SNES revision different colorburst/vsync lengths? Oscilloscope readings don't match video output?

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.
Post Reply
akokun
Posts: 16
Joined: Tue Apr 06, 2021 12:40 pm

SNES revision different colorburst/vsync lengths? Oscilloscope readings don't match video output?

Post by akokun »

Maybe I simply don't understand well enough what's going on here, but the situation is I have several superfamicoms I bought to poke around in, and a couple in my "broken" pile have this problem with some kind of black smearing and poor brightness. Unable to figure out why, I eventually resorted to comparing 1:1 the readings from other working consoles with good video output. Strangely I could find no real difference in the oscilloscope readings, and at this point I'm resorting to getting scope readings right off the yellow RCA cable, and there's nothing between that step and my capture card / tv. The signals coming out of these two appear almost identical, so I really can't figure this one out. Surely the signals should look much different for these or is the different-looking colorburst and vsync I'm seeing somehow causing this?

For my test case I could think of no better than the everdrive menu. It's grey on white so there shouldn't even be any color issues involving this, doesn't communicate with the APU at all so no need to even have the SPC module in, it's a completely still image, and it loads instantly, so I don't even usually need anything plugged in except the power plug and the everdrive cart. And most of all it's super easy to figure out which scanline you're even looking at just by recalling the text on the screen.

Good output - Sfamicom GPM-02
Goodever.png

Blackstreaks & sometimes bad sync on tv - Sfamicom SHVC-CPU-01
badever.png

goodvsbadoutputs.png
1) Top - Blackstreaky Sfami: Scanline of the bottom tails of the "p" and "g" in "page" at the top of the screen. The only real difference is the anemic-looking colorburst.
2) Good Sfami: Scanline of the bottom tails of the "p" and "g" in "page" at the top of the screen. My snapping and aligning isn't the greatest here; when overlapping / measured, they are almost exactly identical except for the colorburst which this time is lengthier and has more amplitude. Freq seems to be the same.
3) Blackstreaky Sfami: The vsync pulses that occur much later in the frame. I'm not really sure if this is normal or why there are more pulses on this one or why my good, working units all match the other signal. The video encoder on this unit is different to the GPM-02, but I believe it's the same one as in my SNES if all SHVC CPU 01's have the same video encoder, and my SNES has wonderful video output.
4) Bottom - Good Sfami
(Ignore absolute voltage levels here; I adjusted all the saved signals and spread them out so it would be a bit easier to read. All the absolute and relative voltages are nearly same and within nominal values when not adjusted/spread out.)

So maybe the rising and falling edges are a bit cleaner on the GPM-02, but I still don't really get where the black streaks are coming from. I certainly don't see them in the scope reading, if anything there should be a bit of luma streakiness from the worse falling edges of the SHVC CPU, or maybe it is the color logic catching that slower falling edge as a bad color subcarrier, but then why is it black instead of some color? Both my capture card (GV-USB) and TV (not CRT, crappy HDTV which often gets sync problems from this system) both exhibit the black streaks.
none
Posts: 117
Joined: Thu Sep 03, 2020 1:09 am

Re: SNES revision different colorburst/vsync lengths? Oscilloscope readings don't match video output?

Post by none »

I have seen this happening with a faulty yellow cable. Sounds obvious, but have you tried a different cable?
akokun
Posts: 16
Joined: Tue Apr 06, 2021 12:40 pm

Re: SNES revision different colorburst/vsync lengths? Oscilloscope readings don't match video output?

Post by akokun »

Yeah, I think the cable must be eliminated as a culprit because the other systems display much better picture with the same exact cable. All I'm really doing between readings is plugging the power adapter and multi out and everdrive cartridge into other test systems, nothing else changes, not even the probe on the other side of the cable which is just sitting there in the same configuration. Unfortunately I don't have any other cables on hand to test either.
akokun
Posts: 16
Joined: Tue Apr 06, 2021 12:40 pm

Re: SNES revision different colorburst/vsync lengths? Oscilloscope readings don't match video output?

Post by akokun »

I probably should have tried this a long time ago, but... it seems it's actually just capacitor C59 which sits at the end of the luma output circuit and is just before the multi out. I guess this 30 years old by now capacitor probably drifted out of spec, but I'm still not sure why the signals look so similar to very good output. Oh well. Case closed.
akokun
Posts: 16
Joined: Tue Apr 06, 2021 12:40 pm

Re: SNES revision different colorburst/vsync lengths? Oscilloscope readings don't match video output?

Post by akokun »

Funnily enough there is now a mystery dot / interference thing traveling down the screen but I think this is a side effect of a not quite correct capacitor replacement. I'll see if it goes away when I get an actual 220uf capacitor, or perhaps I need to replace even more caps. Output still doesn't look great but at least it's not horribly dark and streaky.

https://www.twitch.tv/videos/1053586751

Protip: actually that interference was being caused by some local extreme EMI put out by my... soldering station?
Post Reply