shmups.system11.org

I recently found a PAL 1CHIP-01 and as I had heard could be the case, there was a nasty ghosting issue most noticeable to the right of objects against solid colours. Luckily I knew the fix; replace C11 with a 470nF capacitor. So I grabbed a sheet and swapped it over, to find absolutely no difference?!

I wondered if this fix might not apply to PAL consoles and if there's a different solution, does anyone know? You can see the bottom of this console (in case it helps identify what I have here) with the replaced C11:

I had the same issue as you, after doing this fix on 3 1-chip snes I realized that fix doesn't work the same on each unit for some reason, curious to know why, not specially related to PAL, I saw it on an SFC too.

Hmm interesting to know thanks. I might try some higher values; I remember hearing something around 300nF to 1uF is a suitable range and higher than that can give ringing issues. So if I try values of around 600, 800 and 1uF I wonder if the issue will improve or if I'll start to get other undesired effects. I'll order some and give this a go.

Just as I was about to buy some other values, I suddenly realised it's probably fine to stack the 470nF ones that I have, so I added a second, no obvious difference. I proceeded to add a 3rd, and 4th, and I see no improvement, but I don't see it adding any visual issues either. At this point I should be at 1.88uF +/- 10%. I guess I'm at a loss. If anyone has any other info or suggestions on this, I'm open to trying.

You might also want to consider the possibility that the artifacts you see are not related to the C11 capacitor. Make sure your cables are coaxial 75-ohm characteristic impedance. Bad cables can cause signal reflections due to an impedance mismatch, which can cause ghosting artifacts.

Also, if you're using a CRT this can be a problem with the monitor itself.

Hmm you might be right about bad cables. I have a sync-on-luma cable that I was using with my 2chip and while the 2chip has a smear to the right, I think I can see the same kind of ghosting to the right too, it's maybe disguised by the smear. I disconnected the audio in the cable at the console end in case that was causing the artifacts (I noticed this is true on my cheap PS2 cable when using progressive scan, and fixed by running the audio outside of the cable) but that made no difference here.

I'm not sure what you mean by "coaxial 75-ohm characteristic impedance", to a simpleton does that just mean better shielded? Can I buy my own cable and wire it up? I tried this once before with a PS2 cable, using a VGA cable that had shielded RGB lines but it didn't cure my audio disturbance (also causing ghosting to the right) so I just ran a separate audio cable to fix that issue.

I'm using an LCD with an OSSC, not CRT here.

75-ohm coax cables are designed to transmit video signals to their destination with low loss. There aren't many cable makers in the retro market that offer it though. Last I checked, it was just Retro Access for RGBS and HD Retrovision for YPbPr (someone please inform us if more have come into the market since then). It's possible to wire one up yourself if you are capable enough.

I'm still not sure if C11 is the same between NTSC and PAL consoles. If someone can confirm if they're the same or not, that would be great.

I found some of my old SCART cables are coaxial and actually say 75ohm on them (each cable is "wire, plastic, wire shield, plastic" and the audio cables run inside their own plastic cable with foil around it) so as a test I made up a new cable for my PS2 and I was pleasantly surprised to see the ghosting there completely resolved! So I no longer have to run the audio via a separate cable.

While I had high hopes this would resolve my SNES issue, I know it's not the audio causing the issue there, but maybe there's interference from luma sync, 5v or cross-talk between RGB? With this DIY coaxial cable, I have the shield of sync, R, G and B cables grounded to SNES port 6 and SCART ground pins (5, 9, 13, 18, 21) but the issue has not improved at all.

I wonder if doing an RGB bypass would help at all. If this is down to the C11 cap or similar, I guess not?

I read that there was a ghosting fix involving lifting pin 155 and putting a 20ohm resistor to 5v, so I tried that but still no better. I did notice that shielded cable I made makes a big difference with the audio buzz. I grounded every line first just at the SCART end, and then at both ends (SNES and SCART) (I'm not sure which is technically accurate) but I see ghosting with both.

Before you think it's the OSSC or my TV, I get a perfect picture from my RGB modded NES with a cheap csync cable. Bypassing the OSSC the SNES still has ghosting. I removed R, G and B separately to see if it's a particular channel but all lines had ghosting.

So ultimately I'm wondering if the PAL fat SNES needs a different ghosting fix.

On my PAL 1CHIP replacing C11 did fix the ghosting problem, so it's weird that it doesn't help on yours.

As for RGB bypass, I've tried it on a few SNES and Mega Drive consoles, and I've never seen any improvement whatsoever compared to the original circuits. On my 1CHIP I specifically tried bypassing to get rid of the ghosting but it made no difference so I removed it again. Some people seem to swear by these bypass mods so maybe I was just unlucky that they didn't work with my specific models of consoles.

Using the C11 replacement, adding capacitors to fix the huge jailbar and doing the brightness fix my 1CHIP SNES has a pretty perfect RGB image. Certainly the very best I've ever seen from a pre HDMI Nintendo console, but that's not really saying much

Interesting, that's nice to know it can work on PAL, and that as I suspected it's unlikely to help bypassing RGB. Unless the C11 capacitor I received is the wrong value, but I did order from another supplier and that one didn't help either. Strangely while they should both be 470nF, one was much taller than the other. It's a shame, these surface mount capacitors seem difficult to test and they are unmarked, but I doubt I was sent the wrong value from two different suppliers. I wonder, for testing purposes, if I could install a normal capacitor that is marked so I can be absolutely sure I have the right value, but I imagine the polarisation is then important, and I'd be unsure which way round it should be installed.

I also tried stacking one capacitor 4 high, and saw no difference to picture quality. I was expecting to get ringing or other artifacts, but nothing. I guess it would be nice to see some effect from changing the capacitor there to know that it is doing something.

I believe that a side effect of installing that C11 capacitor is that some games like Street Fighter II will cut off the first few lines of graphics, the higher value capacitor you install the more will be lost IIRC. Maybe that's worth trying to see if your capacitors have any effect on your console.

Ah nice one thanks, yes in a Street Fighter II match, I see about 3 slightly transparent lines at the top with this new capacitor, and maybe 1/4 of a line with the stock one (it starts on the left and fades out towards the middle of the screen) but for example at the start of Mega Man X I can see the life bar ghosted against solid colours with both caps; there's no noticeable difference. OK thanks so at least it is having some effect, but not to the ghosting.

I've also been playing some Super Mario Kart, and on the first track in time trials, if you pause the screen with "Yes/No" on the screen over the grey track, the ghosting is really bad on the right of "Yes". Also as Lakitu swoops in with the lap counters, he has heavy ghosting to his right as his sprite passes over the track. Comparing these most noticeable ghosting issues with my 2chip, it does not have them; it is purely a slight smear and not ghosting.

I think at this point, I would choose the 2chip smeared look over this ghosting, especially as the OSSC can clean that up a little with reverse LPF. I might shelve this project, go back to my 2chip and see if any more information about this issues comes to light in the future.

Right, that's a really nice feature on the OSSC. Haven't seen it personally, but if it actually makes an improvement to the 2chip that might be the way to go. Besides you don't loose the little bit of graphics like you do on a 1CHIP with the mod, so there's that

How are your cables wired up? Ghosting is not generally due to cross talk so much as signal reflection caused by impedance mismatch. A PAL SNES cable should be wired like this, with 75 ohm resistors tied to ground (not inline).

Yes that's right they have resistors tied to ground. I have a cable from consolegoods.co.uk that gives ghosting on my 1chip and not on my 2chip, and then I made up a cable here with better shielding that helps with the audio buzz but does nothing for the video signal (to my eyes). To make that cable I followed the diagram below, and I also verified the cable I bought follows this and it does, apart from taking sync from luma.