11 May
2023
11 May
'23
10:47 p.m.
On 2023-May-11, at 7:41 PM, Cameron Kaiser via cctalk wrote:
it's actually an artifact of the monitor that the
upper 6 were clear. Actually,
the stuck bit is entirely bit 2 (i.e., it goes
0 1 2 3 4 5 6 7 8 9 a b c d e f
0 1 2 3 0 1 2 3 8 9 a b 8 9 a b
and the high nybble is OK). Now that sounds more like a bad RAM chip, but why
would it be *just* those addresses? Does that sound like a plausible failure mode?
So I take it the KIM1 uses 6102 1K*1 RAM chips.
(I'm seeing some modern redrawings, but is there no original schematic online?)
Any RAM chip internally has the bits organized in a matrix.
I haven't found a proper 6102 datasheet, but the most likely array size is obviously
32 * 32.
Your bad address range of xx80::xx9F is a span of 32 on base-32 boundaries.
So a failed internal driver or sense-amp for one row or column of one chip would produce
your fault.
That's a pretty plausible failure mode.
Don't know how much luck you'll have finding a 6102.
Cursorily it looks like it's just a CMOS version of the 2102, and pin-compatible, so
that might be an alternative.
The 2102 comes in low-power flavors (often but not always called 21L02).
It also comes in various speeds so you might have to watch that, something like a 21L02-3
might do.