168 传感器外壳与接插件
传感器外壳的工况千差万别
传感器种类太多，工况没有统一答案：接近开关接触切削液，压力传感器接触液压油，温湿度传感器接触空气，食品传感器接触清洗剂。
选料的第一步是列介质清单，不是看温度。这一点和一般结构件完全不同。
PA66-GF30 是通用解
在没有特殊介质的情况下，PA66-GF30 阻燃是传感器外壳的通用解——强度够、尺寸稳定、可阻燃、成本适中。
玻纤增强的意义在于尺寸稳定：传感器的感应面到检测物的距离是校准过的，外壳尺寸漂移会导致检测距离变化。
现场还原：发动机舱里的 120℃
2025 年 2 月，杭州一家做车用压力传感器的工厂带着问题件来了：传感器外壳装在发动机舱，客户样车路试四个月后，外壳与安装座之间出现微渗漏，介质是机油雾加振动。
品质负责人的原话：「常温测试全过，怎么上了车就不行。」
上车就不行，恰恰是车规件的全部难点。发动机舱的实测工况：局部热点 120-150℃、机油和乙二醇蒸气长期熏蒸、振动按整车谱走。常温测试对这三条一个都没覆盖。
我们解剖了渗漏件：PA66-GF30 本体没坏，坏的是密封面——玻纤取向导致密封面平面度在热循环后漂了 0.15 mm，密封圈压不住。整改走两条：换低翘曲矿物复合牌压平面度，密封面加金属嵌件圈。
重新做热冲击 1000 次加介质浸泡，零渗漏。这家厂后来把「车规件常温测试通过不等于可以装车」印在了试验报告的首页——这句话是用一轮路试换来的。
特殊介质要换料
接触切削液和液压油：PA66 长期会轻微溶胀，精度要求高的走 PPS。接触强酸强碱：PA66 不可用，走 PVDF 或 PPS。食品和制药：要食品接触认证，走 PPS 或食品级 PP。这三条是传感器选料最常见的分岔口。
屏蔽对传感器是硬需求
传感器输出的是微弱信号（mV 级），极易受电磁干扰。
金属外壳传感器屏蔽最优，塑料外壳必须做导电处理。接近开关和光电开关的塑料壳多数内衬金属屏蔽层。这一层不能省，省了现场就是误动作。
深一层：接插件的关键不在本体在配合
传感器外壳聊的是料，接插件聊的更多是「配合」，这个区别值得单独讲。
接插件的电气性能核心是接触电阻，接触电阻由端子镀层和插拔力决定，本体材料的作用是把这些条件稳定住：尺寸要稳（公差带内几十年不漂）、韧性要够（装配和插拔不裂）、耐热要够（工作温度下不蠕变松弛——端子预紧力靠本体撑着，本体蠕变了，接触电阻就慢慢涨）。
所以接插件本体主流是 PA66-GF25 到 GF30，还有一条容易被忽略的要求：不含游离胺类——胺会迁移到端子表面腐蚀镀层，接触电阻漂移就是从这来的。这一条物性表上看不出来，要问供应商的配方体系。
镀层端子配合还有个湿度问题：PA66 吸湿后尺寸微胀，插拔力变化，高频插拔的连接器对此敏感，要做湿热后插拔力验证。电信级连接器选低吸湿改性牌或者 PBT、PPS，就是这个原因。
一句话：接插件选料，本体料只占一半，另一半在「它和端子、密封件、灌封料的长期配合」上。单看物性表选料，在这类件上最容易翻车。
接插件的镀层配套
传感器的接插件（M8、M12 航空插）是失效高发点。插针是铜镀金，绝缘体走 PBT 或 PA66-GF25。
PBT 的优势是吸湿小、尺寸稳定，插拔 500 次后接触可靠。
PA66 在这里的短板还是吸湿——潮湿环境下尺寸变化会影响接触压力。
灌封与密封的隐性变量
传感器内部一般灌环氧或硅胶保护电路。灌封料的收缩率要低，否则会拉伤元件引脚。
外壳和灌封料的热膨胀要匹配，冷热循环后不能开裂。另外灌封前外壳要干燥——PA66 含水率高时灌封会产生气泡，气泡就是以后的渗水通道。
工程实测：4 条强制测试
测试1：介质浸泡。PA66 在液压油中 1000 h 溶胀 0.8%，PPS 仅 0.1%——高精度走 PPS。
测试2：尺寸稳定。PBT 吸湿尺寸变化 0.1%，PA66-GF30 0.35%——接插件绝缘体优先 PBT。
测试3：屏蔽效能。内衬金属层屏蔽 70 dB，无屏蔽 0 dB——塑料传感器壳必须加。
测试4：插拔寿命。PBT 绝缘体插拔 500 次接触电阻稳定，PA66 在潮湿后波动——接插件优先 PBT。
追问三连：采购最常问的三件事
一问：车规认证要哪些材料数据。 常规三件套：UL 黄卡（阻燃与 RTI）、第三方批次报告、IATF 体系下的 PPAP 文件包。具体到传感器外壳，还要介质相容性数据（机油、乙二醇、清洗剂）和热冲击循环数据——这两项没有现成报告，要按实车工况定制做。
二问：外壳和灌封料的配套要注意什么。 灌封固化放热和线膨胀差是两大隐患：灌封料固化温度高，壳体料要扛得住固化温度不变形；工作温区内两者膨胀系数差要控制在密封设计余量内。灌封前壳体先做预热验证，不是多此一举。
三问：屏蔽要求怎么做。 传感器屏蔽的主流做法是金属屏蔽壳加塑料内衬，塑料件做绝缘和定位，屏蔽交给金属。要求全塑轻量化的场合用镀镍碳纤维导电 PA，屏蔽效能 40 dB 上下，成本明显上浮，一单一议。### 算一笔材料账：传感器外壳的车规溢价
车规件的材料溢价经常被质疑，把账拆到单位上看就清楚了。
一只车用压力传感器外壳，车规专用牌比工业牌贵 2.5 元。按年供货 50 万只算，年多支出 125 万元——这个数字在采购会上通常是被攻击的对象。
但车规溢价的构成是：介质相容性全套数据（机油、乙二醇、清洗剂）、热冲击循环验证、PPAP 文件包、批次一致性管控。这些不是奢侈消费，是把「装车后出问题」的概率从百分之几压到百万分之几的成本。
算一笔召回期望值：汽车行业一次小规模召回的成本，单一动辄数百万；若缺陷触发强制召回并进入监管记录，连带的市场损失按千万计。
125 万的年投入，买的是把召回概率压低一个数量级——期望值角度这笔账是正的，而且是明显为正。
车规供应链里流传一句话：便宜的材料贵在售后。这句话在传感器这类安全关联件上不是修辞，是财务模型。给车厂供货的改性厂，报价单上把验证成本单列，客户反而买得痛快——黑箱溢价才是采购最难接受的东西。### 边界声明
| 工况 | 推荐材料 |
|---|
| 通用工业 | PA66-GF30 阻燃 |
| 切削液液压油 | PPS |
| 强酸强碱 | PVDF 或 PPS |
| 接插件绝缘体 | PBT-GF25 |
| 需要屏蔽 | 内衬金属层 |
工程备忘
传感器外壳量产前必须做介质浸泡 + 屏蔽效能 + 尺寸稳定三项。选料第一步是列介质清单，不是看温度。
实战案例：常见踩坑与正解
踩坑一：只看阻燃等级不看 CTI。传感器外壳装在带电回路附近，阻燃 V-0 但 CTI 只有 250 V，长期爬电后表面碳化短路。正解：带电件必须 CTI ≥ 400 V（相比漏电起痕指数），V-0 只解决起火，不解决爬电——这是电气件最常被漏掉的一条。踩坑二：用回收料或副牌料做绝缘件，介电强度批次波动大，耐压测试 5% 击穿。正解：绝缘件一律走正牌新料，批次附耐压报告。踩坑三：端子件装完一段时间扭矩衰减，以为是螺丝松了，实际是尼龙蠕变。正解：传感器外壳承载螺纹连接时必须玻纤增强到 GF25 以上，并在装配 24 h 后复拧一次。
反向案例：被清洗剂放倒的一批外壳
2024 年 7 月，苏州一家做工业传感器的工厂遇到怪事：装配线上用一种新的环保清洗剂洗外壳，洗完没问题，装完测试也没问题，客户用了三个月，外壳表面出现细密裂纹。
追查花了五周。最后在应力开裂这条线上找到了答案：PA66 对某些极性清洗剂敏感，注塑残留的内应力没退干净，清洗剂一接触，应力集中处沿分子链薄弱位开裂——业内叫环境应力开裂，破坏的时候根本不在载荷路径上。
这家厂的整改：清洗剂选型加了材料相容性验证一栏；外壳注塑后加退火工序；外壳料换应力开裂敏感性低的改性牌。三件事做完，再没复发。
这个案例的价值在于把「材料本身没问题」和「材料在这个体系里没问题」区分开了。传感器件是典型的系统件，料、介质、应力、温度四个变量互相作用，验证时少放一个变量，现场就会替你补上这一课——用返工和赔付的方式。
延伸判断：选型前要先确认的三件事
传感器外壳在选料之前，有三件事要先问清楚，顺序错了后面全部返工。
第一：长期使用温度是多少。短时峰值温度和长期工作温度是两回事，物性表上的热变形温度是短时指标，长期工作温度一般要打七折看。
第二：接触什么介质。油、水、清洗剂、汗液、电解液，每一种都会改变料号选择，介质清单比温度表更重要。
第三：有没有认证要求。阻燃、CTI、食品接触、涉水卫生、安规认证，有认证要求的件，换料号就要重新验证，代价远高于材料差价的几十块钱。这三件事问清楚，选料就完成了一半。
把这三件事写成一张表发给供应商，比打十通电话有用——传感器外壳的选型沟通成本，基本都花在这几项反复确认上。
补记：三个现场判断信号
信号一：密封面渗漏、本体完好。 先量平面度漂移，玻纤翘曲或热变形导致，换低翘曲牌或加嵌件，换密封圈只是续命。
信号二：裂纹沿装配应力方向、无外力断口。 环境应力开裂，查介质接触史和注塑残余应力，两头的账都要算。
信号三：接触电阻随时间缓慢上涨。 本体蠕变或胺类迁移腐蚀镀层，前者换低蠕变牌，后者换配方体系，两者都要做湿热后复测确认。### 验证顺序：三步走完再下单
第一步，对工况：发动机舱热点温度、介质清单、振动谱三样拿全，没有工况输入的选型是猜谜。
第二步，验组合：外壳、灌封、密封圈、端子做成套老化——传感器是系统件，单件验证漏掉配合问题。
第三步，锁体系：PPAP 文件、批次报告、变更报备条款一步到位，车规的坑大多出在「后续批次悄悄变了」。三步走完，车规件的供应关系才能长期成立。### 一句补充：别让外壳拖垮传感器
传感器这个产品有个特殊性：外壳坏了，客户不会修外壳，直接换整只传感器。外壳的成本占比往往不到 5%，却决定了那 100% 的更换频率。
所以外壳料的预算逻辑不是「省」，是「一步到位」——低蠕变、耐介质、尺寸稳三条一次配齐，比后续任何补救都便宜。选型沟通时把这句讲透，客户对专用牌的溢价就不再纠结。
结语
有些生意我们不做——选料这件事，越早问越省事。
这类件的选料与试模，可以一起聊。
168 Sensor housing and connector
The operating conditions of sensor housings vary greatly
There are too many types of sensors, and there is no unified answer: proximity switch contacts cutting fluid, pressure sensor contacts hydraulic oil, temperature and humidity sensor contacts air, food sensor contacts cleaning agent.
The first step in material selection is to list the medium, not the temperature. This is completely different from general structural components.
PA66-GF30 is a universal solution
In the absence of special media, PA66-GF30 flame retardant is the universal solution for sensor housings—sufficient strength, stable dimensions, flame retardant, and moderate cost.
The significance of fiberglass reinforcement lies in dimensional stability: the distance from the sensor's sensing surface to the tested object is calibrated, and size drift in the casing causes the detection distance to change.
On-site reconstruction: 120°C inside the engine compartment
In February 2025, a factory in Hangzhou specializing in automotive pressure sensors arrived with problematic parts: the sensor housing was installed in the engine compartment, and four months after the customer's prototype car road test, a slight leakage appeared between the casing and the mounting seat, with the medium being engine oil mist and vibration.
Quality Manager's exact words: "If all the tests pass room temperature, why is it bad once you get in the car?"
Once in the car, it doesn't work—precisely the real challenge for automotive-spec parts. Engine compartment test conditions: local hotspot 120-150°C, long-term fumigation of engine oil and ethylene glycol vapor, vibration following the whole vehicle pattern. Room temperature tests did not cover any of these three criteria.
We dissected the leakage part: PA66-GF30 main body was intact, but the faulty part was the sealing surface—the fiberglass orientation caused the flatness of the sealing surface to drift by 0.15 mm after thermal cycling, and the sealing ring couldn't hold it down. Rectification took two steps: replace with a lower warped mineral composite plate to press flatness, and add metal insert rings to the sealing surface.
Redo the thermal shock and soak it in medium 1000 times, zero leakage. This factory later printed "passing the room temperature test for automotive-grade parts does not mean it can be installed in the car" on the first page of the test report—this sentence was obtained through a round of road testing.
Special media require material replacement .
Contact cutting fluid and hydraulic oil: PA66 will swell slightly over time; high-precision applications should be made with PPS. Contact with strong acids and bases: PA66 is not suitable; use PVDF or PPS. Food and pharmaceuticals: For food contact certification, use PPS or food-grade PP. These three are the most common branching points in sensor material selection.
Shielding is a hard requirement for sensors
Sensors output weak signals (mV level), which are highly susceptible to electromagnetic interference.
Metal housing is optimal for sensor shielding; plastic casings must be treated for conductivity. Most plastic housings for proximity switches and photoelectric switches are lined with metal shielding layers. This layer can't be skipped; skipping it on site will result in malfunctions.
Deeper layer: The key to connectors is not the body itself.
The sensor housing is about the material; the connector is more about 'fitting.' This difference is worth discussing. The core electrical performance of the
connector is contact resistance, which is determined by the terminal coating and the insertion/removal force. The purpose of the body material is to stabilize these conditions: dimensional stability (no drifting within the tolerance range for decades), toughness (no cracking during assembly or plugging), and sufficient heat resistance (no creep or relaxation at operating temperature—terminal preload is supported by the body; when the body creeps, contact resistance gradually increases).
So the mainstream connector bodies are PA66-GF25 to GF30, and there's another easily overlooked requirement: no free amines—the amines migrate to the terminal surface and corrode the coating, and contact resistance drift comes from this. You can't see this on the physical property chart, so you need to ask the supplier's formula system.
There's also a moisture issue with coated terminals: PA66 expands slightly after absorbing moisture, causing changes in plug-in/pull force. High-frequency plug-and-pull connectors are sensitive to this, so you need to test the plug-in and pull-out force after wet heat. For telecom-grade connectors, choose low-moisture modified brands or PBT/PPS for this reason.
In short: For connector materials, the main material only accounts for half; the other half is about 'long-term compatibility with terminals, seals, and potting materials.' Judging solely by the physical property table when selecting materials, these types of parts are most likely to fail.
Connector Coating Matching
Sensor Connectors (M8, M12 aviation connectors) are high-risk failure points. Insertion pins are copper-plated with gold, and insulators use PBT or PA66-GF25.
PBT advantages are low moisture absorption, stable dimensions, and reliable contact after 500 insertion and unplugging.
PA66 The drawback here is moisture absorption—dimensional changes in humid environments affect contact pressure.
Implicit Variables in Potting and Sealing
Sensors are generally filled with epoxy or silicone protection circuits. The shrinkage rate of potting material should be low; otherwise, it will damage component pins.
The thermal expansion of the casing and potting material must match, and no cracking occurs after hot and cold cycles. Additionally, the casing must be dry before potting—when PA66 has high moisture content, bubbles will form, which will become the subsequent seepage channels.
Engineering Testing: 4 mandatory tests
Test 1: Medium soaking. PA66 expands 0.8% in hydraulic oil for 1000 hours, PPS only 0.1%—high-precision PPS is used.
Test 2: dimensional stability. PBT moisture absorption changes by 0.1%, PA66-GF30 0.35%—connector insulators prioritize PBT.
Test 3: Shielding performance. Inner lining metal layer shielded at 70 dB, unshielded at 0 dB—plastic sensor housing must be added.
Test 4: Plug-in/pull-out lifespan. PBT insulator stable resistance after 500 contacts after plugging and unplugging, PA66 fluctuates after wetness—connectors prioritize PBT.
Follow-up triple question: The three most frequently asked questions in procurement
First question: What material data are required for automotive-grade certification. Routine three-piece set: UL yellow card (flame retardant and RTI), third-party batch report, PPAP file package under IATF system. Specifically for the sensor housing, media compatibility data (oil, ethylene glycol, cleaning agent) and thermal shock cycle data are also needed—these two items do not have ready-made reports and must be customized according to actual vehicle conditions.
Question 2: What should be noted when matching the shell and potting material? Heat release and linear expansion difference during potting and curing are two major hidden dangers: the potting material has a high curing temperature, so the shell material must withstand the curing temperature without deformation; Within the working temperature range, the difference in expansion coefficients between the two must be controlled within the sealing design allowance. Preheating verification of the enclosure before potting is not redundant.
Question 3: How should shielding be done? The mainstream method for sensor shielding is metal shielding with plastic lining, with plastic parts for insulation and positioning, and shielding left to metal. In cases requiring full plastic lightweighting, nickel-plated carbon fiber conductive PA is used, with shielding efficiency around 40 dB, significantly increasing costs, and case negotiation. ### Doing the Math: The Automotive-grade Premium for Sensor Housings
Automotive-grade parts is often questioned; just break down the accounts to the unit level and it's clear.
A car pressure sensor housing, automotive-grade brand is 2.5 yuan more expensive than industrial grade. Annual supply of 500,000 yuan is just counted, annual exceeding 1.25 million yuan — this figure is usually targeted at procurement meetings. But
But the automotive-grade premium consists of: a complete set of media compatibility data (engine oil, ethylene glycol, cleaning agent), thermal shock cycle verification, PPAP file package, batch consistency control. These are not luxury purchases; they reduce the probability of "problems after installation" from a few percent to a few million in cost.
Calculate the expected recall value: The cost of a small-scale recall in the automotive industry can easily reach several million; If a defect triggers a mandatory recall and enters regulatory records, the resulting market losses could reach tens of millions.
The annual investment of 1.25 million yuan is to lower the recall probability by an order of magnitude—from the perspective of expectation, this account is positive, and clearly positive
There's a saying circulating in the automotive-grade supply chain: cheap materials are expensive because of after-sales. This phrase isn't rhetoric for safety-related components like sensors—it's a financial model. For modification factories supplying automakers, the verification cost is listed separately on the quotation, and customers end up buying happily—black-box premium is the hardest thing to accept in procurement. ### Boundary Declaration
| Working Condition | Recommended Materials |
|---|
| General Industrial | PA66-GF30 Flame Retardant |
| Cutting Fluid Hydraulic Oil | PPS |
| Strong Acid, Strong Alkalis | PVDF or PPS |
| Connector Insulator | PBT-GF25 |
| Shielding Required | Inner Metal Layer |
Engineering Memo
Sensor housing: before mass production, three items must be performed: medium immersion + shielding performance + dimensional stability. The first step in material selection is to list the medium, not the temperature.
Practical Case: Common pitfalls and correct answers
Pitfall 1: Only look at flame retardant rating, not CTI. The sensor housing is installed near a live circuit, with flame retardant V-0 but CTI only 250V. After long-term creep, the surface carbonizes and shorts. Correct answer: Live components must have a CTI ≥ 400V (compared to leakage trace index); V-0 only solves fire and not creepage—this is the most frequently missed electrical component. Pitfall 2: Use recycled or sub-grade materials as insulation components; dielectric strength batches fluctuate greatly, withstand voltage test shows 5% breakdown. Correct answer: All insulating parts must use genuine new materials, with batch voltage resistance reports attached. Pitfall 3: After installing terminal components for a while, torque decays occurs. Thinking the screw is loose, it's actually nylon creep. Correct answer: When connecting the sensor housing threaded under load, the glass fiber must be reinforced to GF25 or above, and after 24 hours of assembly, it should be retightened once.
Reverse Case: A batch of casings knocked down by cleaning agent
In July 2024, a factory in Suzhou specializing in industrial sensors encountered something strange: the assembly line used a new eco-friendly cleaning agent to clean the casing. After washing, there was no problem; after installation, testing was fine. The customer used it for three months, but fine cracks appeared on the casing surface.
Investigation took five weeks. Finally, the answer was found on the stress cracking line: PA66 is sensitive to certain polar cleaning agents, and the residual internal stress from injection molding hasn't fully dispelled. When the cleaning agent contacts it, the stress concentration cracks along weak molecular chains—this is called environmental stress cracking in the industry, and the failure occurs not even along the load path.
's rectification: cleaning agent selection added a material compatibility verification section; After injection molding, the shell was followed by annealing processes; The shell material switched to modified brands with low stress crack sensitivity. After all three things were done, there was no recurrence. The value of
's case lies in distinguishing between "the material itself is fine" and "the material is fine within the system." Sensor components are typical system components. Material, medium, stress, and temperature interact with each other. If you omit one variable during verification, you'll be taught the same lesson on site—using rework and compensation.
Extended judgment: Three things to confirm before selecting a model
Before selecting materials, there are three things to clarify. If the order is wrong, everything will be reworked.
First: What is the long-term operating temperature? Short-term peak temperature and long-term operating temperature are two different things. The thermal deformation temperature on the physical property table is a short-term indicator, while the long-term operating temperature should generally be discounted by 70%.
Second: What medium is it contacted? Oil, water, cleaning agent, sweat, electrolyte—each will change the part number selection. The media list is more important than the temperature chart.
Third: Are there certification requirements? Flame retardancy, CTI, food contact, water hygiene, safety certification—if there are certification requirements, the replacement part number must be reverified, at a cost much higher than the material price difference. If you clarify these three things, the material selection is half done.
Write out these three matters in a table and send it to the supplier; it's more effective than making ten phone calls—the cost of communication for sensor housing selection is basically spent on repeated confirmations of these items.
Additional note: Three on-site judgment signals
Signal one: sealing surface leakage, body intact. First, measure flatness drift; fiberglass warping or thermal deformation is the cause; replace with low warping or inserts; replacing sealing rings only prolongs life.
Signal 2: Cracks should follow the direction of assembly stress, with no external force fractures. Environmental stress cracking, check the media contact history and injection molding residual stress, and calculate both ends.
Signal 3: Contact resistance slowly increases over time. For body creep or amine migration corrosion coatings, replace the former with low creep brand, the latter with formula systems, and both must be retested and confirmed after wet heat. ### Verification sequence: Complete all three steps before placing an order
Step one: For operating conditions: engine compartment hotspot temperature, media list, vibration spectrum, all three items are included. Choosing models without operating condition input is a guessing game.
Step two: Verify the assembly: housing, potting, sealing ring, and terminals are assembled for aging—sensors are system components, so if individual verification is missed, coordination issues are missed.
Step three, locking system: PPAP files, batch reports, and change reporting clauses are all in place. Most pitfalls in automotive standards come from 'subsequent batches quietly changing.' Only after completing these three steps can the supply relationship of automotive-compliant parts be established long-term. ### One extra note: Don't let the casing drag down the sensor
Sensor products have a special feature: if the casing breaks, customers won't repair the case and will replace the entire sensor directly. The casing often accounts for less than 5% of the cost, yet it determines the 100% replacement frequency.
So the budget logic for shell materials isn't about 'saving,' but about 'getting it in one step'—low creep, medium resistance, and stable dimensions all at once, which is cheaper than any subsequent fixes. If you explain this clearly during model selection communication, customers won't worry about the premium for the special brand.
Conclusion
There are some businesses we don't do — the earlier you ask about material selection, the easier it is.
For material selection and mold trials for these parts, you can chat together
