SuperDisc BIOS released + analysis [attn: nocash]

Discussion of hardware and software development for Super NES and Super Famicom.

Moderator: Moderators

Forum rules
  • For making cartridges of your Super NES games, see Reproduction.
nocash
Posts: 1201
Joined: Fri Feb 24, 2012 12:09 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by nocash » Sun Mar 06, 2016 7:40 pm

Are there any photos showing the complete chipset?
Front and back sides of all PCBs, including whatever daughterboards.

byuu
Posts: 1545
Joined: Mon Mar 27, 2006 5:23 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by byuu » Sun Mar 06, 2016 7:52 pm

I'm afraid this is all we were given by the owners:

http://luigiblood.tumblr.com/post/13273 ... laystation

They're talking about raising $30,000 to put out a book (no idea how you can make an entire book out of "my dad bought a lot at a corporate auction and we found an SNES-CD prototype inside!"), that would supposedly contain high-resolution scans of the PCBs.

But even those low-res pictures revealed something new: the real Sony part IDs of the SMP and DSP chips.

User avatar
LuigiBlood
Posts: 62
Joined: Thu Jul 29, 2010 2:24 pm

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by LuigiBlood » Sun Mar 06, 2016 7:58 pm

No there's even more, I've just put out a quick album: http://imgur.com/a/2PGcU

I believe I can get more from the owner if you want.

byuu
Posts: 1545
Joined: Mon Mar 27, 2006 5:23 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by byuu » Sun Mar 06, 2016 9:55 pm

Oh, beautiful! :D
(Even better if they weren't watermarked, but oh well.)

Looks like the CD audio DSPs (CXD2500AQ) are readily available. Wish we could get our hands on some CXD18000's or at least their datasheet, that would help immensely with research. But it seems that's a phantom piece. It really doesn't seem to share a lot with the PS1 CD controller.

By the way, what's this chip? http://i.imgur.com/ehNHPeY.jpg
EDIT: much better photo: http://www.blogcdn.com/slideshows/image ... 1-20-1.jpg
Seems to be a NEC D75P308GF, a 4-bit microprocessor. Probably used to drive the CD-ROM front panel. It has an internal 8064-byte program ROM (OTP). While it does have a documented read-out mode, I imagine the odds of us obtaining that ROM are slimmer than a snowball's chance in hell :(

In any case, it's quite likely the BIOS communicates with it in some way. Maybe the CD-PLAYER I/F tests this path? So we'll likely have to resort to HLE for this component.

User avatar
LuigiBlood
Posts: 62
Joined: Thu Jul 29, 2010 2:24 pm

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by LuigiBlood » Mon Mar 07, 2016 6:21 am

byuu wrote:Seems to be a NEC D75P308GF, a 4-bit microprocessor. Probably used to drive the CD-ROM front panel. It has an internal 8064-byte program ROM (OTP). While it does have a documented read-out mode, I imagine the odds of us obtaining that ROM are slimmer than a snowball's chance in hell :(

In any case, it's quite likely the BIOS communicates with it in some way. Maybe the CD-PLAYER I/F tests this path? So we'll likely have to resort to HLE for this component.
Well it's even worse if the CD-ROM Drive doesn't even work in our case here.

byuu
Posts: 1545
Joined: Mon Mar 27, 2006 5:23 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by byuu » Mon Mar 07, 2016 7:21 am

> Well it's even worse if the CD-ROM Drive doesn't even work in our case here.

Yeah, there was the very slim possibility we could produce CDs that would boot on the real hardware. But there was never any real possibility that we'd get to reverse engineer the hardware through software testing.

Still, if it at least worked a little, it would have been nice to capture some video of how the front LCD panel operated with a music disc inserted.

Fate really is cruel. If only we had maybe 5-10 of these systems, and a single game for it, we could easily emulate it in great detail*. But with only the BIOS ROM and nothing else, we're only ever going to be able to emulate a very small fraction of the BIOS, and most of it will likely be wrong/incomplete (yet functional.) At this point, I'm confident we'll get to a point of running ISOs/CDs. But most of those CXD18000 commands are going to remain pure guesswork or unimplemented. Not that I even think they'd all be that useful. What more can you really do with a CD other than read its data and/or play its music tracks? It's kind of ridiculous how many registers are exposed to this thing in the first place.

(* it really would be very underwhelming. With no processing power inside the BIOS cartridge, this system as it is right now would just be a stock SNES with constant, crazy-long loading times and possibly some CD audio.)

Anyway, the one single console thing is really the worst part. I'm reasonably confident the OTP data inside that NEC microcontroller is still there, but it's going to decay like all EEPROM data does. And probably relatively soon. But given the console's immense value, there's no chance they'd consider desoldering the chip, and then wiring it up to dump the program ROM contained within it. So instead that data's just going to bit-rot, and then truly be gone forever. Even though the CD system doesn't work at all anyway.

Sik
Posts: 1589
Joined: Thu Aug 12, 2010 3:43 am

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by Sik » Mon Mar 07, 2016 9:30 am

Maybe it's time to start asking them what do they plan to do with it?

Also, does anybody know why the CD drive doesn't even work in the first place? Is it just a mechanical failure as usual? Because if not, then it could be a problem with that memory as you said... though honestly I have my doubts about it. Especially since from what I see it's OTP, shouldn't that be written by burning fuses instead of storing a charge? (there are two models of that controller, the other one is indeed rewriteable)

User avatar
LuigiBlood
Posts: 62
Joined: Thu Jul 29, 2010 2:24 pm

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by LuigiBlood » Mon Mar 07, 2016 9:45 am

Sik wrote:Maybe it's time to start asking them what do they plan to do with it?

Also, does anybody know why the CD drive doesn't even work in the first place? Is it just a mechanical failure as usual? Because if not, then it could be a problem with that memory as you said... though honestly I have my doubts about it. Especially since from what I see it's OTP, shouldn't that be written by burning fuses instead of storing a charge? (there are two models of that controller, the other one is indeed rewriteable)
The kickstarter clearly says the owner wants to repair the hardware, possibly with help from people who worked on this before. I'm trying to talk to some ex-SNES CD devs about that and if anything new actually happened.

nocash
Posts: 1201
Joined: Fri Feb 24, 2012 12:09 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by nocash » Mon Mar 07, 2016 12:57 pm

LuigiBlood wrote:No there's even more, I've just put out a quick album: http://imgur.com/a/2PGcU
Thanks! But there's also this image, http://assemblergames.com/l/attachments ... jpg.15278/ which appears to be another PCB being found in the same console. No idea how many further PCBs might be hiding in there.
LuigiBlood wrote:I believe I can get more from the owner if you want.
Complete set of PCB images with front/back sides would be great!

A dump of the (E)PROM (the chip with the "CDU33A1.0d" sticker, seen on the above image) would be also important.
I've never seen such a "socketed SMD chip" before, but if it's having the same pinout than the chip on the BIOS cart, then one could probably just swap the chips, and dump it by dumping the BIOS cart (with a small mod on the VCC pin to match 28pin vs 32pin; before trying to do that one would better remove the stickers, and check that the chips are really having a compatible pinout).

Without dumps and complete photos of chipset it's a bit ridiculous to try to figure out how the thing should work, or even how it could be repaired ; - )

User avatar
LuigiBlood
Posts: 62
Joined: Thu Jul 29, 2010 2:24 pm

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by LuigiBlood » Mon Mar 07, 2016 1:11 pm

Guess we gotta get a Sony CDU33A (perhaps 31A?) PC CD-ROM drive to help on this. There are DOS drivers apparently, maybe that can help by REing those or something?

Anyway I asked him for the PCB pics, though I'm gonna have to go soon as I'm in the middle of moving house...
Maybe he'll get here and help you all out, I don't know, but I linked to him this topic as well.

Sik
Posts: 1589
Joined: Thu Aug 12, 2010 3:43 am

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by Sik » Mon Mar 07, 2016 2:33 pm

nocash wrote:Thanks! But there's also this image, http://assemblergames.com/l/attachments ... jpg.15278/ which appears to be another PCB being found in the same console. No idea how many further PCBs might be hiding in there.
There's so much off-the-shelf stuff there that now it makes me wonder how hard it is to make a CD drive from scratch.

Also what's the CXD1186BQ? Because "subcode decoder" doesn't tell me much, and this thing has DMA =P (whether that's actually used or not is a different matter) All registers are documented, no idea if in the SuperDisc case the 65816 has access to them.
http://www.digchip.com/datasheets/parts ... BQ-pdf.php

nocash
Posts: 1201
Joined: Fri Feb 24, 2012 12:09 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by nocash » Mon Mar 07, 2016 3:28 pm

Oh, that image with the "CDU33A1.0d" chip was posted only because it's having similar hardware, but isn't part of the console? Okay, then forget about that!

byuu
Posts: 1545
Joined: Mon Mar 27, 2006 5:23 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by byuu » Mon Mar 07, 2016 6:26 pm

> Without dumps and complete photos of chipset it's a bit ridiculous to try to figure out how the thing should work, or even how it could be repaired ;-)

The BIOS alone is enough to bullshit an emulation of CD-ROM program loading, at least. But yeah, the emulation is always going to be basically a giant hack. I imagine 90% of the SNES-exposed CD-ROM interface will likely end up unimplemented.

I'm not even sure how much of the low-level stuff makes sense to emulate. We don't really have to decode subchannel data, we can read it right out of an ISO file, or use the OS function calls to access the data from a real CD-ROM drive. I do want that NEC PROM, but we can probably get by with just simulating the six buttons instead.

Though I certainly agree. It'd be nice to get as much info and data as possible. I just don't see them being willing to perform operations that risk harming this system further, given its priceless value.

I really don't see them being able to repair the system, though. They already looked it over for simple things like caps, supposedly. If one of those ICs are bad, like the CXD18000, it's game over. Those parts aren't sourceable.

> Guess we gotta get a Sony CDU33A (perhaps 31A?) PC CD-ROM drive to help on this.

What we really, truly need more than anything is a source for CXD18000 chips. Even getting the prototype working won't allow us to perform the exhaustive hardware testing necessary to get strong emulation in place. We need our own chip, and we need to hook it up to an Arduino, and see what all these registers actually do.

I'm not even sure how immediately valuable the CXD18000 in isolation will be without programming documentation to make sense of the inputs, outputs, and their designated purposes.

nocash
Posts: 1201
Joined: Fri Feb 24, 2012 12:09 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by nocash » Mon Mar 07, 2016 6:33 pm

That chip should be CXD1800Q, not CXD18000.

byuu
Posts: 1545
Joined: Mon Mar 27, 2006 5:23 pm
Contact:

Re: SuperDisc BIOS released + analysis [attn: nocash]

Post by byuu » Mon Mar 07, 2016 8:06 pm

Ah, good eye, thanks. That puts it a lot closer to the PS1's CXD1815, then. But still, the register names seem to be entirely different for that one.

Post Reply