exxos's Atari, Amiga & retro forum

It seems the noise relating to the colour signal itself, I thought as the pot was basically being rooted around the clock generator circuit, it was picking up interference that way. Long story short, I cooked that signal and created a new pot right on top of the video board, but it did not seem to make any difference.

Adding the series resistors in the clock signals did not have any noticeable effect.

I pretty much think the noise I am seeing is actually generated on the monitor itself, as I know the picture isn't great in the first place..

There is also the fact that this is only composite video that is basically crap to start with

I'll get mine built as soon as I can to do the monitor compare thing.

Something interesting, aside from my dodgy camera work, the first images with colour signal buffered, and the second image is with the signal not buffered.. The one without the buffer I cannot see any distortion in the blue at all... Bizarre...

Even more odd, I now seem to have odd colour bleeding on top of the screen which I didn't have before...So I suspect the buffer is making the signal sharper, but is making it that sharp that is issuing up all the imperfections in the video signal

Another test below,The background of the school at the top as green blobs in the white with unbuffered, but the blue looks more solid with less interference.

Overall the buffered one does look better.


BUFFERED

UN-BUFFERED

Interference on the black seems to be relating to the 4.43MHz gen circuit somehow..

If I load the collector of Q5 with something like a 2.2 K resistor, the black goes blacker, and interference obviously gets less as well as you cannot see it.

It has a similar effect if I increase the value of R46... But basically the 4.43MHz to peak voltage is reduced, however when it is reduced too much, the video just goes nuts. So there is a fine line between working and not working

What I am tempted to try next, is to is basically lift R47 (and remove Q5) and just feed in a dedicated 4.43MHz oscillator to see what happens...

My theory is that reducing the voltage on the oscillator circuit is reducing the interference, but there comes a point where the voltage is simply too low to drive the TIA correctly, so the video goes nuts. but if I can feed a oscillator in directly and remove all those components, This should eliminate a lot of the Interference at least....

...Assuming such a oscillator exists, but a quick look seems not

So, progress, but I didn't have all of the parts I needed, so off to order them now.

Topside:
Underside:
I didn't have any SMD caps, but hoping these should do the trick.

Fingers crossed I'll have it finished and installed next week. Coincidentally, I was given a Dell LCD display with composite and S-Video inputs the other day, so it'll be interesting to test on that.

That bottom 1k you may want to experiment with.. 2.2k might help reduce the black level.. To high value and noise starts to get bad.

What's annoying me is I want black to be black (not grey) but there's no way to reduce it without reducing the colours as well, reduce to much and it goes to monochrome,

I'm wondering if there is a way to check black colour and reduce the voltage on black only.. Though its getting complicated then..

Oddly I spied RGB kit just, they replace the TIA video with a pld. So gonna get me one of those.., they not cheap tho.

Ok, thanks will look at that once built and installed.

Also, I was wondering if R9 is replaceable with 2.2k and R6 with 560r? Only asking as I have these in stock and would need to order some otherwise. No biggie if not, as I have to order some of the others that I definitely don't have to hand.

Don't know which is R9.. You need to use the exact values as I posted a couple pages back. There is a heck of a lot of voltage dividers and biasing going on so changing any values for alternatives will likely not work correctly.

Ahh, I must have missed that post, I'll get the correct parts on order then, thanks.