NES RGB Video Issues

[fleeisme]

So I finished modding my AV Famicom with Tim Worthington's NES RGB board, and everything works quite well (RGB, S-video, and composite), although running RGB over a c-sync scart cable to the OSSC, there appears to be faint low signal like jail bars. This appears more on lighter colors. I used very short leads to wire the board to the AV multi out, and I have c-sync, R, G, B, Video, Y, and C currently wired up, and cut the original video trace.
Basically I was wondering if anyone else had this issue or similar ones and was wondering if there are any common fixes to tune up the signal. For example, should I wire up the ground to the multi out, or anywhere else? Or perhaps disconnect the video line all together. Or even add an rgb amp of some verity. Any suggestions are appreciated. I can provide pictures when I ever I sort the capture device out or just snap a few straight off of the TV.

Note: The cable in question is flawless on my SNES 1-chip Rev 2 with absolutely no noise. And since I have all NTSC consoles, I thought that the capacitors used for SNES SCART cables may be the culprit, but I have no way of confirming this because it is my only Nintendo SCART cable.
Thanks in advance for any help.

[DirkSwizzler]

I have been seeing the same issue with my AV Famicom with Tim’s board. I only started noticing it when I got my OSSC which brought the jailbars into sharp focus.

I also tested against my snes mini since they both used the same spec cable. And the snes mini had no issues at all.

If you google around, you’ll find that people have been hitting this with other RGB amps for quite a while. And there are about a half dozen nutty things to try. Most of the suggestions have been accumulated on this page http://jpx72web.blogspot.sk/2016/11/fam ... w.html?m=1 . The best (but not perfect) fix i’ve tried so far is an adaption of “lift pin 21 of the ppu”. Where I cut the trace on Tim’s adapter board.

I read that Tim’s board still needs pin 21’s output. But we can still isolate it from the main famicom PCB. The av famicom adapter board gives you a very easy line to sever after it’s been assembled.

I’m literally in the middle of this project now trying to find a 100% fix. So let me brain dump my experience so far.

Originally I asked the Retro Roundtable guys for advice on better bodge wire because I was convinced mine was too long.

Voultar mentioned that Tim’s board already has capacitors on the rgb outputs and that people commonly pair it with snes scart cables with caps in the cable. This halves the capacitance so he suggested removing a set of capacitors. I removed/bypassed the set on Tim’s board. No noticeable improvement, but it’s closer to spec so that’s good.

He also mentioned that there’s been an issue with a improper voltage regulator installed for most of Tim’s boards sold in 2017. I replaced my regulator with the correct one. No noticeable effect, but my ocd is relieved.

I cut the trace feeding ppu pin 21 back to the main pcb and that reduced the jailbars quite a lot. But they’re still there.

I wrapped the ppu body with copper tape and grounded it to pin 20. No significant effect.

I added a bunch of different sized capacitors (.01uf,.1uf,1uf,10uf) between pin 40 and ground on the PPU. No significant effect, but that same fix worked for my supergrafx so I’m keeping it anyway.

I desoldered transistor Q1 just in case it was picking up any noise still, no significant effect.

I desoldered the CPU and wrapped the body in copper foil and grounded it. No significant effect.

I added another set of capacitors between pin 20 and 40 of the CPU. No significant effect.

I was reluctant to try the capacitor between pins 20 and 22 of the PPU because I was unsure it would be a purely positive change. It had no visible effect with a 1uf capacitor. I’ve read people try other values but I’m probably going to remove that cap altogether.

All of the above may have had a cumulative effect. So I’m keeping most of the changes. But I still have faint jailbars.

The next thing I want to try is removing the multi-out connector and severing the PCB traces connecting the grounding pins to the ground plane. The using a bodge wire to try alternate grounds. Maybe hook up a second voltage regulator to supply Tim’s board with completely separate power and ground from the main PCB.

I read somewhere that a guy modding an original famicom using a different board noticed that his jailbars only showed up if he used the main PCB’s ground. And they were gone when using the new mod’s independent ground. So it sounds like the two main problems are pin 21 picking up noise and a noisy ground plane. So I’m hoping to supply a different ground that’s somehow cleaner.

I’m not an EE, so I don’t really understand the best practice for removing noise from a ground plane except to maybe stuff as many capacitors as possible around the board between power and ground. But I’m unsure if even that is a good idea.

Anyway, if you don’t want to dive deep. I suggest removing the rgb caps on Tim’s board and severing the pin 21 trace on the adapter board and learning to live with a little bit of jailbars.

[Ripthorn]

I've never heard of jailbars on AV Famicom. Only on regular NES and Top Loader.
I've had 6 manufactured in different years, 3 of them with NESRGB, none with jailbars. And as far as I know, NESRGB helps to solve the jailbars issue.
Maybe the problem is on NESRGB board.

[DirkSwizzler]

While it’s always possible that Tim’s board is the cause. I highly doubt it for several reasons:

1. The fixes page I linked is specifically for Famicoms. And it points to a long discussion on a famicom messageboard. So clearly famicoms have issues as well as the American versions. So your data of only the American versions having problems is already in question.

Additionally, the famicom world thread that the fixes are based on seems to have a loose consensus that the toaster had the least amount of jailbars.

2. The fix of isolating pin 21 from the main PCB had a major effect. And since that’s the unmodified signal from the PPU going into the stock PCB. It’s pretty clearly an interaction between stock components in the stock configuration.

3. The remaining jailbars after the pin 21 fix are identical between what I’m seeing and what they were seeing. A brightened line approximately every 8 pixels.

4. Like I said. I had been playing my modded av famicom for the better part of a year before I started noting the jailbars while playing Super Mario Bros on an OSSC. So they’re not so pronounced that you immediately see them on a crt.

[fleeisme]

Thanks for the replies, I'm going to try severing pin 21, but I've noticed that the jail bars are less pronounced on a heavy duty power strip. That makes me wonder if it's a power problem. What power supply are you using? I'm using Ben Nahum's fantastic RetroDC, although this makes me wanna dig up my Famicom official power supply despite the 10 volt difference. Just a thought.

[Syntax]

If you want to test if its a power issue try using a 10-12v battery as a power supply.

[DirkSwizzler]

Im using one of these https://www.digikey.com/product-detail/ ... 02-3806-ND (possibly with a polarity inverter)

I was waiting for the RetroDC to come back in stock over the summer when I asked HD Retrovision if it was really worth the wait or if they had other power supply recommendations. It turns out that they were already on the cusp of removing it as a recommended power supply because the quality was inconsistent between batches.

Ste recommended a different power supply on amazon (the pwr+ I believe) that I also use for a bunch of consoles. I tested that one with my famicom to make sure that wasn’t the cause. I also split the one I linked above with my snes and it isn’t experiencing any jailbars. So I don’t think that’s the cause.

I mainly like the one I linked above because CUI’s 12v supply is the most stable one I could find.

From what i’ve read the jailbars are from internal components making the 5v line noisy. So sorting out your power supply is definitely going to have an effect.

[fleeisme]

I can't speak for others, but I have had great luck with my RetroDC. The Genesis and combinations thereof have absolutely no noticeable video noise (surprising to me for how notorious they are for jail bars). I can't deny that it could be the issue however. Perhaps I could test it on a Super Famicom, which NY brother currently has. I'll update with anything I find and attempt to fix.

[DirkSwizzler]

RetroDC was definitely good at some point to acquire a recommendation from HD Retrovision. It’s just that it was inconsistently good.

Yours might be great. It might also be less than great.