Galaxian bad ram 1 unusual

[Belike]

I have a Namco Galaxian pcb with an error message "bad ram 1" on screen.
As I've read, it has to be a bad 2114 ram at 7n or 7p.
I swapped these rams with a working Galaxian pcb, both works fine.
Checked all the connections, solder joints, ic socket pins, everything looks ok.
Any idea what else could be wrong?
Many thanks for any help in advance.

[level42]

Hey Béla.....good to see you here buddy !

Maybe a buffer chip between the processor and the RAMs ? Just wild-guessing here... (7D, 7E,8D)

[Belike]

Thanks for the tip my Friend, I try to check those.

[level42]

And........any luck ?


Sorry, found back this thread because I'm working on the Namco Galaxian PCB now.

[Belike]

And........any luck ?


Sorry, found back this thread because I'm working on the Namco Galaxian PCB now.

Unfortunately not, but the cab is in good hands by now, I feel like this problem will be solved soon.

[nilfisk]

Hey Bela,

I've finally found the time to spend a few hours on the PCB. The "Bad Ram 1" was caused by a broken trace between pin 5 of the 7408 at 2N, and pin 15 of the 74138 at 8M. I've discovered this more or less by accident, when following all the relevant traces shown in the schematics with my DMM. With this trace patched, the board now boots and runs for a few minutes. After that, the 2100 rams (sprite rams) get quite hot, and the graphics start to become distorted again. I've ordered new rams today, let's see if these fix the problems.

There were a number of other issues on the PCB as well, i.e. a rotten 7486 at 3B, causing the "Bad Ram 1" error to be shown twice on the screen . I wans't even hunting this one yet, but the ICs legs were so rotten that it literary fell off the board when I touched it

Cheers,

Martin

[Belike]

Nice to hear that the pcb is finally in good hands.
I bought this one from Olly (Alpha1), it was fully working until the "rack sound" became dissorted.
Hope to see it running again in your nice original Namco Galaxian.

[nilfisk]

Unfortunately, fixing the "Bad Ram 1" error only brought up the next problem:

While letting the game run for a while with the test rom, it started to freeze. After a few hours of cooldown, things were back to normal. However, everytime I let it run for more than a few minutes, the freezes returned. And not only did they return, they got more and more frequent.

I checked around with my scope, and what happens, is that the 6 Mhz and Clock signals get more and more noisy, until they more or less disappear - with this, there's obviously no video sync, and the CPU stops running properly.

Tracing the signals starting from the crystal, here's what I found:

- The crystal is pulsing fine all the time
- Signals look ok until they leave the 74368 at 1E
- The outputs of the 74107 at 1D are very noisy already, even though the IC is known good (I swapped in a new one, as I did with the 74368)
- Just for the heck of it, I also swapped the other ICs involved in the signal handling here (the 7474 at 1C, the 7420 at 3C), so they're also known good

I suppose there's something along the bus delivering the CLK signal that's creating that noise. My initial suspects where the 74161 at 3A and 4A, but I've swapped these as well with new ones, with no change.

As an added riddle: as I can leave the PCB off for a few hours, and things will work again for a few minutes then, I thought of heat related issues with one of the ICs. However, none of them gets hot, and also using freeze spray on any IC that's slightly warm does not change a thing - only powering down the game for a few hours restores a working state for a very short time.

Any hints from the experts out there are obviously very welcome

Cheers,
Martin

[level42]

Do all the RAMs on the bottom right edge have heat sinks ?

[nilfisk]

Do all the RAMs on the bottom right edge have heat sinks ?

You are referring to the 2700s on 1N - 1T, right? No, they don't.

That's how the board came, and as it was working before the RAM error occured, I supposed(d) it's fine that way. I did suspect these as well, though (see my first post), and will hopefully receive a new set of 2701s from Andreas / Speleo next week (2701s should work fine as well here)

Martin

[Laszo]

I would suspect a cap. It starts after a few minutes, the noice gets worse and you have to wait for a few hours....

[level42]

That's a good possibility too....

No I meant the RAMs at row 1, 5 of them. They get pretty hot on the Namco board, the Midway board seems to either have cooler RAMs or have much better heatsinks as they are way cooler.

[nilfisk]

You are referring to the 2700s on 1N - 1T, right? No, they don't.

Yeah, me too ... these are the RAMs in row 1, next to the character eproms. Or am I missing something?

Reg. leaking caps: yes, I was thinking about this, too, but from a quick look into the schematics, the only cap directly involved here is the small ceramic cap right after the crystal - and this one is definitely good. Also, the 5V line itself looks steady and good on my scope.

 But I'll check again today - many thanks for the hint, though!

[Belike]

The pcb worked fine for hours before the bad ram 1 error appeard, the only problem was that dissorted rack noise.
Hope you can fix this one soon!

[nilfisk]

While waiting for the beformentioned RAMs: does anyone know what's the correct way to disable the watchdog on Galaxian? I've read about two solder pads near the program rams that need to be connected for this, but the troubleshooting manual doesn't say anything on that.

I suppose it's not sufficient to just lift pin 26 of the CPU, and then ground it to put the Z80 into a reset/initial state, right? As the reset line should be pulled high after that, from what i see.

Thanks again for your inputs!

Martin