RE: FLASH ROM replacement

RE: FLASH ROM replacement

From: M H Stein <dm561_at_torfree.net>
Date: Wed, 6 May 2009 05:55:12 -0400
Message-ID: <01C9CE0F.3A035CE0@host-208-72-122-51.dyn.295.ca>
----------
From: 	Jim Brain[SMTP:brain@jbrain.com]
Sent: 	Tuesday, May 05, 2009 11:09 PM
To: 	cbm-hackers@ling.gu.se
Subject: 	Re: FLASH ROM replacement

Jim Brain wrote:
> M H Stein wrote:
>> Any chance you could also put the two rows of pins for a 6540 
>> replacement
>> on there while you're at it, to use either a 2732 or 1/2 of a 2764?
Would this work:

http://www.jbrain.com/vicug/gallery/ROM-el/6540_ROM_el_PCB

I made some assumptions:

CS3,4,5 are active low
CS1,2 are active high
CLK is active high

I'd appreciate knowing if those are valid or invalid assumptions.

* As the schematic shows, the design uses a 74'138 to decode all of the 
CS lines and CLK into a single CS line for a 2764-27512
* I can add a jumper to allow 24 pin EPROM operation (hook pin 26 to pin 28)
* Extra address lines are pulled up to 5V and available on pads on top 
of adapter (to allow switchable ROMs)
* logic and resistors are hidden inside 28 pin header.

Let me know if there is any interest.

Jim
--------
Well, I can't even figure out how to print your schematic, so I'm probably
not the one to ask ;-). I kinda thought it'd be just another two rows of pads
beside the 24 pin pads, without any additional logic or pullups, but I 
hadn't really thought it through.

As far as I can tell, CS4 is used for the decoded 4K block chip select
to two chips in parallel, and A11 goes to CS3 for the low 2K and to CS1
for the upper 2K; CS2 is tied high, and CS5 seems to be NC (maybe
pulled low internally?), which all matches what you have for the CSs.

I thought that if one accepted replacing both the upper and lower 2KB 
with a 2732 or greater (except for the E ROM which could be a 2716) 
and used the lower socket of each pair, then all you'd need is pin 3 to 
/CS and jumpers for pin 4 to A11 and /OE to Vss or /CE.
For the E ROM connect pin 4 to /OE instead (The upper half of E is
memory-mapped I/O).

But it probably isn't that simple, and no doubt someone will point out
the flaw(s). Maybe there isn't really enough demand anyway to make it 
worth while, especially seeing there's one permanently on eBay...

What's the story on the fancy Welte board? Is it still available, either built 
or PCB and logic?

Maybe I'll just hand-wire something or kludge a 27128 and replace 'em
all while I'm at it... ;-)

mike




       Message was sent through the cbm-hackers mailing list
Received on 2009-05-06 12:06:18

Archive generated by hypermail 2.2.0.