Possible alternative to C=VGA

[adric22]

This is a nifty solution being sold for the neogeo which turns analog RGB into component video, S-video, and composite.   I am buying one to try mounting inside of my Amiga 1000 or maybe the A500 if there is room.  I have confirmation from an Amiga user over on www.amiga.org that this board will work on an Amiga and it looks fantastic.  

Now.  I got to thinking, this might be adaptable to the C128's 80-column output.  I realize that the C128's signal is digital RGBI.  However, I think with the right combination of diodes and resistors the RGBI can be translated into a fixed 16-color set of analog equivilents.  This would allow you to display your C128's 80-column video in three different ways:  Composite, S-Video, and YPrBr (or whatever that is) component video.  So if you have a nice 15" flat-panel TV I think it would make a great monitor for a c128.  

I'm going to be finding out soon how well of a monitor it makes for an Amiga.  really there should be no difference in appearence between using this solution on a 15" flat-panel VGA monitor or a 15" flat-panel TV using component video.  I believe the same signal information is still presented, it just comes in a different way.   In fact, the TV may yield superior results.  I was looking at some of the cheap ones at Wal-Mart the other day.  My dad just bought one that actually has a physical screen resolution of 640x480.   It isn't HD or anything.  But I thought about it and that might be a good thing for guys like us.  That means there won't be any scaling or anti-aliasing going on because it is an exact 1:1 pixel ratio, at least for the vertical pixels anyway.   So when my Neotbitz board comes in and I get it mounted, I'll try hooking my Amiga up to his TV (He just lives down the street) and see how it looks.   If it looks perfect, I might invest myself into a cheap LCD TV with only 640x480 pixels.

Anyway.. any thoughts here before I get started?

[Mark Smith]

It's the "Intensity" part of RGBI that is going to cause grief.

I think what you are going to need to do is have a "widget" sitting in front of the AD724JR IC, the widget is going to translate the 4bits of data coming in over the RGBI lines and then output the correct colour over it's own R G B outputs.

There are only going to be 16 possible inputs to the 'widget":

0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111

You'll have to match those 16 inputs up to the correct colours a c128 shows on it's VDC output, and then have the widget output the correct analog combinations of Red Green and Blue.
The Vsync and Hsync you can use for timing of when to read the RGB lines then pass through to the AD724JR chip.

At least thats how I see it anyway ... corrections greatly appreciated as I'm learning :-)

Actually got me thinking more now ... don't think this is that hard an issue to crack given time ;-)

OK 30minutes later I'm editing this:

How about the RGBI goes into a bit of preprogrammed logic (PAL, GAL, PLD, FPGA whatever), depending on what input is recieved depends on what is output, the output being 3 seperate 8 bit outputs (one 8bit set for Red, another set for green and another for Blue) but only 16 combinations in total that match the VDC colours ... how to explain ...

If you have the RGBI input of 0101 and that happens to be the VDC colour of Red, then you'd want the R 8bit output to be 11111111, B to be 00000000 and G to be 00000000.
If you have the RGBI input of 1101 and that happens to be the VDC colour of Brown, then you'd want the R 8bit output to maybe be 00000000, B to be 11111111 and G to be 11111111.
(8bit values I did are made up ones, have to work out what combination makes brown, what makes yellow etc)

Then those three 8bit RGB values can be fed into a simple RAMDAC (or 2r Ladder), which will give us a nice analog Red Green and Blue, which you can then feed into a whatever you like for YPbPr component, or to push it into a VGA monitor.

So any thoughts on that .. or have I not had enough caffiene lately and come across as rambling ? :-)

Mark

[adric22]

Okay.. here is what I was thinking:



I don't know what values the resistors or diodes need to be. But I see no flaw in the theory here.

Here is an easy way to find out without having the converter board I'm getting. First, build this little circuit and then pump it into the analog-input of your 1084 monitor and see how it likes it. If it likes it okay, then I'm sure the little converter board will like it too.

[Mark Smith]

I'm not so sure to be honest ... on Zimmers there is : http://www.zimmers.net/anonftp/pub/cbm/documents/cables/128-videocable.txt  which is what you are basically describing, I just feel something is being left out is all .. but hey if it does work then I'm all for it :-)

[adric22]

I just examined the diagram there.  I suppose that would work, but I feel as if not having any diodes would allow color bleed-over between the reg, green, and blue.  

I think this weekend I'm going to experiment with just taking the R, G, & B lines and some resistors and going for an 8-color display.  Then I'll start trying to tweak in the Intensity line and see what happens with it.  

One issue I have is that the system I'll be using this with only has 1 sync pin.  so I believe I can just combine them.  I'm going to test it on my Sony PS-One screen since I already have it modded to accept analog video with 4 RCA connectors for RGB and Sync.

[Mangelore]

Okay.. here is what I was thinking:

http://www.my-pc-help.com/images/ebay/schematic1.gif

I don't know what values the resistors or diodes need to be. But I see no flaw in the theory here......

Maybe it's worth while trying to source the schematic diagram of a 1084 or similar Commodore monitor that has an RGBI input. It's possible that these monitors may use a similar circuit for the digital to analog conversion stage... just a thought...

[Mangelore]

Just checked Google and found this link...
http://staff.washington.edu/rrcc/uwweb/1084PAL2/TTL-FACE.GIF