[morganw]

I want to feed a projector over a long component cable from both a DVD player and TiVo, which only has s-video out.


I know that production transcoders ($1k & up) exist that handle this and I know I could just run an s-video cable along side the component one. I also know I won't get any quality improvement: all I'm doing is moving the chroma demod from the projector to my AV rack. I'm just interested in whether this could be done for less than the cost of a good 25' s-video cable ($58.95 @ A2Z).


--


The Y signal in Y/C is the same as the Y in Y Pr Pb (though it might need scaling).


I think the C is QAM I and Q chroma signals, just like in composite video. Separating it into Pr and Pb is not as hard as Y/C separation, but it's a little tricky:


The signals are modulated on 3.58MHz carriers, 90 degrees out of phase which each other. A phase comparator can extract a single phase-modulated signal, but how do you do QAM demod?


There are cheap chips for doing this in TVs. NTE distributes a few. Most are RGB-oriented for feeding guns, so not really appropriate (they take both Y and C from the output of the comb filter).


Any video engineers have some insight into the difficulty?

 

[Glimmie]

Quote:

Originally posted by morganw I want to feed a projector over a long component cable from both a DVD player and TiVo, which only has s-video out. I know that production transcoders ($1k & up) exist that handle this and I know I could just run an s-video cable along side the component one. I also know I won't get any quality improvement: all I'm doing is moving the chroma demod from the projector to my AV rack. I'm just interested in whether this could be done for less than the cost of a good 25' s-video cable ($58.95 @ A2Z). -- The Y signal in Y/C is the same as the Y in Y Pr Pb (though it might need scaling). I think the C is QAM I and Q chroma signals, just like in composite video. Separating it into Pr and Pb is not as hard as Y/C separation, but it's a little tricky: The signals are modulated on 3.58MHz carriers, 90 degrees out of phase which each other. A phase comparator can extract a single phase-modulated signal, but how do you do QAM demod? There are cheap chips for doing this in TVs. NTE distributes a few. Most are RGB-oriented for feeding guns, so not really appropriate (they take both Y and C from the output of the comb filter). Any video engineers have some insight into the difficulty?

You have all the basics. The C is exactly the seperated chroma signal and could be modulated I&Q or equiband which is RY/BY. The demodulator process is as listed and like you found there are countless chips that do all these functions in one package, well almost everything.


First you need to re-generate two 358 subcarrier signals locked to the burst of the incomming chroma. The second signal is simply 90degrees out of phase with the first signal. You will need to seperate sync and make a pulse to gate the burst. This typically feeds a crystal oscillator.


Next the chroma is applied to the demodulators. One is fed the 0 degree 358, the other the 90 deg 350. Out comes R-Y and B-Y respectivily.


Heres's a gotcha. That step took time. Typically seveal hundred nanoseconds. Could be much less with Svdeo only because you don't need a band pass filter on the chroma which is a delay line in it's self. So you need to run the Y signal through a delay line to match the RY/BY. From there it's a resistor matrix to RGB and line drivers. You will need to provide an inverted Y signal to also feed the matrix. Or just buffer the RY/BY if you don't need RGB.


Clamping here would be a good idea but adds a bit more complexity.


SUGGESTION:

Find a old broadcast demod on EBAY or at a surplus delaer. It's going to be much simpler to modify that than to build one from the ground up.


Yes it's less than $50 in parts but you really need a basic 20mhz or better scope to build something like this and a lot of time.