On an automotive electronics test report, the "functional status" field is often filled in as an afterthought. It is in fact a defined judgement field: A to E describe how the sample behaved during the test and after it. Grade it wrong and the conclusion loses its meaning — more trouble than testing the wrong condition.
1. What the five levels actually say
GB/T 28046-1 is an identical adoption of ISO 16750-1, so the two divide functional status the same way. The judgement has two dimensions: whether function is normal during the test, and what it takes to recover afterwards.
| Class | During the test | After the test | Recovery |
|---|---|---|---|
| A | All functions work as designed | All functions normal | No intervention |
| B | Functions work as designed, individual parameters may exceed the specified tolerance | All back within specified limits | Self-recovering |
| C | One or more functions do not work as designed | Normal operation restored | Self-recovering |
| D | One or more functions do not work as designed | Operation restored | Requires a simple manual action, such as a power cycle or reset |
| E | Functions do not work as designed | Cannot be restored | Requires repair or replacement |
Two boundaries are easily confused and worth remembering separately: the point of class B is that the device still works during the test, only the parameters have drifted — it is not an interruption of function. The line between C and D is whether a human action is needed to recover — coming back by itself is C, needing someone to power cycle it is D.
2. Why class C often still counts as a pass
The class itself carries no value judgement. How far down the scale is allowed depends on where the part is mounted and whether a failure threatens safety.
Safety-related parts (braking, steering, airbag control) usually require A — any interruption of function during the test is unacceptable. Comfort and infotainment parts (seat adjustment, ambient lighting, some audio and video modules) allow B or even C: with a self-recovering interruption, all the user notices is that something reconnected.
What actually decides the judgement is not the class table but the "test item by functional status" matrix in the technical agreement. Some OEMs set requirements by functional safety level, allowing different classes for different functions on the same part — and if that detail is not written into the agreement, it can only be guessed.
3. The three most common misjudgements
Writing only "function normal after the test", with no record of the state during it. This is the most widespread problem. With no record during the test, class A cannot be demonstrated at all. Class A requires normal operation throughout, not merely surviving to the end.
No monitoring, so the judgement cannot be made afterwards. If the sample is not powered and not monitored during the test, an interruption of function goes unnoticed. Write the monitoring plan into the test outline at the time of commissioning: continuous power, periodic self-check, logging of fault codes or output states. When TMC LAB handles this kind of work, the monitoring method and sampling interval are agreed before the chamber is started, so the report does not end up able to state only "normal after the test".
Passing off a class D that needs a reset as class C. "It recovers after a power cycle" sounds like self-recovery but is manual intervention. This misjudgement is common in re-testing after corrective action: the circuit is revised, the reset symptom persists, but the class is written from D as C — which buries the problem.
4. How to fix the definitions at commissioning
Three things written into the test outline save trouble later:
- Function list and priority: which functions must be class A, which may be B or C, item by item;
- Monitoring plan: how power is applied during the test, how data is acquired, the sampling interval, and how a failure is recorded;
- Judgement and re-test rules: whether a D or E aborts the run, when to re-test, and how the report expresses it.
These definitions work the same way for electrical load tests; the functional status logic in ISO 16750-2 electrical loads explained can be read alongside this. Running an exploratory round during development to establish the class boundaries is far cheaper than discovering before volume production that the judgement cannot be made — that rhythm is what exploratory development testing versus formal testing describes.
For a full clause-by-clause breakdown see the GB/T 28046 standard page, and the international equivalent on the ISO 16750 standard page.
Turnaround time is arranged according to the actual project. You are welcome to state your target date when submitting samples; we will check it against the current schedule, confirm a feasible plan with you, and update you at each milestone.
