SOLVED!!![finally]: Seiken Densetsu 2 -> Secret of Mana won't save

[Arkanix38]

Hi,
I'm still having trouble with my Secret of Mana repro. I'm using the expanded 32Mbit expanded script extension by FuSoYa, and just can't seem to figure out the SRAM mapping. The game appears to work perfectly fine except that the SRAM won't enable properly.

I currently have my MAD-1 wired up as follows
Pin
1 - N/C
2 - RAM /CS
3 - N/C
4 - ROM /OE
5 - SRAM Vcc
6 - Vcc
7 - VBATT
8 - GND
9 - /RESET
10- Vcc
11-/CART
12-SNES #47 (BA6)
13-A21
14-A14
15-A13
16-N/C

Essentially i have tried to copy how Chrono Trigger has the MAD-1 wired.
So is my wiring bogus or has my SRAM chip died?

EDIT - Board is an original Seiken Densetsu 2 SHVC-2J3M-01 that I have swapped the MAD-R for a MAD-1 (tested, working chip)

[xIceMan]

13 - A21 (#46)
12 - A22 (#47)
11 - SNES /OE (#49)
09 - SNES RESET (#26) and RAM #26 (only in 6264)

Did you connect those as follows?

What EPROM/Flash did you use? Is MaskROM Pin 35 connected to (#48 A23)?

[Arkanix38]

Yes, The MAD-1 is connected to those signal lines through the PCB. I have lifted Leg 13 to connect it to A21, but otherwise all other pins are connected appropriately (with correct continuity checked)

[xIceMan]

Any pictures of the actual board would help alot more.

Also, check the battery if it's 3V?

[Arkanix38]

MAD-1 socketed for testing, all pins showing correct continuity. MAD 13 lifted for A21

Wired connections:
SNES#46 -> ROM A21 -> MAD 13
ROM A22 -> GND
ROM A20 -> MAD 13 socket hole (SNES#45)
ROM /OE -> MAD 4



Battery shows 3.06V on my crappy multimeter

I have tested the MAD-1 chip in the unmodified SD2 cart to the left, and it saved perfectly fine. Could the SRAM on the modded board be dead?

[xIceMan]

Hard to see anything in this picture but:

Is /WE from the 29F032/33 connected to VCC or anything else?

Why did you tie A22 MaskROM to GND? Remove that. Or try connecting it to cart #48


Other than that, try another SRAM if nothing helps.

Also, I don't recommend 32 pin MaskROM boards for TSOP Adapters.

[Arkanix38]

/WE on the ROM is connected to VCC and nothing else.
I removed the wire from A22 to GND. A22 is not needed on the ROM as it is only able to address up to A21 (max 32 mbit), so i pulled the adapter pin A22 to GND as a redundancy.

After checking /WE and removing the A22 wire, the game still won't save. I found a possible SRAM replacement here, please check if this is ok. The only difference to the 6265 in the cart is the lack of a CS2 on pin 26 (which shouldn't be required as its always pulled high during operation anyway)

[xIceMan]

Don't use this SRAM. Use one that is directly pin compatible with a 6264 SRAM.

[Arkanix38]

it is pin compatable, the only difference is the complete lack of need for OE₂ which is NC on the chip I provided. It runs at the right voltage, the data and address pins are ordered correctly and has the correct footprint. I don't see the problem?

[Arkanix38]

PROBLEM SOLVED: SRAM chip was faulty. Replaced with the AMIC chip i linked and the game works perfectly and saves perfectly.

Maybe spoke too soon. Put the cart back in the shell and the save was wiped after testing. Tried saving in game and had the same old issue. I have literally no idea WTF is happening, switched the SRAM and it saves fine. Put it back in the case and suddenly saves won't work again.

[xIceMan]

Sounds like a problem with the /CE lines to the SRAM. Have you checked the resistors and capacitors, too?

[Arkanix38]

I've finally gotten this to work. I've reverted to using an unmodified US ROM rather than the FuSoYa Variable text width hack. I've used a brand new cart and swapped the MAD-R for a MAD-1 from my spare SFC Chrono Trigger (don't worry it still works fine) and the game booted first time.

I think I may have burned out a trace or damaged a chip on the original board I was using, as with the original ROMs the game no longer boots at all. From the lack of ability to save, I believe that perhaps one of the address lines controlling bank switching on the MAD-1 may have not been connecting properly. I'm going to desolder the MAD-1 to use later and keep the board in my parts bin.

I'm glad I finally have an original Seiken Densetsu 2 cart running SoM, but i'm rather sad that I killed a board in the process (even though it only cost me around $3).

I may switch the MAD-1 from the dead board into my Chrono Trigger to see if it still works fine, and if it does then the issue is definitely the PCB. Thank goodness I have a desoldering gun now, makes all this repeated removal of chips MUCH safer as I don't need to reheat a joint 10 times to suck out all the solder with wick.