Selecting sensors for automotive systems
Article By : Jeff Viola
Being labelled as 'automotive qualified' does not guarantee a device's suitability for every automotive application. Specifications, quality processes and test results should also be considered.
Automotive applications require severe operating condistions on electronic components. The demand for reliability and long period of effectiveness is very high. The need for superior characteristics is often neatly summed up in the phrase ‘automotive qualified.’ However, sometimes the only difference between an automotive-qualified part and a non-automotive-qualified part is a wider operating temperature range. And in other cases, an automotive-qualified version of a consumer part will not only have more robust characteristics but will actually have markedly different features, such as an additional set of signal interfaces.
So on what grounds should an automotive manufacturer pay a premium for the automotive-qualified device? Do semiconductor manufacturers all mean the same thing when they describe a component as automotive-qualified? And which characteristics and attributes should a system designer look for when selecting a component for use in an automotive design?
Common characteristics of automotive qualified parts
The first thing to understand about the complicated field of automotive qualification is that silicon is only one of the ways in which an automotive part might be different. Some consumer-grade sensors could be used in a vehicle, depending on the location and the application, and the OEM’s willingness to waive certain standard requirements. For example, a magnetic rotary or linear position sensor might be used to detect whether a display screen mounted in the back of a front seat’s headrest is in the viewing position. An automotive OEM may consider a consumer part for this application, given that it is not safety-critical and the operating conditions are benign. The use of a non-automotive-qualified part will normally save at least a few cents from the system’s bill-of-materials (BOM) cost.
This use case, however, is unusual; in general, automotive applications have ultra-high reliability requirements, and may be exposed to extreme operating conditions. How is this reflected in the typical parameters by which sensor IC manufacturers rate their parts? The most common characteristics of sensor ICs described as ‘automotive qualified’ are:
- Temperature range: −40°C to 125°C or 150°C in most applications, but −40°C to 85°C inside the cabin
- Qualified according to AEC-Q100 standard laid down by the Automotive Electronics Council
- Compliant with the PPAP (production part approval process)
- Specified for tolerance of relatively high voltages caused by ESD (electro-static discharge)
- Compliant with relevant EMC (electro-magnetic compatibility) requirements; EMC requirements refer to both the emissions generated by the part, and the part’s ability to tolerate external interference.
- Carrying an ASIL (automotive safety integrity level) rating according to the provisions of the ISO 26262 functional safety standard
- Supported by an FMEDA (failure modes effects and diagnostic analysis)
- Supporting automotive-specific communications protocols and signalling interfaces
This list is neither comprehensive nor mandatory: there is no single industry standard or test that determines whether a sensor IC may be described as ‘automotive qualified.’ For instance, a climate control system’s cabin temperature sensor might have no functional safety implications, in which case it will not have an ASIL rating, and might have no FMEDA.
Instead, it is better to think of the attributes of automotive parts as falling into three broad categories:
- Special capabilities and features set
- Special testing regimes
- Special verification and qualification requirements
Automotive-specific features
Some automotive parts call for features that are simply not required in consumer or industrial products. For instance, a position sensor used in a safety-critical system such as an accelerator or brake pedal will require certain reliability features in order to support functional safety compliance: diagnostic flags, a checksum, sometimes even a dual-die package to provide redundancy. But an automotive position sensor that is not intended for use in safety-critical applications might not include these features, even if it meets the testing and qualification criteria described below.
Some communication protocols for transmitting a sensor’s output are rarely or never used outside automotive systems. While PWM, SPI, I2C, ABI, and UVW interfaces are commonly used in both industrial and automotive environments, automotive sensors from ams might also offer PSI5, SENT, and CAN interfaces.
A more rigorous testing regime
The recall and repair costs of a vehicle with a component fault are huge, which is why the automotive industry places an extremely strong emphasis on specifying the quality and reliability of the parts it uses. The quality requirement on automotive sensors calls for 100% end-of-line testing: at ams, every single manufactured unit of an automotive qualified sensor will be tested before shipment, to verify that the chip exceeds the minimum and maximum values for key performance parameters stated in the datasheet. This testing requirement extends the time that the chip spends in the factory, and this necessarily entails an increase in its cost.
Automotive qualifications
Compliance with the AEC-Q100 standard will normally be required if a sensor is to be used in an automotive application. Compliance shows that the part has undergone stress testing to quantify the part’s susceptibility to common known failure mechanisms for ICs. But ams guarantees that all automotive products exceed current AEC-Q100 requirements by conducting additional device- and package-level stress tests. These additional tests may include, but are not limited to the following:
- power cycling
- liquid-to-liquid thermal shock testing, an accelerated way to test for endurance of thermal cycling and thermal shock
- instant solder shock
- 100% oxide stress tests
These rigorous tests validate the robustness of ams’ products before they are released to volume production.
In almost all cases, automotive manufacturers will also require suppliers to comply at a corporate level with industry quality standards such as ISO-TS16949. Conformance to this standard does not guarantee the integrity of any given part or manufactured unit, but it does give the customer confidence that the IC manufacturer follows rigorous quality processes, and documents its conformance to them. Similarly, compliance with the industry’s PPAP is intended to give customers confidence that the processes for designing and manufacturing a production part are effective, that they are well documented, and that the part’s manufacturer consistently implements them throughout the part’s production lifetime.
The many varieties of automotive qualification
For sensor ICs, then, the term ‘automotive qualification’ has two different aspects: the part must meet relevant quality, reliability and safety standards, and it must also meet the specific functional requirements of the application. For any given type of sensor, such as position sensors or gas sensors, both aspects are dictated by the nature of the application in which it is to be used. This means that a sensor manufacturer may label its device as automotive qualified, but this does not guarantee its suitability for every automotive application. Automotive designers are thus well advised to look closely at the specifications, quality processes, and test results of the devices that they select for evaluation, and not to take the IC manufacturers’ claims that they are ‘automotive qualified’ as a reliable indication of their suitability for their specific application.
Jeff Viola is a Field Applications Engineer for AMS AG.
