Unboxing and teardown of non-working Atari 5200

[@danielblack6529]

Looking at your 2-port 5200: The pin that's lifted on the big bottom right chip, the GTIA, is pin 11, which is T3 (Trigger 3). Doing a little research on it: On early 4-port systems, T3 normally goes to pin 13 of the 4th controller port which is for the top left/right red fire button and pin 4 (LP/ Light Pen) of the ANTIC (big chip on the top right) is only pulled high with a 4.7Kohm resistor to 5V (that'll be relevant in a bit). On later 4-port systems, T3 is also connected to pin 4 (LP/ Light Pen) of the ANTIC (big chip on the top right) and on 2-port systems it's only connected to LP. So theoretically, pin 11 of the GTIA a.k.a. T3 not being connected to the board may not do anything on a 2-port system, except leaving the LP pin floating and no light pen support if there ever was one for the 5200 (there was one for the Atari 8-bit computers IIRC). The bodge wire on the bottom connects pin 1 (GND) and pin 12 (P2/Paddle 2) of the POKEY (big bottom left chip). P2 leads to pin 10 of the 2nd controller port and controls one of the axis of the analog stick, so (assuming this is just a wire and not an insulated resistor) why is this tied to ground...? As for the bodge wires on the left: I'm not sure what those are. Maybe those are trace fixes? Btw. did you catch the scratch on the board between the chip A2 (big top right, ANTIC)? Also, did you insert that Congo Bongo cartridge the right way? The label on the cartridge was upside down when you inserted it and turned it on. Either way, thanks for uploading this video and I hope you read both of my comments and find them at least somewhat helpful.

[@danielblack6529]

I recently got an original 4 port 5200 sold as untested without the switchbox that does the same thing after I did the switchbox elimination mod: Blackish screen and audio noise coming from the POKEY when I try to play Pole Position. I reseated all the big chips (but not the RAM yet) and both 7805s checked good with a game inserted and it still doesn't boot. Mine has the Rockwell 6502 R-C Network fix likely from the factory, a cap installed on the POKEY chip (the big chip on the bottom left) between pin 1 and 17 (GND and VCC respectively) and on the solder side of the main board, there's a bodge wire between pin 1 (GND) of the POKEY is connected to the GND side of C104.

[@new_oldstock]

I did reseat all the chips (not sure if I showed it here or not). I found a copy of the 5200 Field Service Manual and I ordered some new RAM chips. I'll revisit it soon and hopefully get it figured out.

[@danielblack6529]

​@@new_oldstockGood idea to replace the RAM, because on mine, the culprit that kept the system from booting was the seventh DRAM chip from the top, U24, (XA24 on later boards). The chip itself looks like it got really hot at some point as the die marks are visible on the chip, while the label is barely readable. I found it to be not working by testing each RAM chip, one at a time, on my SG-1000 II (the one with the discrete TMS chip), since it also uses 4516 RAM for it's VRAM. Testing the 5200 temporarily with a RAM chip from my SG-1K II, it did boot and work, but the graphics were blocky, 2 color only, dark/missing detail and the color hue was wrong. The blocky 2 color only graphics were caused by pin 40 of the GTIA (Address line 2) not being connected to the board, because the pin was a bit corroded, which I could fix and the color hue was also fixed by adjusting R8, but the dark/ missing detail graphics couldn't be fixed. It turns out my GTIA is a bit faulty as Luma Output 0, which is pin 31, only outputs a floating signal (~1V) instead of a proper signal (0V - 5V), even when it's not connected to the board, so I need to either replace it or get the Sophia, whenever that is available. TL;DR: One RAM chip was bad and aside from the GTIA being a bit faulty, the system is working fine now.

[@danielblack6529]

Update: I recently got the Sophia 2 for my 5200 and on the RF output it still looked wrong like with its original GTIA. After lots of probing and testing it turns out that the GTIA was fine and that Luma Output 0 wasn't the problem, however I did discover that Luma Output 3 wasn't on the final output circuit, even though said output is fine from the chip itself. So what was going on? It was the CD4050 hex inverter U1 (XA1 on later hoards) whose 4th inverter didn't work right that caused Luma Output 3 not to show up. After desoldering that and stealing a CD4050 from one of my Atari 2600s to put it on my 5200, the video output looks exactly like it should, even with the original GTIA. TL;DR: The GTIA wasn't faulty, the Hex Inverter was partially broken causing the video output to look wrong.