[sdiy] Quadrature VCO??

[Magnus Danielson]

From: Tim Ressel <madhun2001 at yahoo.com>
Subject: Re: [sdiy] Quadrature VCO??
Date: Tue, 1 Jul 2003 16:53:41 -0700 (PDT)

> Yo,
> 
> --- Magnus Danielson <cfmd at swipnet.se> wrote:
> > Now, _PLEASE_ tell us what you need it for... I'm so
> > curious ;O)
> 
> If you insist:

But I do, I do... ;O)

> I was looking at my really old RCA IC
> data book when I spied an IC for decoded color TV
> signals. It uses two balanced (de)modulators fed by a
> quadrature oscillator.

Yeap. In the beginning of each line you have a color-burst, which is actually
a few cycles of the color carrier signal. This is fed to the color carrier
resonator/PLL crystal oscillator so that it keeps in lock. The reason it is
needed will be clear soon.

Both PAL and NTSC uses the phase of the color carrier to convey the color,
the amplitude of the color carrier to convey the color-depth and then the
luminance (strength of the pixel) is the normal TV channel. If you get a phase
error in the local carrier replica, then you will get a color-fault. Another
reason of phase error is due to multi-path errors and other phase errors in the
transmission-path.

This is the technical reason why NTSC has been nicknamed Never Twice Same
Color. In PAL they did an interesting trick so that ever other half-frame
(a field) has the polarity reversed the other way around on one of the
references. This makes the phase error turn the color an equalent amount the
other way. Since it is a half-frame it will be every other line, so the phase
errors causes color errors of opposite sign, so what the eye sees is a mixture
of +colorerror and -colorerror which is more or less without color-error.
Actually a very simple and clever solution to a rather difficult problem to
solve in "simple" analog domain for a quite complex problem.

With the comming of cheap digital processing things like ghost image reduction
also lends itself to avoid incorrect phase of color even in NTSC. ;O)

The luminance signal Y you mix up from RGB like this:

Y = 0.30 * R + 0.59 * G + 0.11 * B

I have been fashinated by all this myself. In my work I still need to recall
important facts about analog television systems even if it usually seems VERY
digital these days.

> I couldn't help but wonder what that might sound like.

Ever listend to a modem? They use QAM-modulation, i.e. quadrature amplitude
modulation or also known as I-Q modulation (In-phase and Quadrature signals,
i.e. Cosine and Sine). The modulation input could however be better selected.

> Then I remembered a project for a direct-conversion radio receiver that used
> the same technique.

It is not only a classic, it's a necessity if you want to do usefull stuff.
Frequency shifters is really what radio-people refer to as SSB (Single Side
Band) modulation. Heck, much of the terms in synthersizer-world still can be
derived from the radio days.

> the article said when heard in stereo headphones it sounds rad. So I have
> decided to give it a try. I'm going to this:
> 
> http://www.maxim-ic.com/appnotes.cfm/appnote_number/196
> 
> --as my inspiration.

OK. Good luck!

Cheers,
Magnus