How I spent my Memorial Day weekend by Sean.
I had been working for a while trying to make an Amiga RGB->Component converter, since pretty much any modern TV has a VGA port that won't take a 15khz signal (even though there is no technical reason anymore why it shouldn't be able to), but Component ports that will.
Being that the RGB port has all the ripe ingredients for making a component signal, I figured it would be easy so I looked online... and came up empty-handed with regard to Amiga-specific circuits. I found a few generic RGB to component circuits utilizing op-amps found only in Europe or Asia or long discontinued or only available in quantities of 1000... in surface-mount form. (Miniaturization has really made the electronics hobbyist's life a pain in the ass.) I also found some instructions from an Australian guy about how to modify an A520 for S-Video. Of course the PAL version of the 520 has component numbers completely different from the NTSC version. So I ended up looking at the datasheet for the MC1377 RGB to Composite chip at the heart of the 520.
Turns out that the 1377 does a couple neat things: It loops the Y and C signals it creates out and back in again for combination. This obviously makes S-Video extraction easy, but also means that it's trivial to unhook the Chroma signal from the output stage to give you a nice, crispy, properly balanced Y for your YPrPb. The other nice thing that the 1377 does is give you a -Y. This makes the mix matrix at your op-amps for the R-Y and B-Y much simpler! I was able to make the whole R-Y/B-Y circuit on a breadboard with just a single LM6182IN dual op-amp that I had lying around, 6 resistors, 2 pots, 3 power filter capacitors, and a 7905 for the Vee (because the A500/A1200 don't have -5v, only -12v). One of the goals for this project was to be able to do it entirely with scavenged equipment or parts from my stash. The R-Y and B-Y circuits are identical, textbook summing circuits, with the pots being used to dial in the mix ratios.
It's not perfect. It was originally a little dark, so I removed the pull-down sides of the voltage dividers that the 520 does at the RGB input, which is kind of a kludge way to fix that problem, and I'm getting some right edge dot crawl from somewhere, which I didn't think was possible without any color modulation. Clearly there is more chroma-related stuff on the 520 that I can strip away and probably clean up the signal even more. The output impedance on my color signals is deliberately wrong because they were running a bit hot and I'm too lazy to do the proper math, so I just dropped in resistors that got me in the ballpark by trial and error. The pots (1K) are a bit touchy. I wish I'd had some with a more gradual taper. Next long weekend maybe I'll dig the scope out of the basement and see if I can tighten it up, but it's certainly good enough for now.
I also learned that CF cards are basically little solid-state PATA drives, so all you need is a physical adapter to connect them to an IDE port. I was able to transfer everything over (at great effort, but that's another story) from my 20 year old 250MB hard drive to a CF card and put it in a safe place. So now I have a silent, component-video A1200 that boots at lightning speed!
All it needs is a new case.
(Yes, the keyboard is a scavenged A500 keyboard.)