2600 with RGB/Component

[Spector]

Is it possible to mod a crusty old Atari 2600 so that it has displays RGB and is connected through the SCART socket so beloved of us European gamers?

 

I know of A/V mods, but an RGB one would create the best and purest picture possible. Has anyone ever done this?

[+SpiceWare]

no, TIA doesn't output an RGB signal.

[+Nathan Strum]

How about an adapter of some sort to go from an S-video mod to your SCART socket?

 

Otherwise, you can buy an S-video to component transcoder, but they aren't cheap.

Edited March 16, 2007 by Nathan Strum

[+Mitch]

Someone was selling an RGB SCART 7800 in the marketplace forum. As long as you don't mind gaming on a 7800.

 

Mitch

[Spector]

It's no big deal really. The 2600 isn't going to benefit from an RGB mod the way a more developed console would, like the PC Engine. I just wanted to find out, because you can't beat RGB.

My AV modded Atari seems to have blown. The picture is buggered, so I might have to get a new one. Anybody you recommend?

[mimo]

Someone was selling an RGB SCART 7800 in the marketplace forum. As long as you don't mind gaming on a 7800.

 

Mitch

 

I was, but it is now on e-bay

Lovely machine, can't understand why the 7800 is so unloved

[supercat]

I know of A/V mods, but an RGB one would create the best and purest picture possible. Has anyone ever done this?

 

The TIA does not generate RGB video internally. To produce an "RGB mod" one would either have to (1) time the first rising or falling edge of the chroma output signal following the proper edge of the system clock, relative to the system clock, and then convert that plus the luminance output into an RGB signal, or (2) build a piggy-back board with a TIA on it and a CPLD sitting between the TIA and the system bus. The CPLD should capture A0-A5 and D0-D7, phi2, and R/W, and output its own version of D1-D2 (the other data bits could be left as-is).

 

The function of the CPLD would be to (1) latch all values written to COLUxx, and (2) cause all writes to COLUBK to store $x0, writes to COLUPF to store $x2, writes to COLUP0 to store $x4, and writes to COLUP1 to store $x6. The luminance outputs of the TIA would then be used to select one of the CPLD's copies of the color registers.

 

Using this approach, the output of the system would be absolutely pixel-perfect. Not sure it would really be worth the effort, though.

[Urchlay]

The function of the CPLD would be to...

 

Using this approach, the output of the system would be absolutely pixel-perfect. Not sure it would really be worth the effort, though.

 

That would be *cool*!

 

Probably not worth it for the 2600, but I bet everybody on the Atari 8-bit forum would want one, if such a design could be done for the GTIA. Correct me if I'm wrong, but that would allow the Atari to directly drive an SVGA monitor, wouldn't it? That's kind of a "holy grail" for a lot of 8-bitters... I'd definitely buy one.

[supercat]

Probably not worth it for the 2600, but I bet everybody on the Atari 8-bit forum would want one, if such a design could be done for the GTIA. Correct me if I'm wrong, but that would allow the Atari to directly drive an SVGA monitor, wouldn't it? That's kind of a "holy grail" for a lot of 8-bitters... I'd definitely buy one.

 

I hadn't thought about the Atari 8-bits. I'm not into such machines myself, but the approach should probably be suitable there though I couldn't say for sure without knowing how the ANTIC and GTIA interact. If a particular monitor can deal with a 15.75KHz scanning rate, it could be connected directly. Otherwise it would be necessary to add scan-doubling logic (which would require either adding over 320 bytes to the controller chip (i.e. using a somewhat advanced FPGA) or else adding an external 512 bytes of RAM (which would require adding an extra 11 pins to the CPLD).