Short lasted ROM-CPU communication symptom

[GelatinCapacitor]

I have encountered two devices which do not function; they do not turn on.
The only signal of life present, is communication between the ROM and the main processor, and the presence of oscillation on crystals.

Using the functions of a multimeter and a handheld oscilloscope, I have monitored the relevant signal lines towards the processor and power supply, tested components and ensured stable connections.

The results are stable and clean supply voltages, behaved relevant reset, standby and power-up signals, good electrolytics, good ceramics, good resistors, stable connections.

All my tests are using the HDS200 oscilloscope in 100ms scan mode, peak detection mode, and only used auto and single shot modes to analyze for details using the provided counters and cursors. This oscilloscope does not allow to decode digital data.



Device 1:
This device has circuitry for processor reset, which is working correctly. Just in case, my tests included manually enabling and disabling the supply voltages for the relevant ICs.

Behavior:

1: Device is disconnected from mains and ensured no voltages are present in the circuitry.
2: On connection to mains, digital activity between the main processor and ROM occurs, lasting for around 100ms. During this time, the processor's quartz crystal oscillates at the correct frequency, then stops.
3: No digital activity or oscillations ever occur again. No other expected signal from the device's expected behavior occurs.
4: Pressing some of the device's keys (also manually activated in case of damaged key buttons), for a period of around 3 seconds, causes logic levels on lines between the processor and ROM to swap states (no oscillations), then everything returns to previous state. The processor's crystal also oscillates at correct frequency during this period.
5: The device does not respond to any of the service modes and functions described in the service manual. No other expected signal outcoming from the the processor are present at any point.

Device name: HTR-2064



Device 2:
This device doesn't have circuitry for processor reset. Just in case, my tests included manually enabling and disabling the supply voltages for the relevant ICs.
When connecting the device to mains, there's like a 1 in 10 chance that the device actually works correctly. The other times, the behavior is decribed:

Behavior:
1: Device is disconnected from mains and ensured no voltages are present in the circuitry.
2: On connection to mains, digital activity between the main processor and ROM occurs, lasting for around 100ms.
3: Afterwards, the processor shows no other activity or signals, ever.
4: The processor's crystal has the corect oscillation frequency present at all times with no interruptions.
5: Manually reseting power to the ROM causes 100ms-lasting CPU-ROM activity to occur again.

Good behavior:
1: When the device decides to work correctly, on connection to mains the communication between the ROM and CPU lasts for an exteded period.
2: When device is in working mode, CPU-ROM communication stops, but very short (unmeasured) bursts occur every around 2 seconds.
3: When device is in stand-by mode, CPU-ROM communication is occuring constantly.
4: The device continues to work correctly until disconected from mains.

Device: A basic 32-inch SAMSUNG LCD TV. (sorry, forgot model since couldn't find service manual). The ROM was a 16-pin 25Q[...] Winbond.



My best guess for this problem is a malfunctioning processor or ROM.