Apple 1 weird crash

2 posts / 0 new
Last post
Offline
Last seen: 8 hours 14 min ago
Joined: Oct 18 2022 - 08:43
Posts: 29
Apple 1 weird crash

Hi,

I'm working on my second Apple 1, and I got a weird behavior: 

First, I was having the boot screen blinking correctly but @ was replaced with D and the whitespace was replaced with $. When I shorted clearscreen, I got still the @ replaced with D (blinking) and all the rest of the screen was all $ (instead of whitespaces, still). I started touching all IC one by one to feel if one was getting hot, and when I looked up, the screen was all black. 

Now when I boot it, all I have is a screen with no character at all, and CS does nothing.

I think it went black when I touched around the 7402 (on C10) but again, not sure. And I tried to replace it with no difference.

So before switching ICs like crasy, I checked the power lines. Everything is still fine. 

Anyone ever got something similar ?

 

Thank you in advance.

Offline
Last seen: 4 days 15 hours ago
Joined: Apr 1 2020 - 16:46
Posts: 996
If a motherboard reacts to "touching" of ICs, then ...

... you have an intermittent contact issue. Which can be a bad IC socket, or a bad solder joint, or a hairline crack in the PCB, or a bond wire issue within an IC, but the latter is a very rare occurence. Bad IC sockets and bad solder joints happen more often. Especially if the "original" TI low profile sockets were used, which were notorious even when new, 50 years ago, and did not get better with age, I'd suspect them first.

 

There is yet another possibility of intermittent contact, and this is a bent IC pin which never entered the contact of the socket in a proper way. This bent contact issue also can happen with IC socket pins, so they never entered the PCB thru hole and never were actually soldered. The latter should never happen with "machined contact" sockets (their pins don't bend easily) but it may happen with "stamped contact" sockets.

 

These intermittent contact issues are difficult and time consuming to track down. I have built fifteen Apple-1 clones, and one had such an issue, which took me more than a weekend to find. I first work with cold spray to see it's a bad IC, and unfortunately, the IC which reacted to the cold spray wasn't the culprit. Instead, it was a bent "stamped contact" socket pin in the row above the "bad" IC, which turned out to be "good". These faults should not happen when paying attention to details while hand soldering (there would be a missing pin on one socket) but this particular example came out of a wave soldered production run, so it wasn't me who planted the fault.

 

If you can see the reaction of the intermittent contact on screen, you may be able to deduct in which section of the circuit the fault must be. If CLR SCREEN does not work properly, but the signal path from the keyboard connector into the logic is OK, then it's worth looking at all the ICs in the cursor state machine, which is the cluster in the right hand side lower corner of the "Terminal Section" schematic. But if characters on the screen change shape you can deduct, from the two ASCII codes, which bit(s) in the video data path have changed and then follow these bits along the data path, beginning with the 74157 multiplexers, the 2519 and the 2513. Any IC (or its sockets / pins / PCB traces) in the data path processing this bit may be the culprit. But be aware that the cursor state machine controls the 74157 multiplexers, so if it spits erratic signals into the video data path, then erratic characters may appear, too. The exact behaviour of the shape shifting characters also is important - it the shape shift comes and goes, then it's not a fault in the 2504 storage loops, but downstream from them, beginning with the 2519. But if characters shape shift and persist in that state, it's in the 2504 storage loops or in the 74157 muxes or  in their control signals from the cursor state machine. Faulty two phase clocks from the DS0025 can also cause issues with the 2504 storage loops.

 

So you can see, the "Terminal Section" is not a trivial circuit, because everything is working together in the most intricate ways, and any fault can lead to the most bizzarre behaviour. I once had one case where characters on the screen were moving in sparse vertical columns (like in the "MATRIX" movie). Weird !

 

Good luck with finding the fault !

 

- Uncle Bernie

 

 

Log in or register to post comments