SNES oscilloscope readings?

Discussion of hardware and software development for Super NES and Super Famicom. See the SNESdev wiki for more information.

Moderator: Moderators

Forum rules
  • For making cartridges of your Super NES games, see Reproduction.
93143
Posts: 1718
Joined: Fri Jul 04, 2014 9:31 pm

Re: SNES oscilloscope readings?

Post by 93143 »

Is the S-CPU well enough understood that a new-production chip or a CPLD version in a pin-compatible package could be sold standalone as a replacement part?

I don't have a solid idea of the potential market size, but I imagine it would be small...
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: SNES oscilloscope readings?

Post by lidnariq »

93143 wrote: Thu Apr 08, 2021 1:37 am Is the S-CPU well enough understood
Probably?
a new-production chip
Unaffordable. You can't get custom silicon without an initial outlay on the order of $80k+
in a pin-compatible package
I've seen clever things with castellated PCBs and mezzanine connectors, such as the UltraPIF
or fancy flex PCBs, like the UltraHDMI

but I don't know how the pin pitch of the S-CPU compares to what castellation pitch is affordable.

One of the bigger annoyances is that the S-PPUs and S-WRAM are assuredly also 5VCMOS logic, so even though the address bus is unidirectionally out of the S-CPU, up-translation is probably needed on every pin, making such a product significantly bigger
creaothceann
Posts: 611
Joined: Mon Jan 23, 2006 7:47 am
Location: Germany
Contact:

Re: SNES oscilloscope readings?

Post by creaothceann »

It would be easier to transplant a CPU from another SNES...
My current setup:
Super Famicom ("2/1/3" SNS-CPU-GPM-02) → SCART → OSSC → StarTech USB3HDCAP → AmaRecTV 3.10
93143
Posts: 1718
Joined: Fri Jul 04, 2014 9:31 pm

Re: SNES oscilloscope readings?

Post by 93143 »

creaothceann wrote: Thu Apr 08, 2021 3:29 am It would be easier to transplant a CPU from another SNES...
Why bother? If you've got another SNES that works, use that one.

If it's usually the CPU that fails prematurely, it's going to be hard to find a busted SNES with a good CPU, so the only way to fix a busted one is to ruin a good one. That's why I suggested a new replacement part - it could stave off a premature decline in the availability of working consoles.

But it's not clear what the statistics are on this. I have a SNES with a bad CPU, and I've heard of others, but I don't know of any attempt to robustly quantify this problem. I may be overreacting. Also, it's possible that only a particular model of CPU is prone to this sort of failure (I seem to recall hearing something about that), which would arrest the accelerated decline once the vulnerable consoles had all failed.
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: SNES oscilloscope readings?

Post by Pokun »

In my experience with Famicom is that people tend to always think it's the CPU that is broken when it's more likely just a dirty cartridge connector or bad AC-adapter. But at least here we have some evidence that the CPU breaks more often than other custom parts. In that picture of broken CPUs it seems 02 is common, but also some 01 chips. Is that the revision number though? It also appears to say A or B on some chips.

Unfortunately only the S-PPU (5C77-01) and various cartridge co-processors are decapped so far, while the S-CPU, S-APU and S-WRAM are not yet, AFAIK.
creaothceann
Posts: 611
Joined: Mon Jan 23, 2006 7:47 am
Location: Germany
Contact:

Re: SNES oscilloscope readings?

Post by creaothceann »

Pokun wrote: Thu Apr 08, 2021 7:22 am Unfortunately only the S-PPU (5C77-01) and various cartridge co-processors are decapped so far, while the S-CPU, S-APU and S-WRAM are not yet, AFAIK.
65c816: https://commons.wikimedia.org/wiki/File ... _metal.jpg
5A22: https://siliconpr0n.org/map/nintendo/s- ... y=8872&z=1

Very Interesting, but still not really helpful until we have a representation of all layers.
My current setup:
Super Famicom ("2/1/3" SNS-CPU-GPM-02) → SCART → OSSC → StarTech USB3HDCAP → AmaRecTV 3.10
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: SNES oscilloscope readings?

Post by Pokun »

Oh they got them.

You mean not useful before a visual 5A22 is made?
tepples
Posts: 22708
Joined: Sun Sep 19, 2004 11:12 pm
Location: NE Indiana, USA (NTSC)
Contact:

Re: SNES oscilloscope readings?

Post by tepples »

It might mean that making Visual 5A22 won't be feasible until the layers below metal are exposed and photographed.
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: SNES oscilloscope readings?

Post by Pokun »

So there are more layers not yet photographed.
creaothceann
Posts: 611
Joined: Mon Jan 23, 2006 7:47 am
Location: Germany
Contact:

Re: SNES oscilloscope readings?

Post by creaothceann »

My current setup:
Super Famicom ("2/1/3" SNS-CPU-GPM-02) → SCART → OSSC → StarTech USB3HDCAP → AmaRecTV 3.10
Pokun
Posts: 2681
Joined: Tue May 28, 2013 5:49 am
Location: Hokkaido, Japan

Re: SNES oscilloscope readings?

Post by Pokun »

I see.
akokun
Posts: 16
Joined: Tue Apr 06, 2021 12:40 pm

Re: SNES oscilloscope readings?

Post by akokun »

Oh hey I didn't realize this got a page 2. I don't know much about it, but I don't suppose FPGA will ever help us with 5a22 replacements since they're all on 3.3v or some such?

Also, I suppose I should look into getting a logic analyzer. I feel I could figure lots of things out if I knew where my games were crashing by cross referencing a disassembly or something. I see some extremely cheap (and probably awful, but maybe okay to futz around on) ones already, $12 delivered... geez. Anyone have opinions if it's worth getting a cheap 8 channel one? As I understand it, it's mostly the software that carries the value with these things. Although if I'm not mistaken, 8 channel would only allow me to capture one byte off the address pins at a time.
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: SNES oscilloscope readings?

Post by lidnariq »

You could get one of the very cheap 8/16-channel ones. Search AliExpress for "fx2lp". You'll need all the probes, too, so it may not work out to a win to buy things piecemeal, though.

It's limited - only 12MSa/sec at 16 channel, but it's what I used for the PPU logic analyzer traces.
creaothceann
Posts: 611
Joined: Mon Jan 23, 2006 7:47 am
Location: Germany
Contact:

Re: SNES oscilloscope readings?

Post by creaothceann »

As lidnariq said, slower signals like the CPU address/data buses, can be sampled with cheaper logic analyzers. Higher-frequency signals would be more difficult.

More theory:
https://www.youtube.com/watch?v=dobU-b0_L1I
https://www.youtube.com/watch?v=xZ5wKYnCNcs
My current setup:
Super Famicom ("2/1/3" SNS-CPU-GPM-02) → SCART → OSSC → StarTech USB3HDCAP → AmaRecTV 3.10
akokun
Posts: 16
Joined: Tue Apr 06, 2021 12:40 pm

Re: SNES oscilloscope readings?

Post by akokun »

Thanks, those were great reads / watches. I've always wondered how exactly the PPU1/2 worked on a physical level as opposed to in the software realm where you just think of them as some registers that you upload stuff to at certain times.
Post Reply