The extra hardware for patching consisted of an FPLA (Field-Programmable Logic Array) and a PROM. The FPLA monitored the address bus, and on certain ranges it rerouted Chip Select from the mask roms to the patch rom. From the processor¡¯s point of view, this replaced a couple of instructions with a jump that led to replacement code in the PROM, which did its thing then jumped back to the mask ROM.
This scheme was used in 1980 in the 4024/4025/4027 terminals made by the Information Display Division at the Wilsonville plant. Later models used EPROMs, 2732 and up if I recall correctly.
Dave Wise
IDD 1980-1995
From:
[email protected] <
[email protected]> On Behalf Of Raymond Domp Frank via groups.io
Sent: Tuesday, April 05, 2022 8:08 PM
To:
[email protected]Subject: Re: [TekScopes] (7854) A Drop-In Replacement for the MCM68766 EPROM
On Wed, Apr 6, 2022 at 04:41 AM, Carl Hallberg wrote:
The MCM68766 can always be replaced by a 2764 using a 28 dip socket to 24 dip
header. I don't have a 7854 and didn't look up the specs, but need to check
the access time.
Physical space is the problem here, not electrical specs; the four 24-pin ROMs are mounted end-to-end lengthwise with only 0.2" in between, which doesn't allow for the longer 28-pin 2764's. on standard converting headers. The two rows of two each that make up the total of four ROMs are 0.3" apart. There would be enough space if one built a flat plug-in piggy-back PCB and mounted the replacement EPROMs without sockets.
Standard MCM68766's (Ta <= 450ns) will do the job. I used them in one 7854. The bad IC's are not EPROMs (as Ram says) but mask ROMs, made by Mostek. Some 'scopes are equipped with Motorola mask ROMs. Those don't suffer from "ROM-rot". BTW, the board contains extra hardware that allows patching of data in address ranges read from the ROMs, probably because it was considered too slow and expensive to run new masks if a change was made.
Raymond
