Hello everyone,
I'm working on a Samsung Sound Tower MX-T70 (1500W Party Audio) that came to me completely dead, showing no signs of life. Since I couldn't find any schematics or service manuals, I did some deep diagnostics and measurements.
Here is a summary of what I have checked and done so far:
At this stage, I suspected a firmware corruption—possibly caused by the power supply pulsing or the short circuit during previous operation.
The firmware resides on a GigaDevice GD25Q32SCP (32Mbit / 4MB SPI NOR flash). I desoldered it, read the content using an XGecu T48 programmer (the read and verify process completed with no errors), and flashed the exact dump onto a brand-new, identical chip. Soldering the new chip back didn't change the behavior.
Next, I hunted for an UART interface to see what the CPU was doing. I found an unpopulated 4-pin header near the Bluetooth module area (XP16). After connecting a USB-to-TTL converter (configured to 3.3V logi levels at 115200 Baud), the terminal output confirmed my suspicions.
Upon plugging the unit into AC, the console prints exactly this and then completely freezes:
Plaintext
[AV] Samsung Party T70
1000.0 >
This output proves that the main SoC is getting the correct voltages, the clock signal is running, and the SPI bus is functionally active, as the CPU successfully reads the model identifier string and the firmware version (1000.0 or 1007.0) from the header of the flash memory.
However, right after printing the prompt > character, the boot process halts completely. The console is non-interactive (does not accept RX inputs/commands), and pressing the physical power button triggers no change or further logs.
It strongly appears that the main firmware binary is corrupted right past the initial bootloader header.
Unfortunately, Samsung only provides standard USB update packages on their support page, which requires a functioning system to initiate. I have searched everywhere but cannot find a raw SPI dump for this model.
Does anyone happen to have a known-good working SPI Flash Dump (.bin) for the Samsung MX-T70?
Any help, dump files, or further advice would be highly appreciated!
Thank you in advance!
Janos
I'm working on a Samsung Sound Tower MX-T70 (1500W Party Audio) that came to me completely dead, showing no signs of life. Since I couldn't find any schematics or service manuals, I did some deep diagnostics and measurements.
Here is a summary of what I have checked and done so far:
- Power Supply: The standby voltage is present and stable at 12.3V–12.5V. The 3.3V rail is also active.
- Power Button: The power key pin receives 3.3V and successfully drops to ~0.3V when pressed, so the signal reaches the board.
- PSU Forced Start: I identified the Power-On pin on the 14-pin cable connecting the PSU to the mainboard. By injecting 3.3V into this pin (with the mainboard disconnected), the power supply wakes up fully, and the main amplifier rail appears, delivering a stable 56V DC.
- Previous Hardware Issues Fixed: The unit previously had a failed primary-side startup capacitor (CE8) on the standby rail, causing a hiccup/pulsing effect under load. Additionally, the ribbon cable going to the sub-board (I/O & BT module area) was physically damaged, shorting Pin 1 (GND) to Pin 2 (12.3V Standby). I successfully replaced the ribbon cable, repaired the standby power supply, and now all voltage rails are completely solid and stable.
At this stage, I suspected a firmware corruption—possibly caused by the power supply pulsing or the short circuit during previous operation.
The firmware resides on a GigaDevice GD25Q32SCP (32Mbit / 4MB SPI NOR flash). I desoldered it, read the content using an XGecu T48 programmer (the read and verify process completed with no errors), and flashed the exact dump onto a brand-new, identical chip. Soldering the new chip back didn't change the behavior.
Next, I hunted for an UART interface to see what the CPU was doing. I found an unpopulated 4-pin header near the Bluetooth module area (XP16). After connecting a USB-to-TTL converter (configured to 3.3V logi levels at 115200 Baud), the terminal output confirmed my suspicions.
Upon plugging the unit into AC, the console prints exactly this and then completely freezes:
Plaintext
[AV] Samsung Party T70
1000.0 >
This output proves that the main SoC is getting the correct voltages, the clock signal is running, and the SPI bus is functionally active, as the CPU successfully reads the model identifier string and the firmware version (1000.0 or 1007.0) from the header of the flash memory.
However, right after printing the prompt > character, the boot process halts completely. The console is non-interactive (does not accept RX inputs/commands), and pressing the physical power button triggers no change or further logs.
It strongly appears that the main firmware binary is corrupted right past the initial bootloader header.
Unfortunately, Samsung only provides standard USB update packages on their support page, which requires a functioning system to initiate. I have searched everywhere but cannot find a raw SPI dump for this model.
Does anyone happen to have a known-good working SPI Flash Dump (.bin) for the Samsung MX-T70?
Any help, dump files, or further advice would be highly appreciated!
Thank you in advance!
Janos
Comment