Camera Instability on Nvidia Jetson Orin Nano Dev Board

Issue Overview

Users are experiencing instability when using MIPI cameras with the Nvidia Jetson Orin Nano Dev Kit (P3767-0005) and the Commercial module (P3767-0003). The problem manifests as errors when attempting to read images from both cameras simultaneously. Specifically, while the P3767-0005 board operates stably under these conditions, the P3767-0003 board quickly encounters errors, leading to frame discards and system crashes.

Symptoms:

• Errors logged include "corr_err: discarding frame" and "uncorr_err: request timed out."
• System becomes unresponsive, affecting network, mouse, and keyboard functionality.
• The issue occurs consistently when both cameras are used at a resolution of 5M@20fps.

Context:

• Both setups utilize the same P3768 carrier board and NVMe SSD with JetPack version 5.1.3.
• The issue arises during image capture using v4l2-ctl commands.

Impact:

The instability significantly hampers user experience by preventing reliable camera operation, essential for applications relying on simultaneous image capture.

Possible Causes

1. Hardware Incompatibilities:

   • Differences in hardware design between the P3767-0005 and P3767-0003 boards may lead to performance discrepancies.

2. Software Bugs:

   • Potential bugs in the JetPack version being used could contribute to instability.

3. Configuration Errors:

   • Incorrect settings in camera configurations or system parameters may lead to performance issues.

4. Driver Issues:

   • Outdated or incompatible drivers for the MIPI cameras could cause errors during operation.

5. Environmental Factors:

   • Power supply inconsistencies or thermal issues could affect camera performance.

6. User Errors:

   • Misconfiguration of camera settings or improper use of commands may lead to unexpected behavior.

Troubleshooting Steps, Solutions & Fixes

Step-by-Step Instructions:

1. Clean Log Files:

   • Clear existing trace logs to reduce clutter:

     echo > /sys/kernel/debug/tracing/trace
     

2. Capture New Logs:

   • Run the cameras and capture new trace logs:

     cat /sys/kernel/debug/tracing/trace > t.log
     

3. Adjust Capture Timeout:

   • Increase the capture timeout in your configuration file:

     #define CAPTURE_TIMEOUT_MS 2500*4
     

4. Test Camera Modes:

   • Change camera output settings to continuous clock mode if currently set to discontinuous mode, as this has shown improved stability for some users.

5. Isolate Hardware Issues:

   • Test with different power supplies or cables to rule out power-related issues.
   • If possible, swap components (e.g., use a different NVMe SSD) to identify any faulty hardware.

6. Update Drivers and Firmware:

   • Ensure that all drivers are up-to-date and compatible with your current JetPack version.

7. Consult Documentation:

   • Refer to Nvidia’s official documentation for any updates regarding known issues with MIPI camera support on the Orin Nano platform.

8. Community Support:

   • Engage with community forums for additional insights or solutions that other users have found effective.

Recommended Practices:

• Regularly check for updates from Nvidia regarding JetPack and driver releases.
• Document any changes made during troubleshooting for future reference.
• Consider testing with a single camera setup before scaling to multiple cameras to ensure stability.

Unresolved Aspects:

Further investigation may be needed into specific error codes logged during system crashes, as well as potential hardware differences not yet fully understood between the two boards in use.