From: Zump <zump@rom.org>
   Date: Thu, 21 Dec 1995 00:49:27 -0800 (PST)
   Cc: port-sun3@NetBSD.ORG
   MIME-Version: 1.0
   Content-Type: text/plain; charset=US-ASCII
   Content-Transfer-Encoding: 7bit
   Sender: owner-port-sun3@NetBSD.ORG
   Precedence: list
   X-Loop: port-sun3@NetBSD.ORG

   Tom Samplonius wrote:
   > 
   > 
   > On Wed, 20 Dec 1995, Mauricio Tavares wrote:
   > 
   > > 	I wonder:  if one were to take those soldered RAM chips and 
   > > replace them by sockets and larger capacity RAMs, would the 3/50 
   > > acknowldge the improvement?
   > 
   >   Would it be worth it?  A 3/60 is really cheap, takes standard 30pin 
   > simms, and is faster than a 3.50.
   > 

   No, but it would be really cool if a new PAL could enable the 3/60
   to take say 4M simms.  Dunno if the address lines for that are connected.

Looking at the board, it would appear not (although short of dismembering it
it's hard to tell).

Going on the number of interesting parts in the memory section of the 3/60
board (the HAL20L8 and the cluster of IDT 6168s in particular), I suspect
that the 'new PAL' would be a considerable amount of work without access
to the original schematic and equations.

Now, if Sun were to make this sort of information public (a gesture of
goodwill and patronage to the many 68K designers around), such an
upgrade would become a _little_ easier.  (Replicating the original 
timing with currently available parts would be nontrivial)

	-Z,

Mike