Macintosh 128K help request

[mrkirkby]

Hello all,

I have a 128K logic board. It's been upgraded to 512K with the modification and new sockets for replacement ram chips installed. I did at one point see this board working but it failed rather suddenly. Not the smoke this time but it makes some rather loud beeping noises on repeat which seem to line up with the video glitching.

It's strange. The rom does seem to be active and CPU does seem to be trying to execute something. The screen state changes in a somewhat similar pattern on start. Pressing reset will repeat that process. It looks as if some of the RAM addressing circuitry has gone bad but I've probed rather a lot to no end so far.

Thought I'd see if anyone has found anything similar. I do have a donor board fortunately but some of the parts are already missing. Here's a video of what it's doing.


View attachment IMG_9538.MOV

I've tried swapping ram around to see if anything changes. Sometimes that raster turns to more of a checkerboard.

Any thoughts on this are appreciated. I might have missed something while probing but if anyones seen anything similar, it might help me to figure out where to poke. I have a ram tester coming today although voltages and data look ok on what I've seen so far. I'd also expect something a bit different if ram isn't ok. Larry's books seem to show other fault indicators for that failure mode.

 

[Byrd]

I had a very similar output when I replaced my 128K 16 x 4164 DRAM with sockets; wouldn't say it was an enjoyable job and found two shorts when I checked it over. The DRAM tester will also be your friend.

 

[mrkirkby]

That's good to hear. I haven't toned out the links between sockets yet but that sounds like a good thing to do as a next step.

It wasn't myself that did the work and given the perf board used for the additional upgrade logic, it looks like it's been done a long time ago.

I think I did pull all of the ram off the board to see what it did ram-less and it was a sort of similar output so a short could well be the case.

 

[Byrd]

Check around the ground pins on the socket, which were very hard to desolder I found. In my case I needed to run a desolder braid over some of these and reflow. Not my most elegant work - did a couple of months back - but another 128K lives and sockets are great for inevitable "bit rot" of DRAM along the way.

 

[mrkirkby]

Had a really good tone through everything and at least all of the common connections are linked well. I couldn't find any shorts between pins to ground which while reassuring takes away an easy fix. I need to work back to the multiplexing logic now I think. The ram tester turned up and told me all was ok so that's out too.

The mystery continues!

 

[cheesestraws]

What manufacturer are the DRAM chips in the machine? I've had a very bad track record with Micron DRAM of that period, it seems to self-destruct.

 

[mrkirkby]

They're the Toshiba ones. They all passed a fairly robust test in the tester showing <110ns each and no shorts. Not that those tests are definitive but I believe it does the industry standard checks. I have swapped out all of them with two other chips too rotating them all around so I'm fairly confident it's something else.

I think I need to spend a bit more time with it powered on with the oscilloscope as my guide. I'd like to trace the ram logic all the way back to the CPU though as it seems to be pointing toward that path.

Hello btw! @Durosity and I were discussing your BGA reflow work the other day. I dipped my toes a while back, got scared and quit!