结论：同一台设备，配置不同，结果可以差出很多
EMC 测试有一个特点：结果对试验配置高度敏感。 线缆的走向改变几厘米、设备的朝向转九十度、接地点换一个位置，测出的发射值可能有明显差别。
这带来两个实际问题：
一是复现性。 如果配置记录不全，复测时无法还原原来的状态，得到不同结果就无从判断是设备变了还是配置变了。
二是代表性。 测试用的配置应当代表实际使用中的典型配置，否则测出来的数据与实际情况脱节。
所以配置的确定和记录，是 EMC 测试中分量很重的工作，不是测试前随手摆一下。
配置的要素
| 要素 | 需要记录的内容 |
|---|---|
| 设备本体 | 型号、序列号、软件版本、朝向 |
| 附件与配件 | 型号、数量、连接位置 |
| 线缆 | 型号、长度、屏蔽、走向、是否盘绕 |
| 接地 | 接地点位置、连接方式 |
| 辅助设备 | 型号、位置、与被测设备的关系 |
| 负载 | 模拟负载还是真实负载、规格 |
| 工作模式 | 设备处于什么模式、参数设定 |
| 环境 | 温湿度、测试场地 |
「线缆走向」是最需要详细记录的一项，因为线缆是主要的发射源和接收天线，它的位置直接影响结果。仅写「按要求布置」是不够的，应当有照片或示意图。
记录的方法
照片。 多角度拍摄实际的布置状态。这是最直接有效的记录方式。
示意图。 标注各部件的相对位置和距离。
文字说明。 补充照片和图表达不了的信息。
清单。 所有设备、附件、线缆的型号与规格清单。
参数记录。 设备的设定参数、软件版本。
建议以照片为主、图文为辅。 照片能记录很多容易忽略的细节，而这些细节可能正是影响结果的因素。
易被忽略的细节
软件版本。 软件更新可能改变设备的电磁特性（时钟策略、功率管理）。这一项经常被漏记，而复测时软件已经更新，结果不同就无从解释。
线缆的盘绕方式。 多余的线缆怎么处理——盘成圈还是折叠，圈的直径多大，位置在哪里。不同的处理方式对结果影响明显。
电池的状态。 电池电量、是否在充电。
可动部件的位置。 可调节的设备，各部件处于什么位置。
外壳的状态。 盖板是否装好、螺丝是否紧固。松一颗螺丝可能改变屏蔽效果。
接插件的状态。 未使用的接口是否有防护盖。
辅助设备的电源。 辅助设备的供电方式是否引入额外的干扰路径。
这几项中，软件版本和线缆盘绕方式是影响较大又最常被漏记的。
典型配置的选择
配置要代表实际使用，选择时考虑：
最常见的使用方式。 多数用户怎么用。
最不利的配置。 从 EMC 角度最可能出问题的组合。
附件的组合。 如果有多种附件，选哪些、怎么组合。
线缆长度。 用最长的还是典型长度。通常选较长的，因为长线缆发射更强。
多台设备的情形。 如果实际使用中多台设备一起用，是否需要考虑。
「最不利配置」与「典型配置」有时不一致，这时要说明选择的理由。通常的做法是按最不利配置测，这样覆盖面更广。
方案中怎么写
试验方案中关于配置的部分建议包含：
配置的描述。 详细到可以照着复现。
选择的理由。 为什么选这个配置。
附件清单。 完整的型号和数量。
布置图或照片。 直观呈现。
工作模式与参数。 设备的设定。
监测方法。 如何判断设备工作正常。
判定准则。 什么情况算通过。
「详细到可以照着复现」是检验方案质量的标准。 写完之后可以自问：如果一年后换个人按这份方案重做，能不能得到相同的配置？
复测时的对照
产品变更或者定期复测时，配置应当与原测试一致，否则结果不可比。对照的要点：
设备的软件版本；附件的型号；线缆的型号与长度；布置方式；以及工作模式与参数。
如果配置必须改变（比如附件停产），应当说明变化并评估影响。 直接用新配置测出的数据与旧数据比较，而不说明配置变了，是常见的问题。
与产品变更的关系
产品变更后是否需要重测，判断依据之一就是配置相关的变化：
结构变化。 外壳、屏蔽、接地的改动直接影响 EMC 特性。
元件变更。 开关电源元件、时钟、驱动器的更换可能改变发射。
软件变更。 时钟策略、功率管理、通信协议的改变。
线缆变更。 型号、长度、屏蔽方式的改变。
附件增加。 新增附件可能改变整体特性。
判断的原则是：变更是否影响电磁特性。 影响的要重测相关项目，不影响的可以论证免测，但论证要写下来。
软件变更是最容易被认为「不影响 EMC」的变更，而实际上软件对时钟和功率的控制直接影响发射。涉及这些方面的软件变更应当评估。
记录的保存
EMC 测试的配置记录应当与报告一起保存，并且：
保存期限要足够。 至少覆盖产品的生命周期。
照片要可检索。 与报告编号关联。
版本要清楚。 多次测试时各次的配置要能区分。
变更要留痕。 配置调整过的，记录调整内容和原因。
这些记录在产品变更评估、复测、以及出现问题追溯时都会用到。
常见的记录痛点
实务中最常因为记录不全而出问题的几种情形：
复测结果不同。 无法判断是设备变了还是配置变了。
质疑时无法自证。 审查方质疑配置不具代表性，而记录不足以说明选择理由。
换实验室时无法延续。 新实验室按自己的理解布置，结果与原来不同。
内部人员变动后无法传承。 原来做测试的人离职，配置细节随之丢失。
这几种情形的共同代价是重测，而重测的成本远高于当时多拍几张照片、多写几行记录。
建议把配置记录做成固定模板，每次测试按模板填写，避免依赖个人习惯。
我们的做法
做 EMC 测试时，我们会拍摄完整的布置照片并记录所有配置要素，软件版本和线缆处理方式会专门记录，因为这两项影响大又容易漏。
对于附件较多的设备，我们会在测试前与委托方确认典型配置和最不利配置，并把选择理由写进报告。 这样将来复测或者别人复核时，依据是清楚的。
如果你有设备需要安排 EMC 测试，想先确定测试配置，可以把设备和附件清单发过来一起讨论，或者直接联系：132 4819 8029。能力范围见服务介绍，流程见检测流程，更多内容见知识库。
English version
Conclusion. Electromagnetic compatibility testing has one distinctive feature: results are highly sensitive to the test configuration. Moving a cable a few centimetres, rotating the equipment ninety degrees or changing an earthing point can change measured emissions noticeably. Two practical problems follow. Reproducibility suffers, since an incompletely recorded configuration cannot be restored and a different result on repetition cannot be attributed to the equipment or to the setup. And representativeness suffers, since the configuration tested should represent typical use, or the data diverge from reality. Determining and recording the configuration is therefore substantial work within EMC testing rather than something arranged casually beforehand.
Elements of the configuration. The equipment itself, recorded by model, serial number, software version and orientation. Accessories, recorded by model, quantity and connection point. Cables, recorded by type, length, screening, routing and whether coiled. Earthing, recorded by point and method of connection. Auxiliary equipment, recorded by model, position and relationship to the equipment under test. Loads, recorded as simulated or real with their ratings. Operating mode, recorded with parameter settings. And the environment, recorded as temperature, humidity and test site. Cable routing needs the most detailed record, because cables are the principal emission sources and receiving antennas and their position affects results directly. Writing that cables were arranged as required is not enough; photographs or a diagram are needed.
How to record. Photographs from several angles capture the actual arrangement and are the most direct and effective record. Diagrams mark relative positions and distances. Written description supplements what photographs and diagrams cannot express. Lists enumerate the models and specifications of all equipment, accessories and cables. And parameter records capture settings and software versions. Lead with photographs supported by diagrams and text: photographs capture details easily overlooked, and those details may be exactly what affects the result.
Easily overlooked details. Software version, since updates can change electromagnetic characteristics through clocking strategy or power management; this is frequently omitted, and when retesting occurs after an update, differing results cannot be explained. Cable coiling, meaning how surplus cable is handled, coiled or folded, at what diameter and in what position, since the treatment affects results noticeably. Battery state, meaning charge level and whether charging. The position of movable parts on adjustable equipment. Enclosure condition, meaning whether covers are fitted and screws tightened, since one loose screw can change screening effectiveness. Connector condition, meaning whether unused ports carry protective caps. And the supply arrangement for auxiliary equipment, which may introduce additional coupling paths. Of these, software version and cable coiling have the greatest effect and are the most commonly omitted.
Choosing a representative configuration. The configuration should represent real use. Consider the most common manner of use. Consider the least favourable configuration from an EMC standpoint. Consider accessory combinations where several exist. Consider cable length, whether the longest or a typical length, with the longer usually chosen since long cables emit more strongly. And consider multiple-unit situations where several units are used together. Least favourable and typical configurations sometimes differ, in which case the choice must be explained; testing the least favourable is the usual approach since it covers more.
Writing it into the plan. The configuration section should include a description detailed enough to reproduce, the reasons for the choice, a complete accessory list with models and quantities, a layout diagram or photographs, operating modes and parameters, monitoring methods for judging normal operation, and acceptance criteria. Detailed enough to reproduce is the test of plan quality: ask yourself whether someone else repeating the work a year later could arrive at the same configuration.
Comparison when retesting. After product changes or for periodic retesting, the configuration should match the original, or results are not comparable. Check software version, accessory models, cable types and lengths, arrangement, and operating modes and parameters. Where the configuration must change, as when an accessory is discontinued, state the change and assess its effect. Comparing new data against old without noting that the configuration changed is a common problem.
How we handle it. We photograph the arrangement in full and record every configuration element, with software version and cable handling recorded specifically because they matter and are easily missed. For equipment with many accessories we agree typical and least favourable configurations with the client before testing and record the reasons in the report, so that the basis is clear for future retesting or third-party review.
Send us the equipment and accessory list and we will settle the test configuration. Phone or WeChat: +86 132 4819 8029.