Atari 800XL Colours (PAL)

[TEBL]

Have been fiddling with an Atari 800XL I bought a couple of months ago from someone in England, so it's a revision D PAL unit that I'm using over composite (from retrocomputershack) on my LCD (no space for the old CRT thingies) with one of Lothareks PSUs. Have been looking at the posting on http://atariage.com/forums/topic/230225-colors-ntsc-vs-pal/to help debug some color issues I'm having, essentially the wrong colors substituted in the middle of the spectrum. I can adjust R38 back and forth and it'll either go totally out of whack or B/W, don't have a scope though so I've tried using small increments. The closest I get on the startup screen is what I can only refer to as DOS blue though my XLF, extreme jailbars aside, looks more like teal.

 

I've replaced all 4146 RAM chips (mostly dead MT4246 originally), GTIA, ANTIC, 4050 with CD4050BE, PAL crystal, 2N3904 transistors as well as all electrolytics without any significant changes to the colors. Picture below is from SALT 2.05 running via a SIDE2 cartridge, have also tried using a basic script to show all colors and hues where I'm seeing the same "repeat" of colors in the middle. Anybody got ideas any ideas on how to explain this?

 

 

NB! LCD works with ZX Spectrum, Acorn Electron, Commodore 64, Commodore 128, Dragon 32, Atari 520STFM, Amiga 500/600/1200 with some duplicates (which is probably why I've got storage problems). Never seen anything like it so I think it's something to do with the circuit.

[TEBL]

Think I'm in the right place atleast, I believe these resistors in conjunction with the transistors create the colour part of the signal. I've tried mapping out where things go according to the schematic while tracing stuff based on continuity so some passive components may be drawn in in the wrong orientation. Can someone with access to a revision D do some simple resistor measurements from the pins on transistors Q2, Q4 and Q5 to +5V and ground, so I have something to compare it too?

[Rybags]

The phase delay aspect of colour appears as a square wave from the COL pin of GTIA. External circuitry converts it to the approximated sinewave.

 

Supposedly the phi1 signal from the 6502 is also used to help provide the necessary voltage - refer recent thread http://atariage.com/forums/topic/276433-greenish-600xl-it-might-be-missing-o1/?do=findComment&comment=3982443

[TEBL]

The phase delay aspect of colour appears as a square wave from the COL pin of GTIA. External circuitry converts it to the approximated sinewave.

 

Supposedly the phi1 signal from the 6502 is also used to help provide the necessary voltage - refer recent thread http://atariage.com/forums/topic/276433-greenish-600xl-it-might-be-missing-o1/?do=findComment&comment=3982443

 

Thanks, don't think I understood the conclusion of that thread unless it was as easy as a break between 6502 pin 3 and capacitor C44, but that at least gives me a new avenue of things to try out in relations to the area around R38 - so far I've only measured and tried replacing the actual pot without looking so much at the nearby circuit. Times like these I wish I had space for an oscilloscope, but gotta check the traces first :-)

[TEBL]

Hmm... Have gone through that entire section of the PCB without finding anything not matching the schematic, it does however look strange that colours appearantly repeat instead of differing - whatever I adjust it to through R38 it doesn't seem to change the fact that I have two sections of the spectrum that end up the same. Spent an evening stabbing at it with different implements. Regardless I still have no clue as to how the various colors are generated and at what point they could possibly end up the same.

 

Aah, dammit... all this Atari stuff is making me miss my C64, and I'm posting more or less to myself so time for bed.

[TEBL]

Looking at the board, according to the schematics I'm supposed to have 33k resistor at R45. The one on there has color bands indicating 270k though I can't seem to get the multimeter to get a reading on it.

 

Is this a fault in schematics and service manual, or was it changed? Hmm... strange that I can't measure it.

[TEBL]

Well, still have no idea which part fixed it since I more or replaced everything in the voltage doubler circuit that I had access to so it might have been one of the capacitors, but still assuming the fix was replacing R45 with a 47K since that was the closest I had on hand. Looks like crap with all those monolithics and a ceramic bent every which way to fit the holes, but it works and hopefully I can get it back into it's case before deciding to do anymore with it. What do you guys do with unobtainiums like these caps?

 

Still, anyone have the ability to check the proper value of R45 and an opinion on wether I should leave the 47K in there?

Edited March 21, 2018 by TEBL