Re: Analog NTSC to RGB converter?

[Michael J. Mahon <mjmahon@aol.com>]

Linards Ticmanis <ticmanis@gmx.de> wrote:
> On 10/22/2012 05:18 PM, Linards Ticmanis wrote:
> 
>> I'll report on my results as soon as I have some - in case anybody is
>> interested.
> 
> OK, the main part is done. It even does auto-detection of PAL or NTSC
> mode, which I didn't know was even possible until Alex Freed rightly
> suggested that I read the data sheet more attentively. All in all it's
> working really well.
> 
> There are two remaining issues: First, the filter that removes the
> chroma info from that signal which goes down the luma path still has to
> be re-tuned by hand for either PAL or NTSC. Not too convenient as it is
> deep inside this beast. As Alex has suggested, this needs to be switched
> between two differently tuned filters just like the two quartzes are
> switched; but currently I'm having a hard time finding out what exactly
> to get for the second filter. According to the schematic the filter is a
> tunable inductor and a capacitor in series (inside a grounded metal can,
> as a single component), but no pF or µH values are given and I don't
> understand the inscription on the component itself either.
> 
> Second, no matter how you tune the filter, the Apple II signal seems to
> be a bit too "violent" for it. With the beam moving from white or black
> into a colored area, it takes about 2-3 chroma cycles until the
> filtering really gets going. No such problem when watching DVB TV, a DVD
> or VHS through it, so I guess the fact that an Apple color signal is
> more or less "pure" color carrier, highly saturated and instant-onset,
> defeats the circuit until it's had time to adjust. I guess it's a trade
> off with the fact that the monitor has a much higher bandwidth than a
> normal video monitor - 80 char text is quite legible on it.
> 
> I'll stop replying to myself now.

The chroma trap is, as you've observed, a series-tuned circuit. If there is
a PAL/NTSC level generated by the decoder IC, you can use it to control a
relay that would switch in an (adjustable) additional capacitance in
parallel with the trap capacitor. Hopefully, the signal goes to the
inductor and the capacitor goes to ground, but it will work either way. 

Align it by tuning for PAL chroma rejection with the relay contact open,
then adjust the additional trimmer capacitor for NTSC chroma rejection with
the contact closed. 

Alternatively, if you don't mind seeing the pixel structure of colors (like
most RGB monitors), just remove the chroma trap entirely.  That will
improve the bandwidth even more. (This option won't work if the chroma
signal is taken from a second winding on the chroma trap.)

The only time that it can get annoying seeing the chroma signal in the
luminance is in fairly large areas of relatively saturated colors (like an
Apple II display ;-)--but RGB monitor users seem to get used to it. 

-michael - NadaNet 3.1 and AppleCrate II: http://home.comcast.net/~mjmahon