To ensure that the myriad electronic systems in today’s cars do not fail prematurely, the AEC-Q100 standard for active components defines stringent tests across all aspects of performance.
In just a few decades, the automobile has gone from a mostly electromechanical transportation vehicle with just a modest amount of electronics (such as the radio) plus some electrical but non-electronic functions such as starter, air conditioner, distributor, and spark coil, to a sophisticated, multiprocessor system on wheels.
Whether powered by an internal combustion engine (ICE), battery-only electric vehicle (EV), or some version of a hybrid electric vehicle (HEV), today’s cars are laden with roughly one hundred processors, smart controllers, and associated networks. These support basic functions, safety roles, advanced driver assistance systems (ADAS), infotainment, internal and external wireless connectivity, and more (Figure 1).
Figure 1. We’ve come a very long way in vehicle electronics from the battery-free Ford Model T with hand-crank starter and self-powered magneto for providing spark-plug energy (Image: Qorvo).
Even a minuscule, near-trivial failure rate of a few parts per million per 1000 or 10,000 hours means that sooner or later — likely sooner — something will fail, resulting in a problem ranging from a nuisance or inconvenience to a problem with the power train or safety function, and that’s a scary thought.
As a result, this accumulation of electronic devices, subsystems, and functions has fostered a new urgency and intensity to the eternal problem of end-product reliability and longevity. Vehicle owners won’t tolerate breakdowns in functions or features, while manufacturers can’t afford the cost of recalls or in-warranty repairs.
At the same time, the car is a harsh environment for electronics as well as mechanical functions. It must deliver its performance over a wide range of outside temperatures and under-the-hood heat. Nor can it demand frequency routine maintenance such as an aircraft must have after a modest number of flight hours. Ongoing maintenance by the user, whether done personally by the car owner or at the dealer, must be kept to a minimum.
Yet today’s cars are generally extremely reliable in design and production. The average age of a car in the US is about 12 years as a result of this reliability (and the high price of new vehicles, of course).
This FAQ will discuss the test standard, which is a large part — but is not the entire story — of how such extreme reliability in electronics for cars is achieved. (Let’s not overlook that materials and mechanical reliability have also improved — you don’t see many “rust-buckets” these days.)
Q: How many parts are in a car, anyway?
A: That’s a good question, and I haven’t been able to find a completely credible answer. On one side, I see the number 30,000 cited frequently. But I suspect this number came from one unknown primary source, and others are simply quoting it — which does not vouch for its credibility.
Further, it’s unclear if this 30,000 number includes just the major parts and subassemblies or all of the parts down to “least divisible” individual electrical and mechanical components such as resistors and screws (Figure 2).
Figure 2. How many parts are in a car? It depends on your level of subassembly or divisibility, but it’s certainly many, many thousands (Image: pictures.4ever.eu).
I have seen one source cite “upwards of a million parts,” but again, without supporting sources for that number. We do know that one ECU (electronic control unit) can have as many as 1000 parts, and the car can certainly have dozens and easily a hundred such ECUs.
Q: Why does the number matter?
A: Basic reliability analysis shows that the prospect of failures increases rapidly with the number of parts. The associated basic math proves that even a small failure rate per part in a car with so many components soon results in a breakdown of some subsystems.
Whatever the number, the number of electronic parts is large and increasing, even with higher levels of integration in the ICs. That’s because there is so much going on, while power devices such as motor-drive MOSFETs and sensors with their analog I/O front ends generally can’t be integrated into higher-level ICs.
In short, the increase in electronic functions drives the number of parts used to increase faster than the higher levels of integration, causing it to decrease.
Q: What’s the governing standard for active electronic components to ensure their reliability?
A: It’s AEC-Q100, “Failure Mechanism Based Stress Test Qualification for Integrated Circuits,” promulgated by the Automotive Electronics Council (AEC).
Q: Who are they? I’ve never heard of them.
A: The Automotive Electronics Council was initially established in the 1990s by Chrysler (now Stellantis), Ford, and General Motors to establish common part-qualification and quality-system standards. The council members include auto manufacturers, Tier-1 suppliers of automotive assemblies, and vendors of individual components, among others.
The next part looks at AEC-Q100 in more detail.
EE World related content
- 4D imaging radar on-a-chip is automotive-grade, AEC-Q100 qualified
- AEC-Q-100-qualified flyback switcher boasts 30 V to 550V DC input efficiency
- Clock generators, buffers, and PCIe clocks/buffers AEC-Q100-qualified for auto apps
- Gate-driver IC family certified to AEC-Q100 Grade Level 1 offer high-power safety and functionality
External references
- Monolithic Power Systems, “Fundamentals of AEC-Q100: What ‘Automotive Qualified’ Really Means” (very useful and informative)
- Qorvo, Inc. “Automotive Quality Standards 101: What Qualification Really Gets You”AEC Council, “AEC Documents” (useful and informative)
- Japanese Motor Works, “The Average Car Has 30,000 Parts. Which Are Most Important?”
- Texas Instruments, “TI Application Report SNOA994: Calculating Useful Lifetimes of Temperature Sensors” (not about AEC-Q100, but a good overview of reliability and lifetime)
You may also like:
Filed Under: FAQs, Featured, Testing and Safety