Key takeaways
A checklist has only two axes - how much disturbance the product emits outward, and whether it keeps working while disturbed. In 12 V systems the most frequently missed group is transient immunity and ESD, because it appears in no BOM and no drawing, yet it decides the outcome of vehicle-level road testing.
The most common request at quotation stage is "send us an EMC test item checklist first". Many of those checklists are simply a standard's annex copied into a table - complete-looking, but unusable in practice. Here is the conclusion: a checklist has only two axes, how much disturbance the product emits outward, and whether it keeps working while disturbed. In 12 V systems the group most often missed is transient immunity together with ESD: it appears in no BOM and no drawing, yet it decides whether vehicle-level road testing passes. The rest of this article sets out how the list is built, in four steps.
1. Two axes: emission and immunity must be scheduled separately
Emission asks whether the part will interfere with others; immunity asks whether it still functions when interfered with. Their acceptance sources differ and their sample operating states differ, so they cannot be compressed into one table.
| Axis | Typical items | Acceptance source | If omitted |
|---|---|---|---|
| --- | --- | --- | --- |
| Emission | Radiated emission, conducted emission (voltage method / current probe) | CISPR 25 limit standard and the class for the mounting location | Radio noise or CAN errors appear only during road tests |
| Immunity | Radiated immunity, conducted immunity, bulk current injection, ESD | ISO 11452 methods and functional status definitions in the agreement | Resets, blank displays or bus dropouts that a bench test never catches |
| Supply-related | Transient pulses, voltage dips, load-dump-related items | ISO 16750 electrical load clauses | Failure at ignition, stop-start or generator load switching |
| Decision basis | Functional status classification, monitoring items, operating mode | GB/T 28046.1 and the technical agreement | Both sides read different conclusions from the same data |
How the radiated emission setup is built, and why the voltage method and the current probe do not measure the same thing, are covered separately in radiated emission setup and the two conducted emission methods. How limit classes are read is in Class 1-5 selection. This article only groups the items and does not restate any condition values.
2. Why 12 V systems form their own group
The transient environment of a 12 V passenger-car supply network is not the same as a 24 V commercial vehicle. Stop-start, energy recuperation and generator load switching all put disturbance onto that line. A dedicated group of supply-line items therefore belongs on the checklist, rather than being folded into the immunity group as an afterthought.
| Supply item | Trigger | Main difference vs 24 V | Reference |
|---|---|---|---|
| --- | --- | --- | --- |
| Transient pulses and load dump | Load switching, inductive disconnection | Pulse shape and clamping approach | ISO 16750-2 / agreement |
| Voltage dip and interruption | Cranking, stop-start restart | Definitions of depth and duration | GB/T 28046.2 |
| Superimposed AC ripple | Generator rectification residual | Frequency band and amplitude basis | Standard text |
| Reverse polarity and overvoltage | Assembly, jump start | Conditions set by agreement | Customer agreement |
| Electrostatic discharge | Assembly, handling connectors | Unrelated to supply voltage, but usually scheduled here | ISO 10605 / ESD page |
One caution: all condition values in this group follow the applicable standard revision and the customer technical agreement. After a revision, the item name may stay the same while the conditions change, so the checklist should state the revision year rather than the item name alone. Method differences on the immunity side - field calibration, modulation and dwell in the ALSE method - are in ISO 11452-2 ALSE; the radiated immunity capability page and the emission capability page also map item ownership directly.
3. Cut by interface and mounting location
The same part mounted in a different location or attached to a different harness should not carry the same checklist. The criterion for cutting an item is not cost, but whether the physical coupling path that item addresses actually exists on the vehicle.
- Power and ground only, no external signal lines: conducted emission deserves attention on the emission side, while radiated emission stays according to mounting location and cannot be cut as a group. The conducted emission page lists the corresponding measurement basis.
- Long harnesses or multiple sensor lines: the harness is both a transmitting antenna and an injection path. Immunity items must be retested with the real harness length and grounding layout; conclusions do not carry over after a harness change.
- Wireless interfaces or high-frequency clocks: confirm band ownership and coexistence items so the part does not suppress other vehicle modules.
- Cabin / exterior / chassis mounting: location sets environmental severity and limit class ownership. The class comes from the specification revision tied to the part number - the laboratory does not assign the class on the customer's behalf.
The result is best recorded as a three-column table of keep / cut / add, with the reference document and revision year on every row, reviewed with the customer before shipping samples. If an audit later rejects the report, this shows exactly which layer of criteria was misaligned, instead of forcing a full retest.
4. Three pitfalls when the checklist goes live
First, item names without a revision. Writing "CISPR 25 radiated emission" is not enough; state the adopted revision year and the class document tied to the part number. Otherwise conditions change after a revision and the report no longer matches the checklist. Second, sample state missing from the test outline. The immunity group must state operating mode, monitoring items, permitted functional status classifications, and whether low-power states such as sleep and wake are in scope - this is the part most often left as a verbal agreement. Third, treating retesting as extra work. A change to harness, grounding or software version means the relevant conclusions are not inherited automatically. Reserve a retest slot in the checklist rather than adding it after the report is issued.
A few plain words
The value of a checklist lies not in being exhaustive, but in being able to say why each item is on it and which document supports it. When TMC LAB takes on this kind of commission, the supply form, interface list, mounting location and applicable standard revisions are written into the test outline first, with the acceptance source marked item by item, and testing is scheduled only after confirmation. Conditions and limits follow the standard text and the customer technical agreement; no values are preset at the checklist stage.
The genuinely time-saving move is to send the interface table, harness definition and the specification revision tied to the part number along with the samples, so the checklist is aligned once - far less rework than adding items mid-test. Scheduling is arranged per project; you can state your expected delivery date when submitting samples, and we will confirm a feasible plan against the current schedule and report progress at each milestone.
