Yamaha Sound Chip for the 6502

[EvilSandwich]

I know that it a bit early for me to be thinking about this kinda thing since I'm quite new to building a computer, but I've been slowly but surely thinking of designs in my head for different pieces of hardware I could use in my future project. I've been sort of fixated on the idea of putting in a Yamaha sound chip into a 6502 machine and seeing if I can get it to play sound effects and music.

My first instinct was to use a YM2612 because of my fond memories of that very distinct "Genesis Sound", but it seems that all the documentation I found on it really seems to insist that I use it on a Z80 or 68000 instead. Is there any reason why a 6502 CPU wouldn't be ideal?

If it doesn't pan out, I found a few alternatives. Like the YM3526, that was used in the Commodore 64's sound expander or the YM3812 that was in the old AdLib cards.

I would hate to just use a SID chip, since just using the same sound chip of a pre-existing 6502 computer would kind of defeat the purpose of the project. But if its all I can do, it can't be helped, I guess.

[GARTHWILSON]

EvilSandwich wrote:

I would hate to just use a SID chip, since just using the same sound chip of a pre-existing 6502 computer would kind of defeat the purpose of the project. But if its all I can do, it can't be helped, I guess.

More than the old SID chips, I just ran into the FPGASID 6581 and 8580 SID replacement available to buy, based on a modern FPGA. It has cycle-exact reproduction of the original parts, exact modeling of the analog parts leading to a result that is almost indistinguishable from the original, but adds extra features like wave tables, stereo sound, more voices, etc, while keeping full compatibility.

[commodorejohn]

I don't think there's any compelling reason you couldn't go with a 2612 - people just discuss it in terms of the 68k and Z80 since it was used in the Genesis. Yamaha's chips are pretty agnostic; the real challenge in incorporating them into a computer is the mildly obscene delay times required between register writes.

[DerTrueForce]

Having found a page out of a datasheet, it looks like it uses an Intel/Z80 bus, which has to be adapted to the 6502 bus, but it's really easy to do that; just a couple of NAND gates, similar to the 68K.
It looks like there's no actual reason you couldn't use a YM2612 with a 65c02(or a 65-anything).

[EvilSandwich]

GARTHWILSON wrote:

More than the old SID chips, I just ran into the FPGASID 6581 and 8580 SID replacement available to buy, based on a modern FPGA. It has cycle-exact reproduction of the original parts, exact modeling of the analog parts leading to a result that is almost indistinguishable from the original, but adds extra features like wave tables, stereo sound, more voices, etc, while keeping full compatibility.

Now I kinda wanna design a computer with that too, just to see which sound I like better. The Commodore 64's sounds are pretty nostalgic in their own right. haha

commodorejohn wrote:

I don't think there's any compelling reason you couldn't go with a 2612 - people just discuss it in terms of the 68k and Z80 since it was used in the Genesis. Yamaha's chips are pretty agnostic; the real challenge in incorporating them into a computer is the mildly obscene delay times required between register writes.

Yeah, I was reading about that and thinking how to get around that. Maybe some kind of data buffer between the controller and the sound chip?

DerTrueForce wrote:

Having found a page out of a datasheet, it looks like it uses an Intel/Z80 bus, which has to be adapted to the 6502 bus, but it's really easy to do that; just a couple of NAND gates, similar to the 68K.
It looks like there's no actual reason you couldn't use a YM2612 with a 65c02(or a 65-anything).

I didn't realize that the 68K had to be adapted to fit the YM2612's bus. I guess the folks at Sega just really liked the power of the 68K. I always found it kinda cool how the Genesis had both the Z80 and the 68K at the same time.