Looking at this, is it safe to assume that the two EPROMs are NOT consecutive in memory?

If one were to replace these two 2516s with a single 2764, the data loaded into the 2764 would be image M-0 / blank / image M-1 / blank, correct?

 

Hi met,
A 2764 would use addresses A0 through to A11.
So you would have to modify the ROM select to A12

E

 

Looking at this, is it safe to assume that the two EPROMs are NOT consecutive in memory?

If one were to replace these two 2516s with a single 2764, the data loaded into the 2764 would be image M-0 / blank / image M-1 / blank, correct?

No, they are consecutive.
If A15 is enabling the LS139, then A12 and A13 select 1 of the 4 possible options. A13 being Low means A12 alone selects either Rom, depending on the state of A12.
You would need to AND the two outputs from the LS139 together to enable the 2764, and tie the extra address pin of the 2764 high or low to just use half its space!

 

Hi met,
A 2764 would use addresses A0 through to A11.
So you would have to modify the ROM select to A12

E

Doesn't a 2764 have address lines A0-A12?

Also, this board does NOT have anything attached to A11, but it would on the new-style board?

 

hi meter,
As these are ROM devices, they will most likely be loaded with the MCU operating code, are you planning to rewrite the code to suit the 2764 and program the 2764?
E

 

looking at your circuit, A0-A10 are connected in parallel so no difference here.
A11 is not used.
A12,A13 are used as chip select (decoded by LS139).
so it is up to you to decide how you want merge them:
1. if you want to write content of two ROMs consecutively, then A12 of your circuit will need to be tied to A11 of your 2764.
2. if you do not want to do that, you will need to leave a gap and then connect A12 to A12.
3. if you are not sure or want to allow for both possibilities, you can write ROM2 twice (ROM1>ROM2>ROM2>whatever)

 

Hi Meter,
This is the way I would connect IF I was rewriting the MCU code.
A12 tied to 0v
E

 

Not necessarily looking to redesign, but to try and see how much I could figure out without needing the board physically in hand (as soon as I can find one for less than $200 WITH the original EPROM).

When the board I have at hand was redesigned, there were three changes made - consolidating the EPROMs, adding two Schottky diodes (likely in the serial data section), and a more robust drive section (my recent question about TO-126 transistors).

 

Final answer on the code: it was indeed one image / padding (FF) / second image / padding (FF). The code compare did not match up, which suggests either an addressing change or they tweaked the code as long as they were doing a board re-spin.

Connections for A11 and A12 check out (direct from the 2764 to the 8031), as do pins 1 and 27 on the 2764 (normally pulled up to 5V).

Will put up an addendum once I see how pins 20 and 22 are wired.

 

The connection at pin 20 on the 2716 carried over to pin 22 on the 2764.

The connection at pin 18 was changed to a wire-AND using a pair of Schottky diodes and a 10K pull-up, the ROM1 and ROM2 lines from the '139 forming the inputs.