2A03 Revision H

[2600]

Does anyone know anything about this revision?
Has anyone swapped a revision H into a toaster NES?

[Xious]

Does anyone know anything about this revision?
Has anyone swapped a revision H into a toaster NES?

I'm pretty sure that it's a tooling revision and while it's instruction compatible, it may have a different pinout. Best to check the schematics on the NES to the pinouts on the Rev H 2A03 before trying it. Even so, it won't gain you any benefits.

UPDATE: It should be the same pinout as the Rev G usuallyh found in "toaster" NES consoles. From what system did you pull it, a Famicom or an NES2? The Rev H was usually used in the Famicom A/V if memory serves.

I don't think the differences between the different Revs are fully documented, but you could always call or write to RICOH and ask for a devsheet

-Xious

[AWal]

I have swapped H and G chips from time-to-time. They seem identical, I'd assume the major difference is process type and method; The H series seems to have a better matte finish on the chip...Then again, all the H chips I've worked with were the same batch 4BM 4V.

[Lord Nightmare]

Can someone check if the 'reading controllers while dmc is playing causes spurious controller clocks sometimes' bug still exists on the 2a03 rev H? I know nintendo fixed it on the pal 2a07, and the 2a03 rev G still has the bug, but I don't know if rev H does.

LN

[2600]

Didn't expect this post to get brought back from the dead.

Someone had mentioned to me the trouble they were having with getting multiple rev H chips to work on a Vs. System. I don't know of a reason it wouldn't and don't have my test rig set up at the moment. None of those minor revision things should cause an issue with Vs. games. I should just set up my test rig and try it out.

[BMF54123]

It's been rumored and hinted at over the years that some ancient CPU revision didn't support looped noise and/or raw PCM, though I'm not sure that was ever verified. Maybe someone needs to find and dissect one of the very first Famicoms (the ones with square A/B buttons and lockup problems) to put this to rest once and for all.

The PPU, on the other hand, does seem to have had at least one noticeable change: I have either a first- or second-gen NES that tends to boot to a purple or brown screen, rather than the usual gray. The picture is also somewhat darker and clearer than newer models, more in line with my AV Famicom (which also produced a purple or brown screen on a few occasions, IIRC), although I'm not sure whether this can be attributed to internal PPU changes or motherboard circuitry.

[Jeroen]

I think Kevtris DID verify it.

[Lord Nightmare]

YES! The "original" RP2a03 revision used on all VS. NES arcade boards (and other non-nes arcade boards by nintendo, i.e. punch out/super punch out, popeye and donkey kong 3) does NOT have the 'short period' noise implemented at all; setting the appropriate bit does absolutely nothing.
Interestingly this revision was ALWAYS used on arcade boards (with the possible exception of playchoice 10 boards), so perhaps the problems people are having with crashes and other bad behavior when repairing vs nes arcade pcbs using 'home' RP2a03H cpus are related; maybe the original part has some special transient protection or better hysteresis on its inputs, while the home version (rev E, F, G, H) does not.

LN

[2600]

YES! The "original" RP2a03 revision used on all VS. NES arcade boards (and other non-nes arcade boards by nintendo, i.e. punch out/super punch out, popeye and donkey kong 3) does NOT have the 'short period' noise implemented at all; setting the appropriate bit does absolutely nothing.
Interestingly this revision was ALWAYS used on arcade boards (with the possible exception of playchoice 10 boards), so perhaps the problems people are having with crashes and other bad behavior when repairing vs nes arcade pcbs using 'home' RP2a03H cpus are related; maybe the original part has some special transient protection or better hysteresis on its inputs, while the home version (rev E, F, G, H) does not.

LN

I imagine they used the "original" on the Arcade boards to get rid of stock since they had them recalled from the Famicom.

The issues one person was reporting on the Vs. board was it wouldn't boot at all which is somewhat common as the sockets are usually bad. I need to try it out and put the analyzer on there to see what is up. I'll check to see if they work on Vs, Punch-Out, and Playchoice boards, but it will still be a couple months. I'm hoping they aren't bootleg chips.

[AWal]

Can someone check if the 'reading controllers while dmc is playing causes spurious controller clocks sometimes' bug still exists on the 2a03 rev H? I know nintendo fixed it on the pal 2a07, and the 2a03 rev G still has the bug, but I don't know if rev H does.

Give me some code (NTSC if that even matters) an I can test this.
RP2A03H
4BM 4V

It's been rumored and hinted at over the years that some ancient CPU revision didn't support looped noise and/or raw PCM, though I'm not sure that was ever verified.

I forgot who, but someone who happened to own a NES with a socketed 2A03 (I presume for a CopyNES) had tested an rev. E chip, which did not have the missing variable, but received a batch of NES CPUs, of which a tested one that was of the very early revisionless series did not have the variable, causing the bug.

I have a revisionless CPU myself, but the sprites (most likely sprite DMA) is shot...If this can be worked around I can test this one too.
RP2A03
4B471

[tepples]

There was a test ROM in this topic.

[AWal]

NES that tends to boot to a purple or brown screen, rather than the usual gray

Cirtuitry inconsistency tends to play a part in this.

I have yet to have a NES this was not reproducable on, although you can theoretically receive any color of the palette.

[AWal]

I'll check to see if they work on Vs, Punch-Out, and Playchoice boards, but it will still be a couple months. I'm hoping they aren't bootleg chips.

If there's a bootleg of the 2A03 on arcade harware, I'd love to see it.
Don't see why that'd be an issue nowadays since a legit 2A03H can be had for $5 US.

...Anyways, I got Mega Man 2 running on my nes with the revisionless chip in question, and it seemed to play the sound identical to my existing CPUs, so the revision on this one must be too late, if that is the case. I posted the date code a few posts back.

I also tried Skate or Die 2 to check raw audio, and it seems fine, as well as DMC.

Now I have the test pack dmc_dma_during_read4 on my desktop. Which of theese do I need to burn to check for the glitch?

Ooh, 100th post; Only took me 6 years w00t!

[2600]

I'll check to see if they work on Vs, Punch-Out, and Playchoice boards, but it will still be a couple months. I'm hoping they aren't bootleg chips.

If there's a bootleg of the 2A03 on arcade harware, I'd love to see it.
Don't see why that'd be an issue nowadays since a legit 2A03H can be had for $5 US.

FYI, I didn't mean bootleg like a NOAC, but a bootleg in that that it looks like a real 2A03, but there is nothing in the package. We see that a lot with 6502s.

[AWal]

...but there is nothing in the package. We see that a lot with 6502s.

I guess that's barely legal, given the "as-is" situation of most arcade boards these days...I suppose I got lucky that my VSSK daughterboard actually had code on the PROMs provided...let alone a working PPUs (ceramic, and of course the heatsink was broken off).

On topic of the glitches...I had a fiasco with my UxROM board but it has been replaced and I should be able to batch all of blarggs tests in one go here soon enough.