[FIXED] NESRGB / Twin Famicom image interferences

Hi all,
I'm done with modding a TWIN Famicom and I have an issue I can't get to troubleshoot.

I'm using the the original RF modulator port.
The issue is I have small interferences particularly visible on the blacks. Like tiny white dots all over the place.
Where it gets strange is that I can remove those interferences by plugging either of the original composite port (image or sound) to any composite port on my tv.

I've ruled out sync issue because the issue is exactly the same whatever sync I use. I have also rewired my mod with bigger wires. Made things better but not perfect.

I even isolated the video composite port by removing r214, but still, interferences are here and disappear when plugging a jack.
To be specific the jack needs to be plugged both ways. If it's plugged to the Famicom but not to the TV the interferences are back.

I just don't understand how those ports I'm not even using could have that effect.
Do you have any idea?

Exemples below:


Twin Famicom W/ NESRGB using RF Out Port. No RCA plugged in :


Twin Famicom W/ NESRGB using RF Out Port. Either RCA plugged in :

... There's two dots per NES pixel? You've got something very weird going on.

1- That's interference at 10.7MHz. There's nothing in the Famicom that should emit energy at that band. (There is a crystal at 21.5MHz, but no-where is there anything that divides it by just two)

2- Composite over-the-air demodulation is supposed to be narrow-band enough that there couldn't be any energy at that frequency mixed in anyway

Maybe check the shield on the RF switch you're using. By plugging in either baseband composite or audio, you add a ground path, so my only guess directly from that symptom is that the coax is somehow damaged.

Do they move if you toggle which RF channel it's using?

Hey Lidnariq, good to hear from you, it's been a while.

I'm not sure I follow what you mean by "There's two dots per NES pixel". The interference I'm talking about are those white dots all over the black area on the top picture.

The bottom picture is in fact perfect but to achieve this I need to plug an RCA between any port of the TV and any port of the Twin Famicom.

Just to make sure we're on th esame page, this is the installation I did :
https://etim.net.au/nesrgb/installation-famicomtwin/

Look under : RGB connection alternative, discreet method.

Yeah, look at this bit:


But you're talking about the random dots, not the grid. I just saw the grid, because I thought it more objectionable.

Anyway, if things get better when you connect composite video or audio, you've got problems with your cable having a lousy ground connection. You don't need to worry about the actual wires that are carrying the signal, but check the structurality of the shield and grounding pins on the connector and the cable itself.

Or possibly the quality of ground connection from the port to the famicom mainboard is lousy.

A problem with the cable would make a lot of sense... It seems like an old one and it's the only Din8 cable I have.

The connector is brand new, (provided by Tim Worthington)

I have tried connecting the ground pin to different point on the NESRGB and the Famicom itself and it did make things better or worse.

Sorry for the dumb question but how do I test the "structurality of the shield".

Also, I'm still not sure I understand why connecting the composite would have an effect on another cable. Does it somehow establish the link between the ground of the TV and the ground of the Famicom (something the other cable doesnt do because it does not work correctly) ?

Again thanks a lot lidnariq.

PS : the double Pixel might be due to the fact that it was one of the latest Sony Trinitron out there so maybe it was some kind of feature.

Useful...

Ideally, your cable should have a braided wire jacket around the outside, like "B" in this picture: https://commons.wikimedia.org/wiki/File:RG-59.jpg

Try measuring the resistance between the two outer metal rings, and especially seeing if it changes as you flex the cable.

Exactly. Although perhaps "doesn't work well enough" instead of "doesn't work at all"

In hindsight, those are probably individual phosphor groups on the trinitron tube itself, not actually a video artifact.

Ok, I have ordered a new cable, better quality, I'll see what happens.

I'll keep you updated. (will be a few days)

Thanks again for your time and knowledge.

Romi

Just to be clear, I suspect this:
is actually the first place to look.

When I did this the white dot were much more visible when the Din 8 ground was wired directly to the nesrgb ground than when it was wired to one of the composite port ground (on the second case the size of the wire is just a few centimeters as the composite port is just next to the Din 8).

In both cases connecting either of the composite port completely removed the artefact.

Does that tell you anything?

Not specifically, other than vaguely validating.

Anything else I could try to confirm your theory?

I mean ground is the obvious culprit. But a faulty cable (or Din) still feel like real bad luck.

Another thing to note. The Twin Fami is supposed to use a 7.5 PSU. I don't have any so I'm using a 9v with the right polarity. I know the Twin can handle 9v but could it somehow generate noise?

Certainly a "better" power supply (smoother) could reduce the amount of noise ... but the noise you see doesn't seem consistent with a lousy power supply. (Switching power supply or transformer-rectifier?)

I think I remember someone saying to be very careful about using a higher voltage PSU with the Twin Famicom because ... something something using the unregulated voltage and it being too high? But that's not a "noise" issue. If my memory is right, that's a "damaging something" issue.

I'd try adding extra ground connections between the A/V jack (both shield and the ground pin on the connector) and various places on the mainboard.

So, waiting for my new cable, I've decided to open the old one to check it.

It seems to be a real mess, no shielding at all, resistance & caps everywhere...


I've followed Tilm Worthington schematic and I've noticed a few strange things


First, the resistance between pin 8 and Pin 16 was disconnected. I've fixed it, and was really hoping it would fix my issue but it did not.

Also, on Tim's schematic, all the Ground pins seem to be connected together. On my cable I have ground on Pin 4 (audio ground) and on Pin 17 (2 separate wires but with continuity between them).

However non of the other grounds are connected (5, 9, 13, 18). I've tried to ground 18, but it did not really change anything.

So that's it for tonight. Any insights?

That's as I'd expect. Ideally each set of signals (audio 1/2/3/6 vs 4 and composite 18/19/20 vs 17) would be separate coaxial cables inside the outermost jacket, to reduce chance of crosstalk.

Similarly I'd hope for coaxial cables on 7/5, 11/9, and 15/13.

I'd try adding a wire between ground and pin 21.

Ho that's another thing, there's no pin 21.

I almost messed up because of this because I thought pin 17 was in fact pin 19. If you look at my picture, you can clearly see the black wire going to pin 17 then next to it, pin 19 is not wired (as expected) and that's it. I have nothing else.

Hm. I'd still try to get something connected to the outer shield, by whatever means.

Ok just did it, it had no impact.

It still feels like this cable is really low quality though. I think I will wait for the new one before experimenting some more.

Thanks again for your help.

Got my new cable this morning. It works perfectly. So much time wasted, for a faulty cable.

Anyway thanks Lidnariq, you were a great help.

Romi