More on the Compupro - test-drb@ccmp.vtda.org

24 Jun 1999

-----Original Message-----
From: allisonp(a)world.std.com &lt;allisonp(a)world.std.com&gt;
To: Discussion re-collecting of classic computers
&lt;classiccmp(a)u.washington.edu&gt;
Date: Thursday, June 24, 1999 6:28 AM
Subject: Re: More on the Compupro
...

On Thu, 24 Jun 1999, Richard Erlacher wrote:

> I've been watching this thread with some interest, as I'm curious how
large
...
 > a board with 512Kbytes in 4116's would be. 
I've got some boards downstairs
...
   which have
sites for 288 4116's on each.  That's 576KB, though it's
 organized as 64-bit words with off-board SECDED which makes it 512 KBytes 
The board in question obviously has 4164s on it or it's only 1/4 the size
indicated.
 That's what I concluded as well, though the originator of this thread has
said at least once they were 4116's.  I'e got an 8MB "ramdisk" board I
built
back in about '85 when the 256k's were new.  This board is populated with a
few LSI samples I got, together with a 4MBx18 section literally hacked (with
a bandsaw) from a client's application circuit and tacked to the S-100
board.  The packages were the "zip" package, which allowed for pretty dense
layout.  If I ever find this circuit again, I intend to rework the logic,
since the memory is actually organized 1MBx72 bits, and I have a
now-obsolete LSI from TI, the 74AS6364A, which is a gigantic LSI used to
perform SECDED.  Since it corrects and propagates in only 10 ns, I want to
build in a hardware error scrubbing circuit and maybe a battery backup
scheme.  The board's still more than half empty, since the memory is on two
mezzanine boards. - Dick
...

Compupro made a M-drive, a 512k ram disk using 4164 devices.  I have one
of these and it works quite well.

Allison