade, quantity and market pricing, so ask the supplier rather than estimating. The point for the RFQ is that a magnet change is a component change, not a tool.
Q4. Should I specify the tightest tolerance I can?
A4. Generally not. A tolerance tighter than the application needs raises sorting cost and yield loss without an audible benefit. Specify the tolerance the product actually requires, and where the requirement is uncertain, ask the vendor what a looser specification saves before committing.
Q5. Does asking for in-box testing add much cost?
A5. It adds test time per unit, so it scales with quantity. It is worth it where the in-box result genuinely gates the product — a voice-prompt level in the production cavity, for example — and less worth it where a standard jig measurement is sufficient. Ask for both options and compare.
Q6. What is the single most expensive line I could put on an RFQ?
A6. A new outline that turns out not to be necessary. It is the only rung where tooling, qualification and schedule all move together, and it is the one where an early conversation with the vendor is cheapest. State the acoustic target alongside any drawing so the vendor can say whether the geometry is doing the work.
Q7. Can I get a stock part with a different cavity?
A7. The cavity is usually a host responsibility rather than a part property for bare drivers, so the same part can be used in different cavities — with a different in-box F0. For box modules the cavity is integral and cannot be changed without changing the module. State which you have, since the two answer the question differently.
More in This Series — Micro Speaker Quotation
This article is the second of a three-part technical series on micro speaker quotations. The first covers what information to provide; the third covers how to compare quotes that come back differently.
· Part 1 — What Information Should You Provide When Requesting a Micro Speaker Quotation? →https://www.hsdz-spk.com/news/535.html
· Part 3 — Micro Speaker Quotation FAQ: Why Quotes Vary and How to Compare Them → https://www.hsdz-spk.com/news/537.html
11. Closing Notes on Spec Sensitivity
Specification lines on a micro speaker RFQ sort into five rungs, and the rung predicts the consequence better than the wording does. Interface and cosmetic requests (termination, leads, mounting holes) usually carry no tooling. Cavity and tuning requests add engineering and validation time but generally no tool. Motor and material requests change components and process, and can add qualification. A new outline is tooling, and it is the only rung where a mistake is expensive. Tolerances and bespoke test conditions add cost without necessarily adding performance, and each is worth marking with the reason it is there.
For a program finalising a request, the highest-value step is to state the outcome the application needs alongside the geometry the drawing shows, so the vendor can say where the request actually sits. The third article in this series takes the next problem: what to do when three quotes come back anyway, priced against different assumptions, and how to put them on a comparable basis before the decision.