nesdev.com

I'd be more than happy to give this a go, but there are three main problems: first of all, there's no guarantee you'll end up with a cartridge with a 5A in it - it could be a 5B instead of a 5A, as has been known to happen since they appear to be interchangeable. Which brings me to the second problem: whichever the case, you'd need to add in your own circuit to the Gremlins cart, because the top row of the chip (which is where the audio pins are located in the 5B at least) isn't connected to anything. Although I'm sure most of you lot could handle that part, it is far beyond my own capabilities (I'm a musician, nothing more). Lastly, I'm not exactly rich myself, and semi-rare Famicom games like these can cost quite a bit of money I can't afford to spend right now, so I wouldn't be the one financing this project I'm afraid.

I do have a TNS-HFC3, however (it's in jarhmander's hands undergoing maintenance right now, but there's a good chance I'll have it back in working order). So if someone is willing to take the plunge and order one of these Gremlins carts, mod it and send it to me for testing, then sure I'll be glad to oblige - but there's probably not much I could do that someone with a Power Pak couldn't, anyway (swap tests etc.).

Hello

I actually bought two Famicom Gremlins 2 carts to use the 5b's audio, but sadly BOTH had 5a's. I've unsoldered the two rom chips on it and was about to put the Gimmick eproms on it but i got bored, and started hacking something else.

If there's real interest i could finish the work and confirm whether or not 5a generates audio.

Not sure why it makes it a sad discovery, but yes, if you can find out what the 5A is capable of doing, that would be great. A pin-out would be nice too.

Its sad because all I wanted was a 5B. In any case I bought 3 more Gremlins carts, so while five units is not a huge statistical sample, if they are all 5A's I would consider the nescartdb Gremlin's 5b a fluke, and I'll abandon the idea.

BTW I've completed my hack and the 5A doesn't seem to generate anything when playing Gimmick roms.
Does it generate anything else if i randomly write to its address space? No clue.

I'll post more info on my blog soon.

I can confirm, that Gremlins(J) has 5A, not 5B.

After opening _five_ Japanese gremlins carts, they were _all_ 5a's.

While bootgod's scan is a 5b:
http://bootgod.dyndns.org:7777/profile.php?id=3805

Huge monster fluke that was...
I think i will now abandon this silly idea

I would bet that using similar schematics to that of the FME-7 audio addition, the AY-whatever chip could be added for this functionality.

I've no doubt. But I wanted the real generator for my own analysis.

Maybe send in the other versions to bootgod's site.

_________________
Here come the fortune cookies! Here come the fortune cookies! They're wearing paper hats!

my research for mapper 69

sunsoft 5a can't bank any to 6000~7fff?
sunsoft 5b can bank prg to 6000~7FFF?
sunsoft fme-7 can bank prg to 6000~7FFF and can bank sram to 6000~7fff？

Given that Gimmick! will run on my FME-7 socketed board, other than the audio does the 5B present much else in differences to FME-7?

Looking at the register layout of the FME-7, only one thing comes to mind to test: given the differences between the Namco 163/175/340, it's remotely plausible that pseudoregister 8 (controlling the bank at $6000) might behave differently? (See also byemu's post) I see no reason to suspect that any other behavioral differences will exist.

It's remotely conceivable that the FME-7 and 5a might do something funny with register writes to $C000 and $E000. (what? who knows. maybe pins 3 and 18 are digital outputs).

It would be interesting to know, since you've got the sockets already, whether pseudoregister 8 still drives lines PRG ROM A13…A18 according to the latched value when it's configured to RAM. (in other words, it would be nice to know if the musing on the wiki about bankable prg-ram is true or not)

I have three carts,Barcode World (J),Batman (J),Gremlin 2 - Shinshu Tanjou (J).
I'm try to debug on kazzo.
The results are:
Barcode World(FME-7) only bank sram to 6000~7fff (can't bank prg-rom to here)
Batman (5a) can't bank anything to 6000~7fff (no sram on this game)
Gremlin 2(5a) can bank prg-rom to 6000~7fff (be marked as 5b in "NES Cart Database")

The circuit of Batman and Gremlin 2 on prg is exactly the same.

So you're saying that two different cartridges, both with a Sunsoft 5a, respond to writes to register 8 differently?

Is it correct to say that the Batman cartridge is entirely ignoring writes to register 8?

Does Gremlins 2 pay attention to the top two bits of register 8? (In other words, can it ever not map prg-rom in?)

CNROM and AOROM use the same IC, but they act very differently. Several cartridges with certain Konami VRC series ICs respond differently based on which lower address lines are connected to the mapper's register select lines. There are about 14 possible combinations, with nearly half of them attested. Likewise, it's possible to have the chips ignore certain enable signals output by the mapper by playing with how the signals are routed on the PCB.