Quote:
Originally Posted by zenith2134
I'd love to hear more about those chroma circuits, old_tv_nut...
|
Ok -
One of Norm Parker's inventions was the idea of directly demodulating the composite video (containing luma with chroma subcarrier riding on it) into red, green, or blue. Now, you can't do this correctly without some manipulation, because, unlike the RCA dot-seqential system, the NTSC subcarrier has the wrong amplitudes of R-Y, G-Y and B-Y components.
In the NTSC transmitter, B-Y is divided by 2.03 and R-Y is divided by 1.14 before modulating the subcarrier (actually, the equivalent factors are in the formulas for the I and Q signals). Also, the luma signal consists of .3 R + .59G + .11B instead of .33 R + .33 G + .33 B as they are for dot sequential. So, direct demodulation gets the colors wrong. It's necessary to adjust the amount of chroma signal going into each color channel. R-Y needs a boost, B-Y needs an even stronger boost, and there is too much G-Y, (because there's already a lot of G in the luma) so G-Y needs attenuation.
What Moto did in these discrete diode demods is to make the luma be a common-mode signal for three demods, and the chroma a differential mode, with the required amplitude for each. So, all three demods have their required chroma floating up and down on the luma signal. At the output of the demods you now have the sum of luma and a demodulated color difference, that is, each one puts out R, G, or B directly.
When they later went to an IC demod, they used the same principle to put a luma current in common mode into the bottom of each differential mode chroma demod, getting R, G, and B as outputs.
Norm also patented a method to restore the upper I sideband before demodulating on the R-Y, G-Y, and B-Y axes, but this was deemed too complex to put in a set design, especially before ICs.
So, the set ended up with equiband direct R,G,B demods.
By the way, when I say "Norm invented" I mean that he developed the theory and some ideas of implementation, but there were others who worked on various ways to realize the implied functions.
Edit: fixed the B-Y and R-Y factors (had them reversed)