Engine control units are present in many applications today, often involving safety-critical applications that demand highly predictable behaviour and reliable operation in environments such as avionics, automotive and the operation of heavy machinery in freight terminals. These environments are highly safety sensitive, the failure of an ECU (electronic control unit) to act in an appropriate manner under emergency conditions could pose a threat to life and/or property. Ensuring these demands are met requires significant investment in test and documentation.
There are many examples of where the safety-critical operation of an ECU is important.
An example of where fault insertion is currently used in the automotive industry, is part of the overall testing of powertrain control modules (PCMs). The PCM is one of the most complex electronic control units in the modern vehicle, requiring a rigorous and thorough testing of its functionality. The consequences of PCM failure will have greater significance in X-by-Wire applications (a collective term for the addition of electronic systems into a vehicle to enhance and replace tasks that were previously accomplished via mechanical and hydraulic systems such as braking or steering), placing increased importance upon these test methods.
SPECIAL TEST METHODSDue to the high level of sophistication and complexity of today's ECU devices, special test methods are required. Fault insertion testing is an important aspect of ECU validation. The idea of testing for system failures is not new—it is an important aspect of ECU testing and involves the introduction of electrical faults into a system. The simulation typically duplicates various conditions, which could occur because of corrosion, short/open circuits and other electrical failures inherited through age, damage or even faulty installation.
Traditional test methods often involve the manual insertion and extraction of cables to and from a patch panel, which is far from ideal. Not only is this method of testing prone to human error, it is time consuming. Targeting ECU validation, the Pickering Interfaces fault insertion BRIC switching solution enables a sophisticated testing approach for these realworld scenarios.
Typically, ECUs under development are exercised by a test system, which simulates the engine that the unit will control—this is sometimes called a hardware-inthe- loop (HIL) simulation test system. Stimulus instrumentation simulating engine behaviour is connected and controlled either by manual operation or computer, with measurement instrumentation used to capture analogue and digital responses from the ECU.
When it is necessary to inject faults many use a patch panel such as that in
Figure 1.The various cables shown may be used to connect any input/output (I/O) line on an ECU to stimulus or measurement instrumentation. The I/O lines may be disconnected to simulate an open-circuit or tied together to simulate short-circuits (to ground, voltage source or between I/O lines). An engineer can move the patch cables to simulate a desired fault and then measure the results. This type of solution has many inherent disadvantages.
An immediately obvious disadvantage is size, i.e. patch panels tend to be large. There are hidden costs such as maintenance, the need for knowledge on the part of the operator, potential human error and the cost of labour required to execute the test and record results. Over a period of time, patch panel maintenance issues are likely to arise through frequent use.
Clearly a disadvantage of any manual method is the lack of repeatability. The ability to quickly reproduce a failed test condition is essential in a test system, either to aid development or to take corrective action. Being able to precisely reproduce the test procedure quickly is a major advantage in any upgrade or verification programme.
The ability to gain software control of both instrument routing and the insertion of real-time electrical faults enhances both the testing process and the recording of the outcome. However, although a standard crosspoint matrix with adequate an specification is capable of handling the instrument routing to the device under test, the insertion of faults requires a different switching architecture that was previously not available.
THE BRIC APPROACHThe Pickering Interfaces BRIC provides a large range of ultra highdensity matrix modules in a 3U PXI form factor. Each BRIC uses a backplane, which accepts a set of daughter cards. The backplane carries an analog bus between the daughter cards and the control lines for the relays. A single PCI interface is used to interface between the BRIC backplane and the PXI backplane in the chassis, so the module uses just one electrical slot in the chassis. Solutions are available that occupy 4 or 8 mechanical slots. The 4-slot matrix
(Figure 2) may be configured to a maximum of 2208 crosspoints, the 8-slot matrix to a maximum of 4416 crosspoints. Despite its modular construction the BRIC is considered as a single matrix entity to a system, making it much easier to control and requiring no user configuration effort.
FAULT INSERTION BRIC SOLUTIONFor fault insertion test applications Pickering Interfaces has developed the fault insertion BRIC, a scalable solution which may be used in place of a patch panel to switch signals from simulated and real-life devices in a HIL system. HIL testing enables the user to put an ECU through test scenarios identical to those carried out in "engine test stand" testing. This BRIC switching solution can help to simplify and accelerate the testing, diagnosis and integration work in HIL applications. The Fault Insertion BRIC is available with maximum switching capacities of 1 and 10A. A typical Fault Insertion BRIC application is to assist in routing electrical fault simulation to high pin-count ECU's in automotive and aerospace applications. Typical simulated faults include those found in cable harnesses such as opencircuits and short-circuits (to ground, to battery or between I/O signal lines).
A simplified functional schematic of a Fault Insertion BRIC is shown in
Figure 3. Fault insertion and measurement is performed via the Y-axis and connection to the ECU is via the Xaxis. The X-axis has a breakout facility (3-pin in this illustration), allowing the interruption of I/O signals to the ECU.
The Fault Insertion BRIC improves the method of error injection, monitoring and self-test in various test and simulation systems. Using the Fault Insertion BRIC both manual and programmatic access to each signal line can be provided. The Fault Insertion BRIC provides the user with a powerful solution for routing simulated faults to the ECU with guaranteed repeatability including faults such as:
- Open-circuits simulating cable breaks between an ECU and its sensors or actuators,
- Short-circuits to ground,
- Short-circuits to either a battery or an external voltage source,
- Short-circuits between I/O signal lines, and
- Partial connection between signal paths.
FAULT INSERTION BRIC MODELSThe 40-592 is a very high density Fault Insertion BRIC solution
(Figure 4), handling small to mid-range signals. Based on instrumentation grade ruthenium sputtered reed relays, this model has 1A (150Vdc/100Vac, 20W) switching capacity and long operating life. It is available in both 4- and 8-slot mechanical form factors. The 40-592 can be specified in a wide variety of matrix sizes to suit the intended application by having only some of the switching modules installed. A total of 24 configurations are available for the 40-592 with either 2- or 3-pin breakout options. Maximum matrix sizes are 2488 for the 2-pin breakout option and 1608 for the 3-pin breakout option. Larger matrices can be constructed by simply daisy-chaining modules.
The 40-595 is a high power Fault Insertion BRIC solution
(Figure 5). Using high quality gold plated electromechanical relays; this model has 10A (125Vdc/ 250Vac, 240W/2000VA) switching capacity and occupies 8 slots of a 3U PXI chassis. It is provided with 3-pin breakout facility as standard and various configurations are offered up to a fully populated 30x8 matrix.
