Access Official Technical PDF

Share Summary:LinkedInEmail

In-vehicle chip testing presents a practical challenge for automotive engineers: electronic control units (ECUs) communicate across vehicle networks, but the integrated circuits inside those units may have no accessible path to their structural test circuitry once installed. Reusing controller area network (CAN) or FlexRay connections as a test access mechanism offers a way to carry test data to and from embedded chips during maintenance.

The approach connects external test equipment to the vehicle control network and uses a test controller at the target ECU to route data to the selected chip. It is intended for offline, low-level maintenance and diagnosis, not as a replacement for normal network communication or manufacturing test.

Why in-vehicle chip testing matters

Modern vehicles rely on multiple ECUs to manage functions related to safety, operation, and comfort. An ECU may contain a host processor, communication controller, memory, interfaces, and other intellectual property (IP) blocks. Sensors and actuators connect to these systems, while a vehicle control network enables communication among nodes.

Functional testing checks whether a circuit performs its intended tasks. Structural testing instead looks for faults in the circuit’s implementation, including defects introduced during fabrication or those associated with aging. This distinction matters because a chip can experience degradation even after it has passed post-fabrication tests.

Scan testing is a widely used design-for-testability (DFT) method for structural testing. In scan mode, selected storage elements act as a shift register. Test equipment shifts in stimuli, the circuit processes them, and scan cells capture the response. The response is shifted out and compared with an expected result. This gives test engineers a way to control and observe internal circuit behavior that ordinary functional operation may not expose.

However, access to DFT circuitry is often disconnected after manufacturing tests. Limited input/output pins and security concerns can make it impractical to retain dedicated external test connections in a deployed system. Vehicle networks can provide a communication path, but their usual network-level diagnostics do not, by themselves, perform structural tests on the semiconductors inside an ECU.

Using vehicle networks as a test access path

The proposed mechanism treats the network as a transport channel for test stimuli and responses. External test equipment sends data over the vehicle control network to the target ECU. An on-node test controller then directs the data to the selected chip and collects its response. The equipment compares that response with the expected result.

This is distinct from conventional onboard diagnostics, which monitor subsystem parameters and report status information. Extending diagnosis to the chip level can make existing structural-test infrastructure accessible while a chip is embedded in a larger system. The method is aimed at maintenance and diagnosis; it does not claim that routine network diagnostics alone can identify every semiconductor fault.

CAN test access: identifying test messages

CAN is a message-oriented, event-triggered network. Nodes broadcast frames, and receiving controllers use message identifiers to decide whether to accept them. Identifiers also determine arbitration priority: a lower-valued identifier has higher priority. CAN’s non-destructive bitwise arbitration allows higher-priority messages to continue while a lower-priority transmitter gives way.

For CAN test access, dedicated message identifiers distinguish test traffic and identify the chip to be tested. When an ECU receives a test message with the identifier assigned to one of its chips, it can switch that target into test mode. Test stimuli travel to the node in CAN frames, and the node sends response data back using a corresponding response identifier.

For a system with n target chips, the described scheme requires 2n message identifiers: one for stimuli and one for responses for each chip. In the illustrative network arrangement, test equipment sends a frame addressed for a chip at one node; that node’s test controller runs the operation, then returns the response in a frame accepted by the test equipment.

FlexRay test access: using scheduled communication

FlexRay organizes communication around recurring cycles and allocated time slots. Its communication cycle can include a static segment, a dynamic segment, a symbol window, and network idle time. Because transmission is scheduled, test data must use the target node’s appropriate communication opportunity rather than compete for access in the same way as event-triggered CAN messages.

The proposed FlexRay method uses the null frame indicator (NFI) to distinguish test frames from ordinary functional data. The payload begins with a byte identifying the target chip on the target node; the remaining bytes carry test data. The NFI setting differentiates test stimuli from test responses in the described frame scheme. The target ECU’s test controller routes received stimuli to the chip, then returns the captured response through a response frame during the relevant allocated slot.

Engineering considerations for deployment

Using a network that already connects ECUs can reduce the need for a separate physical test interface, but the method depends on access to the target’s on-chip test circuitry and an ECU-side test controller capable of routing data. The network protocol also governs how test traffic is identified and transported.

  • Separate functional and test operation: The target ECU is switched from functional mode into test mode for the procedure.
  • Respect network rules: CAN identifiers and arbitration, or FlexRay’s scheduled slots and frame structure, must be handled according to the relevant protocol.
  • Compare responses: Test equipment evaluates returned data against expected responses to determine whether the tested circuit behaves as anticipated.
  • Plan for access: DFT circuitry must remain accessible through the on-node controller even when dedicated external test pins are unavailable.

The approach is presented as operating through the existing network protocols, with no overhead during functional operation. In practice, its use is for a dedicated test mode rather than ordinary vehicle control traffic.

Conclusion

Reusing CAN or FlexRay for in-vehicle chip testing can extend structural diagnosis beyond the manufacturing stage. CAN uses dedicated identifiers to direct test stimuli and responses, while FlexRay carries test data within its scheduled communication structure. For automotive engineers and system integrators, the central implication is that vehicle networks may serve not only ECU-level communication and diagnostics, but also as a pathway to embedded DFT infrastructure during maintenance. Achieving that access requires suitable on-node test control and careful adherence to each network’s protocol.

Access Official Technical PDF

About The Author