SNES: converting 2 mask rom board to 1 flash rom

Just a quick question about the process explained here in converting a Jap Mother 2 cart to an Earthbound one.

I understand the rewiring regime: reconnect A21 of the new chip to the SNES A21 line, bypassing the MAD-1, then connecting the new chip's /OE line to pin 4 of the MAD-1 (which I assume keeps it always active? I think that pin is used in single Mask ROM arrangements that use a MAD-1).

My question regards the fact that the mod doesn't remove the P1 chip, nor does it disconnect it in any way from the PCB - won't there be a conflict when the SNES looks for data that is in the top part of the ROM? I know that the new chip contains the entire game, but won't the MAD-1 still do it's job and turn on P1 when the top 8Mb of rom is being accessed? Won't that mean that both the new chip and the old P1 chip would be trying to assert the data lines simultaneously? Maybe I am just not understanding this properly.

It's weird, because I know this mod works, and I've done it on several other games (like FFVI), but I don't quite understand why...
Cheers

 

Wouldn't it be easier to just use a single MaskRom PCB?

 

Modding the BJ3M SNES PCB this way makes no sense to me.
The 2MB 29F016s are just $1.5 a piece. So two 29F016 would be $3.
A 4MB 29F032 sets you back at least $6.
Meaning you pay twice the money and then also have to rewire the pcb? Just use two 29F016 instead.

But to answer your initial question: The old chip is inactive after rewiring. It won't do anything.

 

Well, you'd have to solder 2 tsop chips, so there's a bit more work involved. And iirc Jaska had a reasonable source for the 4MB chips
On this board I would cut the upper maskrom P1 leg 33, bend it up and solder it to the 5V pin 34 adjacent to it. That'll ensure this ROM isn't enabled ever.

 

I think once you remove the connection to PIN 16 of MAD1 it will never try to switch to the second Mask ROM, so essentially turns it into 1XM-XX

 

Interesting, does the MAD chip detect that the ROM has been disconnected from Pin 16 and behave as a single chip MAD? It just seems weird to me that the 2nd chip is still connected as it was in the 2 chip arrangement.

This is what I would assume needed to be done to deactivate the 2nd chip, but apparently it isn't necessary and not I don't really get why.

 

It's not a very accurate guide if you ask me.

Pin 16 on the MAD-1 is an output, whether this floats or is hooked up doesn't affect how the MAD-1 behaves.

I have seen occassions where incorrect configurations that allowed multiple chips to drive the same bus actually worked on the SNES. Sometimes when I forget to cut the maskrom /OE and tie it to 5V, the cart might still work. My best guess is that the EEPROM drives the data lines stronger than some maskroms. But the only proper way is to tie the unused chip's /OE to a permanent high signal.

 

That's what I thought.

Seems reasonable to me, I think you're right that the only way to be sure it will work is to tie the unused chip's OE line high. Thanks for your input.

 

Any updates?