Yet another RGB modded PAL N64

It has resistors and capacitors inside??? First I've heard of it.

 

No.

PAL SNES RGB uses resistors.
NTSC SNES RGB uses capacitors.

 

THS7314 wired internally to the N64. I was thinking the same thing about the current being wasted but I only have the one SCART cable to use with SNES and N64 hardware. Tempted to try swapping out the THS7314 for the THS7374 now that I have a hot air tool to use.

 

It has capacitors inside for sure. Because it doesn't work with the old model SNES (non 1CHIP-models). When I tried one of those consoles with the cable the picture would get darker and almost entirely black within a few seconds. Using a different cable without caps solved this.

Anyway I made another RGB cable for the N64 with no resistors or capacitors and it doesn't make a difference in the flickering. But I tried the N64 on an old CRT I have and it worked perfectly with no noticeable flickering at all. So I guess it's just the XRGB being overly sensitive. Wouldn't be the first time...

 

PAL SNES RGB cables have resistors AND capacitors.

And apparently some NTSC are straight-through without any components... These are the only kind I was aware of, but I've never opened an official cable before to know there were capacitors inside.

Then this is totally wrong--when you use a video amplifier the signal is doubled and you're supposed to put a series termination resistor at the source, which effectively divides the signal back to the correct level and impedance matches the coax.

1 V * 2 / (75 + 75) * 75 = 1 V signal (correct) and the current is 2 V / 150 = 13.333 mA

Because you're using a PAL cable with resistors to GND, this means both 75 ohm resistors--the one in the cable and the one in the TV are in parallel, and are delivering the signal at twice the correct value, improperly terminating the source, and overloading the amplifier (usually amps are designed to handle 2 loads = 26.666 mA)

1 V * 2 / (75 // 75 == 37.5) * 37.5 = 2 V signal (wrong) and the current is 2 V / 37.5 = 53.333 mA -- brutal for a video amp.

Since you have 3 channels that's 160 mA, when you could be using 40, and have the proper termination and signal level. Pull up the resistor legs to GND and put them in series and all is fixed.

Huh? This would make sense because I thought the original SFC/SNES had the capacitors internally--putting two in series would reduce the value by half which mean bad low-frequency response.

Or do you mean PAL old model SNES? If you removed the resistor to GND from the PAL circuit nothing would work either (or use a NTSC cable).

But what you do mean non 1CHIP models, 1CHIP aren't supposed have RGB wired at all, right? Maybe no SNES models have the capacitors internally.

Where are you getting sync? Composite or Csync off the DAC and how are you wiring it? If the flickering is losing sync, then that's where to look.

 

Yes I mean the old 2CHIP PAL models.

The 1CHIP are wired for RGB. As far as I know only the SFC Jr and SNES 2 (or whatever you want to call it) needs modifying for RGB.

I'm getting sync from the Csync pin on the DAC. Loss of sync is not a problem, the picture is rock solid. The problem is the flickering waves across the screen. Fudoh on shmups suggested that it might be a power supply problem as his friend tried this mod too with his XRGB-mini, and it too had these flickering waves. They apparently disappeared when he changed the PSU.

 

I see what you mean about the "1CHIP", I thought you were referring to Jr models, since "1CHIP"/SoC revisions are rare or maybe don't exist for NTSC non-Jr systems. All PAL models need the PAL cable, which has the 75 ohm resistors to GND, that's the key.

Remember the CPLD is a digital component, it only outputs ~3.3V or ~0V where sync is a pseudo-digital signal of 0V and -300 mV with termination etc. It's very improper to directly wire sync to the CPLD. Sync is a 75 ohm load, logic inputs are typically 5k-100k+ ohms! 3.3 V / 75 ohms = 44 mA = WAY more current than the CPLD can comfortably supply. The proper way to interface a digital output to sync is: CPLD output -> 220 uF series capacitor -> resistor network which attenuates 3.3V to ~600 mV with 75 ohm output. At the very least use one of those 560 ohm resistors in series. CPLD output -> 560 ohm -> sync in on TV. That way the sync voltage will be 390 mV (not correct, but OK), and it'll only need to supply ~5 mA.

 

1CHIP is an earlier completely different board than the Jr.
Jr has further changes, such as built-in controller ports. At last no more daughterboard.

 

I guess I should wrap this up. I modified another US console and used three 100uFs like Calpis suggested but it made no difference to the noise. I also made a RGB cable with no resistors and caps inside like suggested, but that too didn't make any difference in the noise. The picture was slightly brighter of course.

A 680ohm resistor on the sync line turned out to be a good idea though as the picture was crazy unstable on my Samsung plasma when connected directly

I can live with the noise for now, but if I get the change to pick up an old US or JAP machine that qualifies for the old mod I'll probably do that

 

hello,
maybe this is useful

here is my first mod with the vq44 package i made a while ago.
i attached a few jed files

9536xl
pc44 and vq44 package

9572xl
pc44 package


here is a picture of the nice pcbs desinged by someone of our national "community-board" ^^








greetings, saturnu

 

Those boards look pretty slick!

 

Nice... Tons of PCB for a bad design -_-

-Logic level sync output?
-The standard R2R DAC = wrong levels, high output impedance so bad source termination and poor bandwidth
-The 9572 actually has enough macrocells to register R, G and B components simultaneously instead of sequentially so there isn't pixel smudging

 

Maybe the design isn't perfect but there is no noticeable picture-quality difference between the NUS-001 (FRA) SMD-Mod and the replacement DAC.

<font size="2">

 

So will these mod boards work with denc-nus rs5c282 If so pm me please.

 

Those PCBs are beautiful. Is it still possible to get one? Someday I'd like to perform this mod on a latter revision of the NTSC-J N64.

 

I wonder... Does this RGB mod (replacement DAC) produces image as good as the old amp mod (with THS7314) or is it better? Or worse?

 

I've only seen pictures of the image quality the old style mods outputs, but from what I've seen the old mods are better (as in sharper) than the new DAC. The DAC-mod is still much better than composite and a little better than S-video, so if you have a newer N64 where the NJM/transistor amp isn't possible, then by all means go for the DAC.

Keep in mind that the DAC is much more expensive in materials and also quite time consuming, especially the first time you build it. For me it gave a great sense of accomplishment when I finished the first DAC and it worked flawlessly the first time. Never tried to program an IC or CPLD, but I was surprised how easy and painless it was (with help from members posting in this thread of course ).

 

Of course the CPLD DAC isn't as good, every pixel bleeds into the next with the given logic! Fixing the logic and building a better output circuit would allow it to match/exceed the silicon DAC's quality.

 

I noticed you pointed that out several times before. Maybe you could fix it? Just joking, I know it's time consuming. I have an NTSC-U console with an amp mod and I like the picture. Colors are great but maybe a little dark. Not that the image is dark or something, I just think that it's a tiny amount darker than it should be. I did compare it to the composite output and the brightness is roughly (if not completely) same. So maybe that's how every N64 outputs its image.
That's why I was wondering if the custom DAC produces better image. Now that I know it doesn't I can finally enjoy the RGB mod as I understand it's (nowadays) the best mod available.