Hardware
Programmer: GBxCart RW v1.4c
Firmware: R42+L15
Software: FlashGBX v5.0.1
Platform: Windows
Flash cart: FunnyPlaying MidnightTrace Flash Cart (current revision, no DIP switches)
Purchased directly from FunnyPlaying.
Flash Cart Components:
*Flash ROM: Spansion S29GL064N90TFI04
*FRAM: Ramtron FM28V100-TG
*FPGA: Gowin GW1NZ-LV1
*RTC battery installed (also tested with battery installed throughout testing).
ROM Tested:
Pokémon Silver (MBC3)
ROM Size: 2 MiB
Save Type: 256 Kbit SRAM (32 KiB)
Mapper: MBC3
I also tested with a second Pokémon Silver ROM from a different source with identical results.
FlashGBX Configuration:
*Profile: FunnyPlaying MidnightTrace Flash Cart
*ROM Size manually set to 2 MiB
*Save Type manually set to 256 Kbit SRAM (32 KiB)
*Mapper manually set to MBC3
Problem:
The cartridge erases successfully every time.
Programming begins normally but fails during the write process with:
"Writing at 3.3V failed. Do you want to retry at 5V?"
I have not retried at 5 V because Analyze Flash Cart reports the flash device as 2.7–3.6 V.
The failure point is inconsistent.
Observed failures:
~4%
~10%
~11%
~24%
~78% (once)
After a failed write, the ROM header is partially programmed and FlashGBX can identify the ROM title as POKEMON_SLV, indicating that programming begins successfully before failing.
Analyze Flash Cart
Detected profile:
FunnyPlaying MidnightTrace Flash Cart
CFI:
*Device size: 8 MiB
*Voltage: 2.7–3.6 V
*Buffered write supported
*Sector erase:
*Region 1: 0x00000000–0x0000FFFF
*Region 2: 0x00100000–0x07FFFFFF
Flash ID after erase:
ROM:
FF FF FF FF FF FF FF FF
WR /AAA/AA:
01 01 7E 7E 00 00 0A 0A
WR /4AAA/AA:
01 01 7E 7E 00 00 0A 0A
After a failed write:
ROM:
F3 C3 00 01 00 00 00 00
WR /AAA/AA:
01 01 7E 7E 00 00 0A 0A
WR /4AAA/AA:
01 01 7E 7E 00 00 0A 0A
Troubleshooting Performed:
Updated to FlashGBX v5.0.1
GBxCart RW firmware updated to R42+L15
Multiple USB-C data cables tested
Different cartridge shells tested
Cartridge contacts cleaned
Multiple Pokémon Silver ROMs tested
Cart erased before programming
Analyze Flash Cart run before programming
Correct MBC3 configuration used
Did not retry programming at 5 V
Question
Is this behavior expected with the current revision of the FunnyPlaying MidnightTrace Flash Cart, or does this indicate a compatibility issue or defective cartridge?
I'm new to Game Boy ROM programming, so if I've overlooked something obvious, please let me know. I'm happy to perform additional tests or provide more screenshots, logs, or hardware photos if they would help diagnose the issue.
Update: I was able to successfully program the MidnightTrace using FlashGBX v4.3. I manually selected 3.3 V in the voltage prompt. The initial write completed but verification failed in 9 sectors. I rewrote the failed verification ranges individually until the ROM passed verification completely. The same cart/programmer/ROM combination consistently failed during programming under FlashGBX v5.0.1. This suggests a possible regression or behavioral difference between v4.3 and v5.0.1 rather than a defective flash cart.
Hardware
Programmer: GBxCart RW v1.4c
Firmware: R42+L15
Software: FlashGBX v5.0.1
Platform: Windows
Flash cart: FunnyPlaying MidnightTrace Flash Cart (current revision, no DIP switches)
Purchased directly from FunnyPlaying.
Flash Cart Components:
*Flash ROM: Spansion S29GL064N90TFI04
*FRAM: Ramtron FM28V100-TG
*FPGA: Gowin GW1NZ-LV1
*RTC battery installed (also tested with battery installed throughout testing).
ROM Tested:
Pokémon Silver (MBC3)
ROM Size: 2 MiB
Save Type: 256 Kbit SRAM (32 KiB)
Mapper: MBC3
I also tested with a second Pokémon Silver ROM from a different source with identical results.
FlashGBX Configuration:
*Profile: FunnyPlaying MidnightTrace Flash Cart
*ROM Size manually set to 2 MiB
*Save Type manually set to 256 Kbit SRAM (32 KiB)
*Mapper manually set to MBC3
Problem:
The cartridge erases successfully every time.
Programming begins normally but fails during the write process with:
"Writing at 3.3V failed. Do you want to retry at 5V?"
I have not retried at 5 V because Analyze Flash Cart reports the flash device as 2.7–3.6 V.
The failure point is inconsistent.
Observed failures:
~4%
~10%
~11%
~24%
~78% (once)
After a failed write, the ROM header is partially programmed and FlashGBX can identify the ROM title as POKEMON_SLV, indicating that programming begins successfully before failing.
Analyze Flash Cart
Detected profile:
FunnyPlaying MidnightTrace Flash Cart
CFI:
*Device size: 8 MiB
*Voltage: 2.7–3.6 V
*Buffered write supported
*Sector erase:
*Region 1: 0x00000000–0x0000FFFF
*Region 2: 0x00100000–0x07FFFFFF
Flash ID after erase:
ROM:
FF FF FF FF FF FF FF FF
WR /AAA/AA:
01 01 7E 7E 00 00 0A 0A
WR /4AAA/AA:
01 01 7E 7E 00 00 0A 0A
After a failed write:
ROM:
F3 C3 00 01 00 00 00 00
WR /AAA/AA:
01 01 7E 7E 00 00 0A 0A
WR /4AAA/AA:
01 01 7E 7E 00 00 0A 0A
Troubleshooting Performed:
Updated to FlashGBX v5.0.1
GBxCart RW firmware updated to R42+L15
Multiple USB-C data cables tested
Different cartridge shells tested
Cartridge contacts cleaned
Multiple Pokémon Silver ROMs tested
Cart erased before programming
Analyze Flash Cart run before programming
Correct MBC3 configuration used
Did not retry programming at 5 V
Question
Is this behavior expected with the current revision of the FunnyPlaying MidnightTrace Flash Cart, or does this indicate a compatibility issue or defective cartridge?
I'm new to Game Boy ROM programming, so if I've overlooked something obvious, please let me know. I'm happy to perform additional tests or provide more screenshots, logs, or hardware photos if they would help diagnose the issue.
Update: I was able to successfully program the MidnightTrace using FlashGBX v4.3. I manually selected 3.3 V in the voltage prompt. The initial write completed but verification failed in 9 sectors. I rewrote the failed verification ranges individually until the ROM passed verification completely. The same cart/programmer/ROM combination consistently failed during programming under FlashGBX v5.0.1. This suggests a possible regression or behavioral difference between v4.3 and v5.0.1 rather than a defective flash cart.