Page 1 of 3

SNES oscilloscope readings?

Posted: Tue Apr 06, 2021 12:49 pm
by akokun
I am trying to diagnose why this snes sits on a black screen and doesn't boot. Got it off ebay, and the first obvious issue was no power. Replaced the fuse with the correct type and I get a nice 4.81v off of the output of the voltage regulator, light turns on, and so on, so I really doubt it's a power issue. I'm afraid that a lightning strike or something actually damaged the CPU though, because I get no readings off of any the address or data lines of the S-CPU. Not even a little. If I'm thinking about this correctly, even if some other chip is damaged, the S-CPU should still be incrementing its own addresses, right? Even if it's sitting in a loop waiting for the APU. But I don't see any kind of activity at all on any of the address or data pins. I've checked the clock lines and they come in loud and clear at around the usecond scale, so am I overlooking something obvious? Btw this is one of the first times I've ever used an oscilloscope so maybe I'm just overlooking something stupid.

Also yeah game cartridge is in and I have continuity tested a lot of the cart connector pins. That thing actually looks in really good shape considering.

Re: SNES oscilloscope readings?

Posted: Tue Apr 06, 2021 12:52 pm
by Quietust
akokun wrote:
Tue Apr 06, 2021 12:49 pm
I get no readings off of any the address or data lines of the S-CPU. Not even a little.
When you say "no readings", do you mean that they're outputting a constant value that never changes, or do you mean that they're not outputting any values at all (i.e. they're high impedance)?

If the former, have you checked the S-CPU's RESET input (pin 50, if I'm reading the schematic correctly)?

Re: SNES oscilloscope readings?

Posted: Tue Apr 06, 2021 12:58 pm
by lidnariq
Unlike the NES, the SNES is just held in reset until the CIC lets it work.

Check reset from the CIC to PPU2 ("/RESET"), from PPU2 to PPU1 ("/RESOUT0"), and from PPU2 to everywhere ("/RESOUT1": CPU, power good IC, WRAM, cart slot, APU, EXT port)

Re: SNES oscilloscope readings?

Posted: Tue Apr 06, 2021 1:03 pm
by akokun
Okay thanks guys I should have known to check that first. Also hey lidnariq, when doing some research I think I saw a thread of you helping someone with a similar problem for like 10 pages, lol... I think it was a broken stack register or something which sadly was unfixable. Be back at some point.

Re: SNES oscilloscope readings?

Posted: Tue Apr 06, 2021 2:56 pm
by akokun
My USB scope is acting up and crashing the software (I'm guessing it's because of signal loss in my USB extension cord. stops acting up when plugged directly) so I just reverted to multimeter for this one. Those 3 pins give 4.90V which I'm assuming is what we want. Also tried powering on without the cart in and all three were low. Sadly I'm not gonna know what's going on now without the scope until I can get a better usb extender.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 12:26 am
by calima
You probably need a powered usb hub, not a better extension cable. I had a similar issue, and hooking a powered hub at the end fixed everything, otherwise the voltage was too low for things to work.

Ie
comp -> 5m usb extension cable -> powered hub -> usb device.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 2:47 am
by akokun
calima wrote:
Wed Apr 07, 2021 12:26 am
You probably need a powered usb hub, not a better extension cable. I had a similar issue, and hooking a powered hub at the end fixed everything, otherwise the voltage was too low for things to work.

Ie
comp -> 5m usb extension cable -> powered hub -> usb device.
After a lot of reading on this stuff, yep, I now see how finnicky it could be to get a longer USB going, with the whole 380ns response time required and max 500mA per usb port. From what I read ports will be straight up shut down if they try to use more than 500mA, and this scope seems to use at least that much.

I actually have this somewhat cheap scope https://www.owon.com.hk/products_owon_v ... cilloscope but it came with a Y connector for the PC side, so it's 1 male end into the scope and 2 male ends into the PC. I'm guessing it needs the extra power from the second PC connection.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 4:02 am
by creaothceann
USB 3 can deliver more, but then you need the entire chain to be USB 3 - plug, cable and device.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 7:19 am
by akokun
I've simply moved the SNES into a place where the scope can reach it now. I noticed I'm getting some actual readings off the cpu data pins after power on

CPU D0 60 HIGH
CPU D1 61 LOW
CPU D2 62 HIGH
CPU D3 63 LOW
CPU D4 64 HIGH
CPU D5 65 LOW
CPU D6 66 HIGH
CPU D7 67 LOW
(AA with game Cybernator in)

CPU D0 60 HIGH
CPU D1 61 LOW
CPU D2 62 HIGH
CPU D3 63 LOW
CPU D4 64 HIGH
CPU D5 65 LOW
CPU D6 66 HIGH
CPU D7 67 LOW
(AA with game Supernova in)

CPU D0 60 HIGH
CPU D1 61 LOW
CPU D2 62 HIGH
CPU D3 63 LOW
CPU D4 64 HIGH
CPU D5 65 LOW
CPU D6 66 HIGH
CPU D7 67 LOW
(AA with game Zelda LTTP in)

I'm not sure if that's significant in any way

Other stuff I checked: the scope reports that the crystal is going at 21.47427 MHz and SYSCK pin is pulsing at 2.606 MHz. Close enough I guess?

I've also ordered a super everdrive, I think those can boot up without communicating with the APU so that would isolate any APU issue.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 10:24 am
by akokun
Actually my super famicom arrived just now. Also currently blackscreening but this one seems to have a much easier to diagnose cart connector issue. Poking around inside of it, I see the expected activity of lots of digital stuff on pins such as address 0 / 1 / 2 off the CPU after boot which I wasn't getting off the snes.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 11:19 am
by lidnariq
akokun wrote:
Wed Apr 07, 2021 7:19 am
(AA with [any] game[] in)

I'm not sure if that's significant in any way
As you can tell, the data bus isn't particularly informative. The address bus would be more helpful.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 12:34 pm
by akokun
Indeed, but no matter what I do all 24 address lines seem to always be 0. I just thought it was odd how I was getting something on the data pins for whatever reason. And that's seemingly the only activity I get off of this thing other than clock waves and such. And alas, I have temporarily boxed up that snes for now. I figure I have a better chance of fixing this super famicom. At least it sometimes shows a garbled green screen and activity on the address pins. Unfortunately it's one of those with the soldered on cart connector and I have no idea how to actually get it off yet, but it looks rather corroded.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 12:36 pm
by lidnariq
Sure sounds like the S-CPU is toast there. You might find this: http://www.projectvb.com/nss/logs.htm informative

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 1:53 pm
by akokun
lidnariq wrote:
Wed Apr 07, 2021 12:36 pm
Sure sounds like the S-CPU is toast there. You might find this: http://www.projectvb.com/nss/logs.htm informative
That was an interesting read, and it reminded me... I should put the burn-in cart on my everdrive. The first snes is still totally dead but at least this super famicom can seemingly boot the everdrive.

I guess it's a bit depressing that one of the hardest to replace parts is the most likely to break. I wonder where he got all of his replacement CPUs. It seems like even if you build up a supply of junk SNESes, most of them are going to be junk because of bad CPUs.

Re: SNES oscilloscope readings?

Posted: Wed Apr 07, 2021 1:55 pm
by lidnariq
In DogP's case, he was repairing the rarer Nintendo Super System arcade cabinets, so just buying SNESes for parts was a reasonable choice.