I’d never intended to work on one of these PCB’s. But a request came in and it is a very close cousin to Gyruss which I have repaired a number of times.

Board #1 – In for repair

Triage:

• Board reported to have graphics display issues

Once I got it all connected up:

Notice the characters are bad, but the star field seems very clean.. I figured it was a logic issue when I first saw it.

I did spend some time creating a FPGA Catbox config to test all the RAM and ROM. It took a bit of time, but that all checked out. After that I spent time trying to figure out how the board worked and tracing the video signals. Nothing was really jumping out at me..

At one point I powered on and the video was ok(ish).. but then started to deteriorate over the course of a few minutes.. hmm.. Thermal? I broke out the can of freeze spray and hit a few chips..

Got to the 74LS240@A2 and everything cleared immediately! The LS240 just happens to control RGB outputs. Could it be that simple? (turns out no) Up until this point I had not run into 240’s – they must not be commonly used on Atari boards where I spend most of my time. So I ordered some. Popped in the replacement a week later – same exact issue.

Just for fun.. I hit the new one with the freeze spray and it cleared it up too.. really?!

At this point I checked everything in this area – resistors and all the ceramic caps.. Everything seemed fine. The 555 timer @D3 had an odd looking signal coming from it so I replaced it.. But that was not it.

Now I focused on these chips.. checking inputs, outputs, etc. All of the issues were here as it turned out.

Finally I found what was happening – Here is the video sync signal going into the LS367@C1 Pin 14 and the same signal coming out Pin 13. There are resistors there changing the output a bit.. But the key is that it is not getting to zero volts. Replaced the chip – same output.. hmm…

Then I determined the ground pin was floating at .8v when I measured it.. I thought maybe a leaky cap and pulled off the nearby ceramic caps, etc. There were only three chips that registered the floating ground and just over 4vdc on the vcc pin. At least I’m down to a small area on the PCB. None of that stuff checked out or made sense.

Fast forward a bit – I checked this every way and came to this conclusion: I found a design flaw in this PCB

The wide trace in the red outline is the ground plane that connects LS367@C1, 555@F0, LS04@E0. All 3 had the voltage issue. All of the ceramic caps and a couple of electrolytics around it were good. The problem is – this ground plane isn’t attached to the ground plane of the board. There are no traces on the parts side that were cut or missing – they never connected it. I connected a ground pin from this area to another on the actual ground plane with a mini-grabber during testing and fixed it all. Clean signals on the scope, vcc up at 5vdc where it belongs and the ground at 0v where it belongs.

I’m not an EE – but maybe the ceramics drifted over time and the video got progressively worse? The actual fix is the small jumper wire to connect the ground planes. Those three chip’s voltages all snapped into line and problem solved. The LS240@A2 that I initially hit with the freeze spray was outside all of this – maybe the cold was just enough to change the voltage/current or something to clean up the video? Not sure.

I was expecting some weird analog passive component issue – but this is a design issue IMHO. Once I found it and knew what to search for, there are a number of references to this game with graphics glitches and lines through the graphics.. Or just a star field and no graphics.. But I didn’t see that anyone to date had found the answer – this is it – add a ground wire.

Board works!