More broken Apples...

[Tony Duell]

> 
> Unfortunately, the other two weren't quite so cooperative. I decided to 
> look at the other clone first, as its behaviour was pretty interesting. 
> The screen was entirely full of apostrophes to begin with, but randomly, 
> blocks of them would change to lower case 'p's, and back, flickering 

IIRC the Apple ][ text display was upper case only. The fact that your 
clone does lower case means there must be hardware differences between it 
and a real Apple. Which means, alas, the schematics for the latter aren't 
going to be a lot of use.

> very quickly. It responded to a reset by changing the pattern of ps, 
> though they tended to appear in the same place. I found the ASCII values 
> interesting:
> 
> ' = 0x60 = 0110 0000
> p = 0x70 = 0111 0000
> 
> So, on reset, perhaps it's clearing half the bits per byte, and the 
> other four have a problem. Reseting the machine tended to lead to random 
> behaviour for a bit, such as random display changes, and speaker clicks. 
> At one point, the display switched to high-res mode, and I could see 
> that a large amount of memory had the same sort of pattern through it, 
> and was flickering the same way text was; I guess the entire memory 
> space is like that. Perhaps bits 5 and 6 are permanently stuck.

I would have expected ti to clear the screen to spaces at startup. That 
_may_ be 0010 0000 if the machine uses true ASCII codes in the video 
memeory. So maybe just bit 6 is playing up. 

As others have suggested, reseat all the socketed ICs. If in doubt, 
replace the sockets!. Then, assunming this thing uses 1 bit wide DRAMs 
(4116s, or similar), change the arounds. See ig you can make the stuck 
bit move somehwere else. If so, you know it's a RAM chip problem. 

-tony