13 Aug
2015
13 Aug
'15
2:40 p.m.
If you can write your own code, usually a simple write and read
sequence will find a dead RAM. Just write 0xFF to a location
and expect to read it back and then 0 and read it back.
You write it so that it will stop at the bad RAM location and
continually loop that location.
Probing with a scope on the chip enables will usually find the
byte that is failing. Then just look at the data pins to see which
chip does not match the others.
It helps to have a two channel scope so one can trigger on the
enable while the other looks at the data.
This is not a high quality manufacture type test but most RAMs
fail completely, making such a test useful.
Dwight
To: cctalk at classiccmp.org
From: thrashbarg at kaput.homeunix.org
Subject: Wang PC260 memory fault
Date: Thu, 13 Aug 2015 17:23:17 +0930
Hi List,
So a good friend of mine recently resurrected his childhood 286 Wang
PC260 after I suggested he replace the keyboard controller. In the 14 or
so years it's been unused, it has developed a memory fault in one of the
conventional memory banks. The memory is soldered to the board so we're
looking for anything that might assist in tracking down the dud memory IC.
He can get it to boot if the memory size is set to 256kB, so I'm
guessing the second 256kB bank is bad. Is there software that exists
that can identify individual dead or faulty memory IC's, or at least the
corresponding bit(s) which may not be working?
According to Dr. Google, this machine was built by Tandy for Wang.
Perhaps there's a Tandy diagnostic disk that'll work with it, or better
yet a Wang diagnostic disk.
(I'm wondering if this question has been asked/answered before too)
Cheers,
Alexis.