88 智能音箱外壳用什么改性尼龙
智能音箱的工况不像外壳
智能音箱外壳看着是塑料壳，实际工况非常密集：长期通电（内有电路板散热）、WiFi/蓝牙射频、触摸按键磨损、外观高光、阻燃 UL94 V-0、超声波焊接强度。
五个要求同时成立，选料失误一个，整批报废。智能音箱外壳不是家电外壳——它同时是结构件 + 外观件 + 散热件 + 屏蔽件四件合一。
现场还原
去年一月，深圳一家方案商的旗舰音箱在量产前两周出了问题。外壳喷完漆上网罩一装，开孔边缘全是毛刺，良率掉到六成。项目组在楼道里开会，结构负责人的原话是，样机的时候明明没问题。
对比三批料才发现，量产料的熔指比样机料高了一档，流动快了，薄边处的分子取向变了，喷漆一盖，毛刺全暴露。最后不是改料，是把网罩开孔的模具倒角修了一轮，再把料号锁死。
音箱这种产品，外观件的良率就是利润率，材料批次的漂移要在打样阶段就摁住。
外观 + 阻燃的双重要求
智能音箱最常见的失效是阻燃剂析出导致外观发白。普通阻燃 PA66 加溴系阻燃剂后 3 个月表面析出，外观报废。
必须用无卤阻燃 + 表面光泽协同配方。无卤阻燃 PA66-GF15 + 高光泽改性是主流方案。红磷阻燃剂析出较少，也是常用方案——但红磷有颜色限制，只能做深色外观。
超声波焊接的边界
智能音箱上下壳超声波焊接是装配主流工艺。焊接强度对玻纤含量敏感——GF15 焊接强度最佳，GF30 焊接强度下降 40%。
选 GF15 不要选 GF30——这是外观件和结构件的差别。焊接强度测试要看 ASTM D1002 标准的搭接剪切强度，目标 > 30 MPa。
射频透波的特殊性
智能音箱有 WiFi/蓝牙天线，外壳不能含碳黑或金属填料——会屏蔽射频信号。必须本色或浅色配方，不能用导电/抗静电填料。
介电损耗 < 0.02 是透波硬指标——含碳黑的 PA66 介电损耗 > 0.15，完全屏蔽 WiFi。
触摸按键的耐磨
智能音箱顶部触摸按键区域要耐磨——手指 1 万次摩擦后不能磨损。必须加耐磨涂层或表面硬化处理。
PA66-GF15 + 表面硬化涂层是当前主流。涂层厚度 5-10 μm——太薄耐磨不够，太厚影响触摸灵敏度。
延伸判断：智能音箱的隐性变量
智能音箱外壳有三件容易漏掉的隐性变量。一是模具温度——PA66-GF15 的模温要求 80-100℃ 比通用 PA66 高 20℃，模温不够表面会出现玻纤裸露。
二是湿度管控——PA66 粒子注塑前必须 80℃ 烘干 4 小时，没烘干的粒子表面会有银纹和气泡。三是配色稳定性——多批次生产时配色 ΔE < 1.0 是合格线，> 2.0 客户会投诉批次色差。
深一层：几个数字的来历
阻燃是送检的强制项。音箱接市电，壳体要求V-0，阻燃体系的填料会拉低表面光泽。外观件和结构件分模，是这类产品模具设计的默认动作。一套模具多花的钱，远抵不上外观返工的损失，这笔账在开模前就该算清。
射频透波要提前对齐。2.4吉赫和5吉赫的天线如果藏在壳体内，壳体的介电常数直接影响信号。碳纤增强体系基本无缘外壳，玻纤体系要在天线位置留净空区。这些约束在选料阶段就要和天线工程师坐下来对，等开机测试再对就晚了。
触摸电极隔着壳体感应，壳体厚度每增加0.5毫米，灵敏度就要重新调。料的介电常数一变，量产机上按键时灵时不灵。这是智能硬件售后里最玄学的一类投诉，根源往往在材料批次漂移，不是算法问题。
织物包覆的音箱，壳体外面的胶层长期受力。料的表面能决定胶接强度，表面能低的料要等离子处理。这道工序的成本和良率损失要在打样时算进去，不能等量产才发现贴不牢，撕下来重贴的工时能吃掉一个点的毛利。
成本结构上，音箱整机物料里壳体材料占的比例不高，但壳体决定的外观良率能差十几个点。把材料升级的钱花在这里，比花在喇叭单元上更能改善毛利，这是很多硬件团队算错的一笔账。
跌落测试是1米六面，音箱落地后网罩和壳体的卡扣先受力。卡扣材料韧性不够，测试过了运输还是裂，因为运输环境的震动频率和测试台不一样，冬天低温运输更脆。北方的冬季订单，这项要加做低温跌落。
工程实测：四条强制测试
测试1：阻燃析出测试。溴系阻燃 PA66 在 60℃ 1000 h 后表面析出 0.5 mg/cm²，无卤阻燃 PA66-GF15 析出 < 0.05 mg/cm²——必须无卤。
测试2：焊接强度。GF15 焊接强度 45 MPa，GF30 焊接强度 28 MPa——外观件选 GF15，结构件才能选 GF30。
测试3：射频透波。本色 PA66-GF15 介电损耗 0.015，含碳黑 PA66 介电损耗 0.15——后者完全屏蔽 WiFi。
测试4：触摸耐磨。加硬化涂层 1 万次摩擦后表面无可见磨损，无涂层 1 万次后明显磨花——必须加硬化涂层。
边界声明
| 工况 | 推荐材料 |
|---|
| 主流中高端 | PA66-GF15 + 无卤阻燃 + 高光泽 |
| 低端 | PA6-GF15 + 溴系阻燃 |
| 旗舰款 | PA12 + 阻燃 + 高光泽 |
| 触摸按键区 | 加硬化涂层 |
| 必须本色 | 不能用碳黑或金属填料 |
工程备忘
智能音箱外壳量产前必须做阻燃析出 + 焊接强度 + 射频透波三项测试。任何一项不达标整批报废。
模温和烘干是两个最容易漏掉的隐性变量——外观不良的隐性原因。
追问三连
问一：外壳用什么体系最多？阻燃改性的PC和ABS合金占主流，内部支架走增强PA，两套料各管各的。混着用一套料去应付检测和外观两头的方案，最后两头都吃亏。
问二：能不能做金属漆？金属漆里的导电粉会屏蔽射频。天线外置的机器随便喷，天线内置的机器喷完漆信号掉一格，返喷的成本比换漆贵三倍。打样阶段就要做整机信号测试，不能凭感觉。
问三：打样料和量产料怎么对齐？把牌号和批次写进打样报告，量产只认牌号不认价格。任何替换必须重新做全套外观和射频验证，这一条写进采购制度比写进合同管用，制度管的是每一次下单。
反向案例与收尾判断
有个团队为了赶发布，把送检的阻燃料临时换成库存里的非阻燃料，想着检测完再换回来。结果送检样和量产样的燃烧等级对不上，认证被卡了一个月，发布直接跳票。
硬件产品的每一张认证都绑着一个具体料号，材料的任何变动都要走认证变更，这个流程没法绕。绕过去的，都会在后面加倍还。音箱类产品的选料逻辑是三件事三把尺，外观件看良率，结构件看可靠性，检测件看认证，别用一把尺子量所有件。
实战案例：常见踩坑与正解
踩坑一：按家用件物性表直接套到商用场景，结果商用智能音箱2 年内集中漏水 / 变形 / 失效。
正解：商用和家用是两条产品线，料号玻纤含量、抗水解剂、阻燃等级全部要重新选——这是 80% 售后投诉的根因。
踩坑二：为了省成本用同一种料做整件，结果铰链 / 卡扣 / 阀座这些关键件先坏，整件报废。正解：结构件、连接件、外观件分别选料，不要图省事用同一种料。
踩坑三：耐候件没加 UV 三件套，结果户外安装半年就黄变脆裂。正解：任何户外或窗边安装必须加 UV 吸收剂 + HALS + 抗氧剂三件套，这是 5 年寿命的基本盘。
这三个坑都是量产前必须自查的清单，少一项量产后集中爆，修一颗螺丝的成本是新件的三倍。
补记：交接、样板与涨价的三件事
方案商和品牌方之间的料单交接，是智能硬件最容易断的一环。方案商打样用的料号写在BOM里，品牌方量产时换了代工厂，代工厂按价格重新询料，牌号变了，外观和射频的表现跟着变。责任边界上，方案商说料单交付了，品牌方说没说不能换。
解法很机械，把关键件料号写成冻结项，替换要走书面变更，双方签字。这一页纸的流程，能省掉的是量产前两周那种楼道里的会。
外观检验要有限位样板，不是文字描述。喷漆件的色差、光泽、橘皮，写再多文字都不如签一套三块的金样板灰样板黑样板。样板放在产线首件台上，每两小时对一次。
有家厂靠文字标准验货，验货员和设计师对颜色的理解差了一个色阶，整批返喷，返喷的资材够签三套样板用五年。样板制度是最便宜的质量工具，没有之一。
涨价周期里怎么守料号，也是个现实问题。上游原料涨的时候，料厂会推荐便宜的替代牌号，销售话术是性能差不多。差不多这三个字，在外观件上就是差一档光泽，在射频件上就是差一格信号。
守得住的厂有一个共同点，替代牌号必须重新走全套验证，验证周期两周，这个时间成本摆在台面上，销售端自然会把涨价谈判拉回正轨，而不是拿材料开刀。
补记二：从一台样机到一万台的距离
样机和量产之间隔着的，是批次的方差。我们见过最典型的一幕，试产的五十台全数通过，量产的第一批里百分之十的机器按键不灵。追下去是壳体料换了批号，介电常数漂了那么一点，触摸算法的阈值没留余量。
解法有两个方向，一是把算法阈值放宽，二是把料的批次公差锁死，多数团队选择改算法，因为改代码不要钱。但阈值放宽的代价是误触率上升，两条曲线挤在一起，最后还是回到材料。
硬件团队最好在立项的时候就知道，材料参数不是成本表上的一行，是产品定义的一部分。
还有一件事值得写给初创团队，外观件的公差链要留给喷涂。塑料件出厂的尺寸是裸件的，喷一道漆厚二十到三十微米，配合位就吃掉这么多。设计的时候按裸件公差配合，喷完漆装不上或者过松，模具返修一轮一个月。
把喷涂厚度写进图纸的公差链，是打样阶段就该做完的作业，别等量产。
延伸：三个来自产线的细节
细节之一，喷涂遮蔽。按键孔和接口孔的遮蔽治具磨损之后，漆会渗进孔内，孔径缩了零点几毫米，充电线插不紧。治具按使用寿命强制更换，这条写进工装台账。
细节之二，网罩磁吸的公差。磁吸座嵌在壳体里，注塑的缩水让磁力间隙忽大忽小，吸力时强时弱，用户对着灯找原因。嵌件的定位柱要按模具实况修配，不是按图纸理论值。
细节之三，包装前的静置。喷完漆就装盒，漆面没干透，开箱有味道，退货理由清一色写的气味大。产线末端加一段静置晾干的时间，比客服处理退货便宜得多。
结语
有些生意我们不做——这是我们做的每一份选料建议的验收线。
家用 / 卫浴 / 智能家居整套生活件的选料与试模，可以一起聊。
What type of modified nylon is used for the casing of the 88 smart speaker?
The working condition of the smart speaker is not like the casing
The smart speaker casing looks like a plastic shell, but in practice it faces very intensive conditions: long-term power-on (with circuit board heat dissipation inside), WiFi/Bluetooth radio frequency, wear on touch buttons, high-gloss appearance, flame retardant UL94 V-0, and ultrasonic welding strength.
All five requirements must be met simultaneously. If there is a mistake in material selection, the entire batch is scrapped. The smart speaker casing is not just a household appliance casing—it simultaneously serves as a structural part, an appearance part, a heat dissipation part, and a shielding part, all in one.
On-site restoration
Last January, the flagship speaker of a Shenzhen solution provider ran into a problem two weeks before mass production. After the casing was painted and the grille installed, the edges of the openings were all burrs, and the yield dropped to 60%. The project team held a meeting in the hallway, and the structure manager’s exact words were, 'There was clearly no problem during the prototype stage.'
After comparing three batches of material, we found that the melt index of the mass production material was one grade higher than that of the prototype material, it flowed faster, and the molecular orientation at the thin edges changed. As soon as we applied paint, all the burrs were exposed. In the end, it wasn't the material that was changed, but the mold for the mesh cover holes was chamfered again, and then the material grade was locked.
For products like speakers, the yield rate of the appearance parts equals the profit margin, and the material batch drift must be controlled during the sample stage.
Appearance Dual requirements for flame retardancy
The most common failure of smart speakers is the whitening of the appearance caused by the precipitation of flame retardants. Ordinary flame-retardant PA66 with added brominated flame retardant shows surface precipitation after 3 months, resulting in appearance scrap.
Halogen-free flame retardants must be used with a surface gloss synergistic formula. Halogen-free flame retardant PA66-GF15 high-gloss modification is the mainstream solution. Red phosphorus flame retardants precipitate less and are also a common solution—but red phosphorus has color limitations and can only be used for dark appearances.
The boundaries of ultrasonic welding
Ultrasonic welding of the upper and lower shells of smart speakers is a mainstream assembly process. Welding strength is sensitive to glass fiber content — GF15 has the best welding strength, while GF30 sees a 40% decrease in welding strength.
Choose GF15, not GF30 — this is the difference between appearance parts and structural parts. The weld strength test should refer to the lap shear strength of the ASTM D1002 standard, with a target of > 30 MPa.
The specificity of radio frequency wave penetration
Smart speakers have WiFi/Bluetooth antennas, and the casing must not contain carbon black or metal fillers—it will block RF signals. The formula must be natural or light-colored, and conductive/antistatic fillers cannot be used.
Dielectric loss < 0.02 is a strict requirement for wave transmission — carbon-black-filled PA66 has a dielectric loss > 0.15, completely blocking WiFi.
Durability of touch buttons
The touch button area on the top of the smart speaker must be wear-resistant — it should not be worn after 10,000 finger rubs. A wear-resistant coating or surface hardening treatment must be applied.
PA66-GF15 surface hardening coating is currently mainstream. The coating thickness is 5-10 μm — too thin and it is not wear-resistant enough, too thick and it affects touch sensitivity.
Extended Judgment: The Hidden Variables of Smart Speakers
There are three easily overlooked hidden variables in the smart speaker casing. The first is mold temperature — the mold temperature requirement for PA66-GF15 is 80-100℃, which is 20℃ higher than that of general PA66. If the mold temperature is insufficient, the surface will show exposed glass fibers.
Second is humidity control — PA66 pellets must be dried at 80℃ for 4 hours before injection molding; pellets that are not dried will have silver streaks and bubbles on the surface. Third is color stability — during multi-batch production, a ΔE < 1.0 is the pass line for color matching, while > 2.0 will result in customer complaints about batch color differences.
A deeper look: The origin of a few numbers
Flame retardancy is a mandatory item for inspection. When the speaker is connected to mains electricity, the enclosure is required to meet V-0, and the fillers in the flame retardant system will reduce the surface gloss. Appearance parts and structural parts are molded separately, which is the default action in mold design for this type of product. The extra cost of a set of molds is far less than the loss caused by reworking the appearance, and this calculation should be done before starting the mold.
RF wave penetration needs to be aligned in advance. If the 2.4 GHz and 5 GHz antennas are hidden inside the casing, the dielectric constant of the casing directly affects the signal. Carbon fiber reinforced systems basically cannot have an outer shell, while fiberglass systems need to leave a clearance area at the antenna position. These constraints need to be discussed with the antenna engineer during the material selection stage; it's too late to adjust during machine testing.
The touch electrode senses through the casing, and for every 0.5-millimeter increase in casing thickness, the sensitivity needs to be recalibrated. When the dielectric constant of the material changes, the buttons on mass-production machines may work intermittently. This is the most esoteric type of complaint in smart hardware after-sales service, and the root cause is often the batch drift of materials, not an algorithm problem.
Fabric-covered speakers, with the adhesive layer on the outside of the housing under long-term stress. The surface energy of the material determines the bonding strength; materials with low surface energy need plasma treatment. The cost and yield loss of this process need to be calculated during the prototyping stage, not discovered only during mass production when the adhesive won't stick properly. The labor of peeling and reapplying can eat up a point of gross profit.
In terms of cost structure, the proportion of housing materials in the overall speaker components is not high, but the appearance yield determined by the housing can differ by over ten percentage points. Spending money to upgrade the materials here can improve gross profit more than spending it on the speaker units, which is a miscalculation many hardware teams make.
The drop test is from a height of 1 meter on six sides. When the speaker falls to the ground, the mesh cover and the casing buckles bear the force first. If the buckle material lacks toughness, it may still crack during transportation even after testing, because the vibration frequency in the transportation environment is different from that of the test platform. Transportation in cold winter makes it more brittle. For orders in northern winter, this requires an additional low-temperature drop test.
Engineering field measurement: four mandatory tests
Test 1: Flame retardant exudation test. Brominated flame retardant PA66 exudes 0.5 mg/cm² on the surface after 1000 h at 60°C, while halogen-free flame retardant PA66-GF15 exudes less than 0.05 mg/cm²—must be halogen-free.
Test 2: Welding strength. GF15 welding strength 45 MPa, GF30 welding strength 28 MPa — exterior parts should use GF15, while structural parts can use GF30.
Test 3: Radio frequency transparency. Natural PA66-GF15 dielectric loss 0.015, carbon black-filled PA66 dielectric loss 0.15 — the latter completely blocks WiFi.
Test 4: Touch wear resistance. After 10,000 rubs with a hardened coating, there is no visible wear on the surface; after 10,000 rubs without a coating, noticeable wear occurs — a hardened coating is necessary.
Boundary Declaration
| Operating condition | Recommended materials |
|---|
| mainstream mid-to-high-end | PA66-GF15 Halogen-Free Flame Retardant High Gloss |
| low-end | PA6-GF15 Brominated Flame Retardant |
| Flagship model | PA12 Flame Retardant High Gloss |
| Touch key area | Hardened coating |
| Must be true to oneself | Cannot use carbon black or metal fillers |
Engineering Memo
Before mass production of the smart speaker casing, it is necessary to conduct tests for flame retardant exudation, welding strength, and RF wave penetration. If any of these tests fail, the entire batch will be scrapped.
Mold temperature and drying are two hidden variables that are most easily overlooked—the hidden reasons for poor appearance.
Three consecutive follow-up questions
Question 1: What system is most commonly used for the casing? Flame-retardant modified PC and ABS alloys dominate, while internal brackets use reinforced PA; each set of material handles its own part. Using a mixed single material to deal with both testing and appearance ends up disadvantaging both aspects.
Question 2: Can metallic paint be made? The conductive powder in metallic paint blocks RF. Machines with external antennas can spray it freely, but internal machines lose a notch of signal after painting, and the cost of respraying is three times higher than repainting. During the prototyping stage, you need to test the whole machine's signal, not just by feel.
Question 3: How do you align sample material with mass-produced material? Write the grade and batch in the sample report; for mass production, only the brand is recognized, not the price. Any replacement must be redone with a full set of appearance and RF verification. This clause is more effective than the contract in the procurement policy, as the system governs every order.
Reverse Case and Final Judgment
A team temporarily replaced the resistor fuel submitted for inspection with non-resistive fuel from inventory, planning to replace it after testing. But the combustion levels of the submitted samples and the mass-produced samples didn't match, certification was delayed for a month, and the release was immediately delayed.
Every certification for hardware products is tied to a specific part number; any material change must go through certification changes, and this process can't be bypassed. Anything that goes around will be paid back later. The logic for selecting speaker products is three things and three standards: appearance parts look at yield, structural parts look at reliability, testing parts depend on certification. Don't use one ruler to measure every part.
Practical Case: Common pitfalls and correct answers
Pitfall 1: Applying the property table of household components directly to commercial scenarios, resulting in concentrated leaks, deformation, and failure of commercial smart speakers within 2 years.
Correct answer: Commercial and household are two different product lines; part numbers, glass fiber content, anti-hydrolysis agents, and flame retardant ratings all need to be reselected—this is the root cause of 80% of after-sales complaints.
Pitfall 2: To save costs, using the same material for a whole piece results in key parts like hinges, latches, and valve seats failing first and the whole piece is scrapped. Correct answer: Structural parts, connectors, and appearance parts should be selected separately; don't save effort by using the same material.
Pitfall 3: Weather-resistant parts don't have a UV three-piece set, so after half a year of outdoor installation, they turn yellow and crack. Correct answer: Any outdoor or window-side installation must include a UV absorber + HALS + antioxidant three-piece set, which is the foundation for a 5-year lifespan.
These three pitfalls are all checklists you must check before mass production. If one is missing, it will collapse all at once, and repairing one screw costs three times more than new parts.
Supplement: Handover, prototyping, and price increases .
The handover of material lists between the solution provider and the brand is the easiest link for smart hardware. The solution provider's sample number is written in the BOM. When the brand changes the OEM during mass production, the OEM renegotiates the price, and the grade changes, so does the appearance and RF performance. At the boundary of responsibility, the solution provider said the material list had been delivered, but the brand said it wasn't said and no exchanges allowed.
The solution was mechanical: key part numbers were written as frozen items, replacement required a written change, and both parties signed. This page of the process saved the hallway meetings two weeks before mass production.
Appearance inspection requires limited samples, not written descriptions. No matter how many words are written about the color difference, gloss, and orange peel of spray paints, it's not as good as signing a set of three gold samples, gray samples, or black samples. Samples are placed on the first assembly line and checked every two hours.
A factory relies on textual standards to inspect goods, but the inspector and designer differ by a color level in understanding, so the entire batch of resprayed materials is enough to sign three sets of samples for five years. The sample system is the cheapest quality tool, bar none.
How to maintain material numbers during price hike cycles is also a practical issue. When upstream raw materials rise, suppliers recommend cheaper alternative grades, with sales pitches saying performance is similar. The word 'about almost' means a drop of gloss in appearance parts, and a signal difference in RF parts.
Factories that can hold firm have one thing in common: substitute grades must undergo a full set of validation, with a two-week validation cycle. This time cost is obvious, so sales will naturally bring price negotiations back on track rather than cutting over materials.
Supplement 2: The distance from one prototype to ten thousand units
The gap between prototypes and mass production is the variance of batches. We've seen the most typical scene: all fifty trial units passed, but in the first batch of mass production, 10% of machines had malfunction. Tracking down, the case material changed the batch number, the dielectric constant drifted a bit, and the touch algorithm threshold left no margin.
There are two approaches to the solution: one is to relax the algorithm threshold, the other is to lock the batch tolerances of the material. Most teams choose to change the algorithm because changing the code costs nothing. But the cost of lowering the threshold is higher mistouch rates, the two curves squeeze together, and in the end, it all comes back to the material.
The hardware team should know from the start that material parameters are not just a line on the cost sheet, but part of the product definition.
One more thing worth writing to the startup team: the tolerance chain for appearance parts should be left for painting. Plastic parts are factory size bare parts; spraying a coat of paint 20 to 30 microns takes up a lot of space. During design, fit according to bare part tolerances; if the paint doesn't fit or is too loose after spraying, mold repairs take a month after a round.
Write the coating thickness into the tolerance chain on the drawings; this is a task that should be completed during the prototyping stage, not mass production.
Extension: Three details from the production line
One of the details is spray masking. After the fixture for button holes and interface holes wears out, paint seeps into the holes, the hole diameter shrinks by a few millimeters, and the charging cable cannot be plugged in tightly. The fixture is forcibly replaced according to its service life; this is recorded in the tooling ledger.
Detail Two: Mesh Cover Magnetic Attachment Tolerance. The magnetic base is embedded in the housing, and the shrinkage from injection molding causes the magnetic gap to fluctuate between large and small, with suction sometimes strong, sometimes weak. Users look up to the light to find the reason. The positioning posts of inserts must be adjusted according to the actual mold conditions, not according to the theoretical values on the drawings.
Detail Three: Resting before packaging. After painting, the paint is packed in boxes; the paint surface isn't fully dry, there is a smell when opened, and the return reasons are all written as strong odors. Adding a period of sitting and drying at the end of the production line is much cheaper than customer service handling returns.
Conclusion
Some businesses we don't do — this is the acceptance line for every material selection suggestion we make.
For the selection and molding of complete household items for home / bathroom / smart home, you can chat together
