More SNES mainboard repair!

As with RetroJunkie, I'm having trouble with messed up graphics on a SNES. My problem is a bit more colorful though:


I suspect it's the connector, as it looks like the garbled mess you see on a NES with a bad connector. Audio works fine though! As with RetroJunkie's board, mine is also the type with the connector soldered directly to the board.

Here's a photo of the mainboard:

(Click for full-sized photo!)

I've looked around the board, and I didn't notice any corroded traces or anything like that. I've also cleaned the crap out of the connector and the game used to test it. Any ideas guys?

 

That doesn't look like John Madden Football... I thought we'd all agreed that John Madden Football would be the test cartridge of choice! :biggrin-new:
That could be a bad sign if there's no signs of corrosion. I guess PPUs can die on their own due to natural causes...
Any chance of a picture in the daylight looking down at the board?
I can't even read the model number off that board due to the angle and lighting! :friendly_wink:

 

This more like it?


(click for big!)

 

Oh yeah, that's much better! Why is it so clean, there's no corrosion anywhere! :dejection:
Is that gunk or camera blur on the left hand side of PPU1? Try using a dry toothbrush and scrubbing away at any gunk there.
If still nothing, I have no idea short of swapping chips around...

Don't give up on it though by all means! I may have a spare board that looks to be that age.
What was the serial out of curiosity? I know my 1CHIPs start with 17X and my one like yours to be 159.
Here's what my one looks like for comparison

And I'd be more than happy to send a PPU your way if need be, unless we can work out what else it might be.

If I had to guess though and there's no damage externally, one of the PPUs might be having issues.
Surely someone more versed in Super Nintendo can lend a hand :smile-new: After all I am a Sega kinda guy!

(Although I'm really liking the chances that an LM324 could be a suitable replacement for S-MIX)

 

Looks like an open address line to VRAM (U4/5).

 

Just dust by the looks, here it is after a scrub:

If we work out that a bad PPU is the problem, that would be awesome! I'd probably need to invest in a hot air station to do this though.

Er, and how do we un-open the address line to VRAM? :\

 

use a multimeter to trace all connections to/from the vram and make sure they are connected

 

Okay, I've checked continuity between U4/U5 and U2/U3 and most of the legs seem to be connected. Some are not and these ones just lead to a via, is that normal? There were also a few legs that didn't have any traces running from them, yet were still connected to U2/U3. I did find one of those "no trace" legs that didn't have connectivity to either U2 or U3, might this indicate something amiss?

 

Vias are making connection from 1 layer of the board to another.

You need to trace the vias too, connections dont just goto a via and end there.

 

Ah that makes sense. But have you seen the reverse side of a SNES? There's so much lacquer, I can't make out any traces! Best I can do is angle it so as to catch the light and looks for any ridges. But this only makes it look like it's made up of a bunch of giant slabs of copper, any tips? :\

 

Bad PPU, VRAM, whatever it is - chances are I'll have a spare off the dud I bought. Unless I repair that board...
Then I'd be happy to trade for that board (I still need a reset switch!) plus the case is the worst yellow of yellows.
And the text on the front of the console looks to have been half cleaned right off! At least the case can be a candidate for a spray paint.

And from what I'd been tinkering with, I reckon flux, desolder braid and a steady hand would be best if we went down that path.
Of course, if we can't determine which component is at fault... there's a lot of pins on a lot of chips! :wink-new:

 

Have you tried Retr0bright? Though if it's heavily yellowed, the plastic might have become brittle, which could be trouble down the road.

Supposedly, a hot air gun is the best way to solder/remove this style of chips. Hopefully it won't come to that though!

 

Not even going to try Retr0bright on this one! Actually looking at it... is that just grime? How the hell?



I guess if you'd prefer we can heat gun chips off. Did we work out which chip was toast yet? :biggrin-new:

(Also happy 1000th post!)

 

Scrub the goop off with some zippo lighter fluid and Retrobright it, it will look as good as new when your done.

As for the card slot, remove the connector, throw it into a bowl the methylated spirits for hour , then grab a credit card wrap a thin piece of cloth around it (no more than one layer of cloth), Remove the Pin header from the spirits, and insert the creditcard+cloth into the pins as streight as humanly possible and remove it after a second or so, repeat from left to right, till dry, then leave in a well ventilated area overnight, you can allso use a q-tip to clean the bottom of the header if its dirty. I do this on all nintendo consoles wher the connector is removable.

 

It looks like it was found at the bottom of a pond. I'd try some orange-based "goo remover" see if that gets rid of it.

Also, thanks for the 1000 post congrats, it's all downhill from here!

 

Bad chips are possible, but unlikely. Far more likely is a open connection.

Checking the VRAM is easy. For each address line, and only address lines, make sure both VRAM chips have continuity between them, and then swipe the PPU pins for continuity. Every address line should be accounted for.

 

Address lines being the legs you've underlined I take it? Are you saying that each VRAM is linked? This whole time I've been testing from VRAM to PPU independently :\ I'll go give that a try now! Also, "swiping" is a legitimate form of continuity testing the PPU? I thought I was being clever by running the probe along the legs of the PPU and listening for a beep

 

Address lines being the ones that start with A

Ones with the lines ABOVE just mean they are active low.

Without looking, I would say the VRAMS are on the same bus (so yes, address and data lines linked). Each chip is selected as the current one in use via the CS line (chip select).

 

Damn, I tested the underlined ones! Will have to try the ones starting with A tomorrow! As to the two VRAMs being linked, I found continuity from VCC.

If I'm reading Calpis' post correctly, I should find that all the address lines have continuity between both the two VRAMs and both PPUs?

 

If they are on the same bus, which from calpis's comment, I assume they are - yes.

And none of them are underlined. The line is ABOVE, not below (i.e. /CS, /WE and /OE)

Edit:

Further reading if you wish to understand a little more:

http://en.wikipedia.org/wiki/Chip_select

Not exactly the same (as thats talking about serial SPI chips), but the theory is similar with how the chips are selected and that the data lines are all linked.