The composite video signal in a Duo is made up of the following circuit (see attachment, Pxx = VCE pin, CVBS = composite video output, actually goes through some transistors and bypass capacitors beforehand)

What's interesting to me is that several VCE signals are combined to form the composite video output prior to amplification. I wonder the pins correspond to something like R-Y, G-Y, Y, and sync+blanking or burst. The regular composite sync pin isn't used.

If this was the case, it may be possible to get S-Video output from the individual pins, some trickery would be needed to coax R-Y,G-Y,Y back into Y/C.

The other reason I think this might be the case is that the VCE patents describe the internal operation as working entirely in a split R-Y,G-Y,Y model for the composite video output section rather than RGB, which is handled in a different part of the chip. Considering other video chips (TMS9918 series) had this kind of output, it doesn't seem like too far of a stretch.

UPDATE:

U.S. Patent 5,389,949 has an almost identical circuit for the description of the composite video output at Fig.14, but the wording indicates it's internal to the VCE. Also the SuperGrafx looks to have the same circuit as the Duo.

[Updated on: Thu, 08 February 2007 19:21]

Administrator wrote on Thu, 08 February 2007 20:20

The composite video signal in a Duo is made up of the following circuit (see attachment, Pxx = VCE pin, CVBS = composite video output, actually goes through some transistors and bypass capacitors beforehand) What's interesting to me is that several VCE signals are combined to form the composite video output prior to amplification. I wonder the pins correspond to something like R-Y, G-Y, Y, and sync+blanking or burst. The regular composite sync pin isn't used. If this was the case, it may be possible to get S-Video output from the individual pins, some trickery would be needed to coax R-Y,G-Y,Y back into Y/C. The other reason I think this might be the case is that the VCE patents describe the internal operation as working entirely in a split R-Y,G-Y,Y model for the composite video output section rather than RGB, which is handled in a different part of the chip. Considering other video chips (TMS9918 series) had this kind of output, it doesn't seem like too far of a stretch. UPDATE: U.S. Patent 5,389,949 has an almost identical circuit for the description of the composite video output at Fig.14, but the wording indicates it's internal to the VCE. Also the SuperGrafx looks to have the same circuit as the Duo.

I tested the pins today. Pin 20 is the color burst signal only. Pin 40 is Y and composite sync. Pins 33 and 26 are color subcarrier signals (weird that there's two of them). I looked at an s-video signal from my gamecube and the chroma signal had the color burst part. Looks like we should be able to build a s-video output from these 4 pins of the VCE, just by combining a few pins with the right resistors.

You could probably build a component output too, but you need to convert the phase differences of the two chroma signals into linear voltage levels (which is a bit beyond my experience ).

[Updated on: Sun, 30 May 2010 21:53]