Quote:
Originally Posted by ChrisW6ATV
Since a component (Y/R-Y/B-Y, also called Y/Pr/Pb) input will need to be matrixed differently from the original Y/I/Q demodulated signals...
|
That might not be entirely true. My own tests on my Philco 17MT80 seem to indicate that as long as you get the red and blue levels more or less correct, the green will follow. Remember: regardless of decoding method the amount of each color required to make white on the screen remains the same, and this will be reflected in the matrixing circuitry. As long as you don't fool with the values of the passive parts of the color amps, it stands to reason that as long as the proportion of red/blue are correct so will the green also be correct. But none of this is the difficult part...
The tough part of all of this is making 2 custom delay lines, because with component the color gets to the CRT before the luma does. Blue is less noticable than red is, since there's a greater proportion of red in almost every scene. What's needed are a pair of properly terminated delay lines, whose delay is custom tailored to the set in question. Since there will be a voltage drop across the line, another stage of amplification will likely be needed. I'm experimenting with 6GH8 (yeah yeah, they are laying around) pentode sections to preamplify the signal before it's sent through the line, but it's not an exact science. Ghosting is a big problem, as is crosstalk. The entire thing must be designed correctly or the result will be horrible. But if and when I get it right, the result will be nothing less than spectacular. Perfect convergence, luma to 4Mhz with no dot crawl, and color with digital accuracy. It will rival the picture quality of any NTSC set, including my CT55. Pretty ridiculous for a set I got for free.
