SNES vertical bar: Quick and easy fix.

This is not the case here though. The SHVC-CPU-01 uses the same encoder with same pinout as the console on which i first made this fix. And the user who tried it on a 1chip took into consideration the different pinout and wired the cap properly.

1chips apparently do not suffer from the bar as much. I cannot see it at all on my 1chip-03. If you cannot see the bar on your console, do not try to look for it. Because it is cursed. Once you see it you cannot unsee it


While my solution was not engineered through careful considerations and was mostly a chance find, It was not a completely clueless move of the "hey this looks like it works and nothing has blown up yet" kind either. One of the reason I shared this is because I found that out by accident while trying to fix something else and wanted this to be known and from there more people can work on it to further refine the fix and make it work hopefully 100% on all board revisions.

A small ceramic cap will not filter the same type of noise and in the same amount as a large one. Also, because a small ceramic holds very little charge, they are not as good to regulate low voltage spikes.
The 220 uf value was assessed by trial and error to find the sweet spot, Retrojunkie made a video of this. But I know by experience that large-ish caps are often used on power rails. I used this to fix audio noise issue on my consolized game gear and I had to use substantial amount of capacitance for this to work.

Possibly that using a Cap in the empty spot before the regulator would work as well, And is better for smooth startup of console because it is before the regulator, but also makes working in the SNES more hazardous because of the dangling switch ( Read the last thread I just posted: http://www.assemblergames.com/forum...e-the-same-mistakes-I-did&p=708718#post708718 ) ... With this cap in the console, the switch becomes a Suicide trap and It might be just why they left the spot unused.

I may yet try a cap before the regulator anyway and see if it helps with the bar. I think this would depend on what the source of the noise is. But looking at how the early SHVC-CPU-01 has this cap and has little in the way of bar already indicate this might be a substitute.

IT seems TmEE over at Sega 16 had already made a more "engineered" version of this fix using various small value caps at key locations, but he did not document it much or took pictures of the result. He casually posted this in the middle of a big thread and it went completely under the radar. After I had fixed the bar on my console I could not believe no one had not done just that before so I looked around the internet and this is how I found TmEE's posts. What I did is a bit more brute force but also very simple, this is why I called it "Quick and Easy fix".

 

Quick and easy is the type of fix I like. There are those who always need to go that extra mile though. Same applies for the Crystal Clear Audio Mod for the Mega Drive, I can't tell the difference so I don't worry about it. But the SNES with that band... Two caps and it is fixed? Perfect!

 

No I know that, but I was just mentioning why some people would say that it wasn't working for them. Was trying to give them a head's up that 5V and ground are different pins on the 1CHIP's encoder.

I am assuming the same sized capacitor should be used in the 1CHIP? We've determined the best one is around 220uF 16V? But some people are worried for whatever reason and want to use a smaller one?

 

No, right now we are at 220 uf, but slightly lower might also do. In this case, the smaller the better.

 

I have a 1CHIP-03 and it doesn't have the bar or ghosting. It is a very beautiful, crystal-clear picture, especially through my XRGB-mini.

 

Do you have pocky and rocky? I was wondering if it crashed on your 1chip-03. It does on mine.

 

What is your serial number, It would make for a good reference point for finding a 1-chip-03 for myself.

 

Try an official Nintendo power supply. I bet you are using a 3rd party one, am I right?

 

I am already using official PSU.
If this was a power issue, I would be expecting to see it on the old SNES rather than on a 1chip, which is likely not using as much power.

 

Believe it or not, there's a lot of people with a 1CHIP SNES (or the SNES Mini) that report graphical glitches in Super Ghouls N Ghosts and Pocky and Rocky 2 and when they switch PSUs, it goes away.

 

I can believe that actually.

I am designing a PCB for SuperCIC/IGR and 1chip addon. I had loads of problems with the power supply on the phat 1chip (mini was OK). Had to add extra decoupling cap to my PCB to get the power supply stable enough for it to work properly.

 

fair enough, but all I have are official SNES bricks. Next time I take the 1chip out I will try testing with different adapters. Unfortunately, because the US SNES has that pretty unique connector, I cannot try anything else than official bricks.

 

Can't you mod your SNES to accept more common PSUs?

 

I could, that's what I do when I have a snes with a broken DC input. But My 1chip-03 is quite minty and see no reason to butcher it, especially when you consider its possibly the rarest SNES version out there. I could solder a temporary input to the board for testing but im not that motivated about the Pocky and rocky issue right now.

 

Or use an adapter. There's a laptop jack that's the same size

 

Hmmm, actually, I think I might have a defective SNES brick. I recall checking the connections and the continuity was fine but I still did not get voltage out of that particular brick. If I cannot get it fixed I could convert it to an adapter for standard PSUs on SNES.

 

As for the value of the caps, MaxWar is right. If the C is too low, the high frequencies will be shorted right to ground, but the "slow" (compared to RC time) wobbles won't get through the cap. Though, if C is too big, signal real close to DC will be shunt to ground, maybe dividing the power too much.

In the end it all depends on the R you're shorting to ground using the cap. The frequencies you'll cut off are the ones over 1/(2π RC), you should figure out those "bad" frequencies (oscilloscope/spectrum analyser) and choose C accordingly if you want the optimal C. Though if it's not broken, why fix it?

Cheers!

FG

 

would it be theoretically possible to mod a snes to use 1 volt?

 

Yes... although not that useful...

You'd require 1V with many amperes and a step-up converter...

 

My idea was that since the added Capacitance is after the power switch, the caps get drained when the console is turned off. By opposition, when the console is turned on, the caps will need to be charged, which will result in a slower than usual transition from 0 to 5v on the power rail. If this transition is slow enough it will result in the console glitching. You then need to press reset for it to work normally. I have observed this when using too large caps. Fortunately, with the two 220 uf caps used here, it would appear that the console is able to boot normally.