Recommended Troubleshooting Steps
1. Verify Power
Check first because most “dead” units are power related.
Confirm:
- Input voltage is correct and stable
- No voltage sag during power-up
- Power rails sequence correctly
- PMIC rails are present
Look for:
- No current draw → missing power
- Current spikes then drops → short/reset loop
- Continuous reboot cycling → unstable rail/watchdog
If possible:
- Use oscilloscope instead of multimeter
2. Connect Serial Debug Console (Critical)
Connect to the debug UART:
- 115200 baud
- 8 data bits
- No parity
- 1 stop bit
Observe:
- ROM boot messages
- U-Boot output
- Kernel logs
- Error messages
This immediately tells you whether:
- CPU is alive
- Boot ROM is executing
- Bootloader is corrupted
- Linux is failing
3. Check for Any UART Output
Case A — No Output at All
Most likely:
- Power issue
- Reset held low
- Boot straps incorrect
- Hardware fault
Check:
- RESET_B
- POR_B
- Clock oscillator
- Boot configuration resistors
Case B — U-Boot Appears
Hardware is mostly OK.
Problem is likely:
- Corrupted Linux image
- Bad rootfs
- Bad DTB
- eMMC corruption
Proceed to reflashing.
Case C — Continuous Reboots
Possible causes:
- Watchdog resets
- DDR init failure
- PMIC instability
- Thermal issue
4. Attempt USB Recovery Mode
The i.MX8X ROM supports USB recovery.
Use:
- NXP UUU tool
- Digi recovery image
If recovery mode works:
→ CPU and ROM are healthy
→ eMMC or bootloader likely corrupted
If recovery mode does NOT work:
→ likely hardware/power/reset issue
5. Reflash the Unit
Reflash:
- SPL
- U-Boot
- Kernel
- Device tree
- Root filesystem
Use:
- Digi Embedded Yocto recovery image
- Known-good production image
A corrupted eMMC image is one of the most common causes.
6. Verify Boot Configuration
Check:
- Boot mode resistors
- eMMC boot selected
- Recovery mode not forced unintentionally
On custom carrier boards:
- Wrong pull resistor values are common
7. Reduce to Minimal Hardware
Disconnect:
- USB devices
- Ethernet
- PCIe peripherals
- Displays
- Expansion modules
Boot only with:
- Power
- SOM
- UART
This isolates peripheral-related hangs.
8. Inspect Hardware
Check for:
- Overheating PMIC
- Damaged capacitors
- Shorted rails
- Mechanical damage
- Poor SOM seating
Use thermal camera if available.
9. Compare Against Known-Good Unit
If possible:
- Swap SOM onto known-good carrier
- Swap carrier board
- Compare UART logs
This quickly isolates SOM vs carrier fault.
Recommended Fastest Diagnostic Path
The fastest sequence is usually:
1. Verify rails
2. Attach UART
3. Capture boot log
4. Attempt USB recovery
5. Reflash image
That typically identifies the issue within minutes.
If you want, I can also help with:
- interpreting UART boot logs
- USB recovery procedure using UUU
- reflashing ConnectCore 8X
- identifying boot strap settings
- debugging custom carrier boards
- diagnosing watchdog/reset loops
Last updated:
Sep 30, 2026