A1200 Dying? Odd Colours

[Badders]

I've used a screen capture of Workbench 3.1, opened up the image in Gimp and used the colour picker to determine the RGB values of the gray background. All 3 are 170.

Booted up the Miggy and went into palette. The default background is set to 170 for RGB and shows as pink. I've changed the sliders and can get various other colours. It just seems to be missing Gray which is well weird.

Would replacing the Lisa chip also fix the no signal on the composite connector? I have no desoldering experience and these chips are all surface mounted, not socketed. I'm likely to totally destroy the board if I have a go lol.

The revision 2B board should be here the next couple of days. Hopefully that will work fine.

Thanks for all the help and suggestions.

Pete.

[FOL]

Originally Posted by Badders

I've used a screen capture of Workbench 3.1, opened up the image in Gimp and used the colour picker to determine the RGB values of the gray background. All 3 are 170.

Booted up the Miggy and went into palette. The default background is set to 170 for RGB and shows as pink. I've changed the sliders and can get various other colours. It just seems to be missing Gray which is well weird.

Would replacing the Lisa chip also fix the no signal on the composite connector? I have no desoldering experience and these chips are all surface mounted, not socketed. I'm likely to totally destroy the board if I have a go lol.

The revision 2B board should be here the next couple of days. Hopefully that will work fine.

Thanks for all the help and suggestions.

Pete.

Does RF work?

Also try the old trick, push firmly down on LISA, see if brings the colour back, maybe a dodgy joint.
Obviously, not to hard if you still have the lower shielding attached.

[Jope]

Originally Posted by Badders

Booted up the Miggy and went into palette. The default background is set to 170 for RGB and shows as pink. I've changed the sliders and can get various other colours. It just seems to be missing Gray which is well weird.

Try all of the primary colours R G B one by one to see whether they affect the preview colour on the screen. You should be able to identify the faulty/missing colour component then. If it's pink/purple, it sounds like you're missing green?

There is no such colour component as grey, grey is a mixture of red green and blue all at the same value (such as 170 for the default 2.x grey background).

[8bitbubsy]

Originally Posted by Jope

If it's pink/purple, it sounds like you're missing green?

Uh what? White = (RGB 255, 255, 255). There's a lot of white in his picture -- he's obviously not lacking an RGB component.
Also I meant Lisa in my previous-previous post, not Alice. I need to practice on their names

[Jope]

Ah yes, I forgot all about the pictures.

Nevertheless, grey is still not a colour component and he can still try the sliders one by one to see if the colours work as expected when he starts from black and tries them one primary at a time.

I see Toni already recommended this before.

Either Lisa or the DAC just below the RGB port is where I'd look in this case. Sounds really odd that some colour values make it through ok but others don't.

[Badders]

Originally Posted by Jope

Ah yes, I forgot all about the pictures.

Nevertheless, grey is still not a colour component and he can still try the sliders one by one to see if the colours work as expected when he starts from black and tries them one primary at a time.

I see Toni already recommended this before.

Either Lisa or the DAC just below the RGB port is where I'd look in this case. Sounds really odd that some colour values make it through ok but others don't.

Well, I like to be different

Pete.

[Toni Wilen]

I remembered something similar that happened to my A1200 long time ago..

RGB output had yellow tint, composite output didn't work at all.

I poked around randomly and got normal RGB image by shorting two composite encoder chip pins (I don't remember which ones, I think it grounded chip's clock input or something) I guess the damaged chip can pull up/down DAC's RGB output lines.

[FOL]

Originally Posted by Badders

Well, I like to be different

Pete.

Too many people commenting now, its getting confusing.

I had already said if problem is on RGB and you have nothing on Compsite.
Then its 90% to be the BT101 (Video DAC). Else it will be before that.

Oviously we are all guessing, cause we dont have the hardware infront of us, only a manual.

[Photon]

It's not confusing unless you read all the replies... ;p

If the DAC or Lisa did this, and the composite encoder works, you would get the same colors on composite. (He doesn't, he gets nothing. ok.)

I can't account for palette colors going primary, but even if Badders doesn't tell, my guess is that colors can be changed smoothly, but they all seem to slide smoothly along the same hue range (blue/red, only way to leave it is to go all 15 or all 0. Is that correct?)

A digital RGB problem after Lisa is ruled out, because 15 15 15 and 0 0 0 show correctly. If we raise RGB to 170, suddenly green WB bg is purple! In Transplant, a black background is less red and darker purple. This means that RGB is wrong inside Lisa, incorrectly converted by the DAC in a way not relatable to any RGB component, yet predictable (right? same wrong colors in the same game?), or something after the DAC messes up the RGB going to scart.


To remove the composite encoder (marked "U12") from the equation, you can unsolder one leg of caps C211, C212, C213 (or just one cap) and see if the picture changes at all. If it doesn't, we can move to the DAC and then Lisa. If it does, the picture might go back to normal if you unsolder all three legs.


As a bonus, you get a test: If you suddenly get no picture via scart when you unsolder all three capacitor legs, your TV is default-switching to use the composite video signal in the cable, not RGB.

Limited hue range points to both chroma and an existing problem in the machine, this is why I think this is worth testing.

[Edit: the caps were not very easy to find, they seem to be on the underside --PICTURE--. And of course they're SMD, so temporarily unsoldering them completely works - but I'd lift legs 2,3,4 of U12 instead.]

[Badders]

Well, new board arrived this morning. Installed and Workbench is gray, space is black and all is well with the world again. Now for WHDLoad!

Thanks to all who offered advice. I'll take a really close look at the faulty board now that it has been removed.

Pete.